Award Recipients: 2024 Exploration

Introduction. Site-specific drug delivery aims at controlling the drug dose at a specific site in a tissue without causing systemic side effects. This type of drug delivery is especially important and challenging for applications that require very high dose of the drug at a small biological space of few hundreds of micrometers. For example, delivery of thrombolytic drugs at high doses with high spatiotemporal resolution is a potential thrombolytic treatment in stroke management.

Objectives. The main objective of this application is to design and develop a novel magnetically actuated drug delivery micro-vehicle for high-resolution site-specific drug delivery, and to demonstrate proof of in vivo efficacy in an animal model of vascular occlusion. The target microparticles are aimed to be fast swimmers under magnetic actuation, be capable of drug loading and burst release at the target site, and enable controlled mechanical stimulation of the target site for enhanced drug permeability.

Research Approach. The fast swimmer micro-vehicles will be made using a microfluidic approach. The asymmetric particles will be hollow Janus microparticles, made of a soft magnetically deformable polymeric membrane in one hemisphere and a brittle non-deformable polymeric membrane on the other hemisphere. The special design of the microparticles allow them to be precisely actuated and navigated in any direction under the influence of a time-varying magnetic field, which is controlled by an electromagnetic actuation system. These microparticles will be loaded with a thrombolytic drug and will be tested in a small animal model of vascular occlusion. The particles will be injected in the animal intravenously and will be guided to the blood clot site with magnetic actuation while being observed by ultrasound. The drug release will occur under magnetic as well as ultrasound actuation of the particles and the state of the blood clot will be monitored.

Significance. Ultra-high resolution site-specific drug delivery is a significant technological challenge that can have extensive applications in various branches of medicine. Our proposed strategy will demonstrate how this non-invasive treatment can replace invasive and high-risk thrombectomy surgeries, and offer a new treatment option for removing blood clots in the small arteries in the brain where surgical intervention is not possible.

Ovarian cancer is one of the most lethal gynecological malignancies, often diagnosed at an advanced stage due to its subtle symptoms. Despite advancements in treatment, the five-year survival rate remains dismally low, particularly for high-grade serous ovarian carcinoma, the most common subtype. The introduction of PARP inhibitors, such as Olaparib, has marked a significant breakthrough in the management of ovarian cancer. These inhibitors target tumors with defects in homologous recombination repair, such as those with BRCA mutations, and have been shown to significantly extend progression-free survival in these patients.

However, identifying patients who will benefit from PARP inhibitors requires genetic testing for homologous recombination deficiency (HRD), which is currently very expensive and not universally accessible. This creates a barrier to the widespread use of these life-saving therapies. To address this, innovative strategies are urgently needed to make HRD testing more accessible.

In the current study, we propose to develop a novel data-driven platform to determine HRD status through advanced analysis of routine H&E-stained histology slides. State-of-the-art explainable attention-guided deep-learning algorithms will be designed and developed as the analytical engine of the proposed platform. These algorithms analyze the input whole-slide digital histology images to discover associations between the cellular morphology, distribution, heterogeneity, and texture within the tumour microenvironment with the HRD status. The platform will be developed, optimized, and evaluated using digital histology images and the corresponding HRD information acquired from 300 patients from an established study of ovarian cancer in Ontario by our team, with available clinical information and surgical tumor specimens. The histology slides of surgical tumor specimens will be digitized at 40X. The gold-standard HRD status of each tumor will be determined using HRD Score genomic test, leading to a unique dataset required for the breakthrough analytical strategy proposed in this project.

The platform to be developed in this project is a potential game-changer for equitable access to effective treatments in ovarian cancer. It could provide a cost-effective, widely available alternative to current expensive testing methods, allowing more patients to be accurately selected for targeted therapy with PARP inhibitors, thereby improving survival outcomes in ovarian cancer.

We are developing Serpento, a revolutionary robotic endoscope addressing a longstanding challenge in minimally invasive surgery. Current tools are limited: laparoscopes offer precision but can't reach all anatomical sites or cause iatrogenic trauma, while bendable instruments lack stability for precise interventions. Serpento breaks this paradigm, uniquely combining deformability, dexterity, and stiffness, enabling unprecedented navigation and precision within the body. Our innovative mechanical design, inspired by dense and yet precise DNA packaging in the nucleus, utilizes a unique helical stacking approach to achieve full structural definition while maintaining extreme dexterity. This represents a significant leap forward, overcoming limitations that have persisted despite numerous attempts in the field. We now aim to develop equally groundbreaking actuators and electronics to fully realize Serpento's potential. This high-risk, high-reward project integrates the fields of advanced robotics, surgery, and community health. Our research focuses on four critical areas:

1. Actuator Innovation: Develop novel actuators to power the mechanical design's dexterity. 2. Advanced Control Systems: Create an intuitive interface integrating robotics and medical imaging. 3. Team Integration: Establish methods to seamlessly integrate surgeons' expertise into the day-to-day engineering process, enhancing the synergy between clinical insight and technical development. 4. Clinical Integration Strategies: In the design, account for procedures with high community health impact across diverse healthcare settings.

Our interdisciplinary team crosses tri-council boundaries, integrating engineering with surgical and health sciences expertise. This collaboration considers healthcare economics and equity, focusing on broadly applicable interventions. Success could revolutionize minimally invasive procedures, improving outcomes, reducing costs, and expanding access to advanced surgical techniques across diverse healthcare settings. We aim to redefine medical robotics through a fundamental novel concept that enables unprecedented flexibility and precision in surgical instruments. At the same time, we are developing innovative collaborative methods between surgeons, engineers, and equity experts. The challenges and high risks of this transformative project are met by a highly skilled team, well-prepared to address its technical and interdisciplinary complexities.

Prostate cancer (PCa) is now mostly diagnosed at a localized stage, when the cancer can be treated with curative intents. This can notably be achieved using the prostate-specific antigen (PSA), a highly sensitive PCa biomarker. However, PSA lacks specificity and has a high rate of false-negatives. More specific biomarkers are thus required to avoid unnecessary biopsies, clinical procedures, and anxiety in patients.

We suggest using the unique prostate metabolic program as a novel biomarker for PCa. Unlike other cells that have a tricarboxylic acid (TCA) cycle with citrate at its core, prostate cells exhibit a truncated TCA cycle that allows the production and secretion of citrate in semen. This distinctive citrate metabolism is reprogrammed in PCa, leading to a decrease in secreted citrate levels.

Our hypothesis is that a bioassay using semen samples will distinguish men with low citrate levels, indicative of PCa. We believe that this test, complementing PSA testing and the digital rectal exam, will significantly increase the specificity of PCa screening. Therefore, we propose two specific aims.

Aim 1. To engineer a new device for rapid point-of-care citrate measurement in semen. We will engineer a new device that automates the liquid handling processes required to dilute and filter semen samples prior to testing, and implement citrate detection in semen. We will optimize this device with a limit of detection and coefficient of variation to meet Health Canada reporting standards. We will then compare citrate measurements from this device with in-house mass spectrometry (MS) measurements.

Aim 2. Perform a prospective clinical study investigating the relationship between citrate levels in semen and the presence of PCa. We have already performed a retrospective pilot study that demonstrates the feasibility and acceptability of our approach. We will now perform a prospective clinical trial and quantify citrate in semen from patients with and with PCa, both using our newly developed device and MS. Embedded in this clinical study, patients will answer standardized questionnaires to further validate the acceptability of the current approach.

Overall, we aim to define a new paradigm in PCa screening and allow clinicians to avoid invasive unnecessary biopsies, maximizing their effort on patients with PCa, and increasing quality of life for patients without disease.

Opioid dependence and opioid-related deaths in Canada have risen exponentially in the last decade. In the process of rehabilitating patients struggling with opioid dependence, compliance with the therapeutic methods designed by the physician remains a challenge. Poor compliance can negatively impact a patient’s relationship with their physician and increase incidences of relapse and the attendant risk of overdose. In recent years, the rapid growth and uptake of mobile technologies have resulted in various applications developed to improve patient compliance. However, these have only yielded modest success rates and the majority share similar functions of manual reminder alerts and access to resources. Research has also found the negative potential of mobile apps to introduce new risks, for example, adding burden to patients’ daily lives or being perceived as an invasion of privacy. The objective of this project is to create a digital platform for patient-physician interaction that can enhance compliance without introducing intrusive aspects to therapy. The compliance app will utilize elements of user experience (UX) design to foster user engagement on the platform, particularly drawing on gamification techniques and variable reward algorithms. Patients will be given goals to pursue for rewards, with the goals and reward ratios ultimately being decided by the physician. The main strength of this approach involves the incorporation of both algorithms and UX design to target reinforcement learning in its users. By leveraging the same processes that contribute to the development of addiction, it can disrupt the cycle of reinforcement in opioid misuse and encourage compliance behaviour. The novelty of this app is that it operationalizes tangible, hedonic, and personalized rewards that can positively inspire patients to work in collaboration with their physicians. If found to be efficacious in future trials, the platform could be modified for use in fields outside of addiction, including but not limited to transplant patients, for whom medication compliance is critical. This is a novel attempt to develop a compliance management app and test its feasibility in preventing relapse among patients with opioid use disorder. This work is significant to address the growing rates of opioid overdose and difficulties achieving therapeutic compliance with the substance-using population.

We aim to develop MEMS (micro-electro-mechanical sensors) based photo-acoustic quantum biosensor for early detection of neuronal degeneration diseases, such as Alzheimer’s and Parkinson’s.

Functional lysosomes are crucial in neurons, which degrades incoming material and ridding of damaged cellular material. Accumulation of undigested material in the lysosomes, a common feature in nearly all neurodegenerative diseases, signals catabolic dysfunctions. As a result, proteins, phospholipids and cholesterol are crowded in lysosomes, leading to an increase in the viscosity. By measuring the viscosity of the lysosome, neuronal degeneration diseases could be detected at early stage.

We envision to develop a quantum sensor that (1) utilizes squeezed light in a dual-core-fiber photo-acoustic interferometer (PAI) with surface acoustic wave (SAW) excitation and (2) enables coherent detection of a single polarization mode. Our devices are: (i) intrinsically quantum, allowing unprecedented sensitivity by oversampling SAW at coherent optical frequency with low phase noise for minimum number of nanoparticles; (ii) much smaller in size with low power of mW, hence minimal disturbance for cells comparing with current technology using optical fluorescent detection. Our new approach stands at the frontier of innovation, leveraging acoustic pumping with below quantum noise optical detection, enables highly sensitive detection of nanoparticles and lysosomes. It also leads to the measurement of lysosome viscosity related to the photo-acoustic attenuation of PAI at SAW frequency, and a key indicator of lysosome fitness.

Our PAI-SAW-based quantum bio sensing technology aims to deliver revolutionary capabilities of gauging lysosome fitness with high detection sensitivity. Early detection of lysosomal dysfunction is of paramount importance for effective interventions, as well as boosting lysosomal functions as a viable therapeutic strategy to delay neurodegeneration and promote neuronal recovery.

This high-risk, high-reward project is driven by our multidisciplinary expertise in fiber photonics, disease biology, and quantum device engineering to develop proof-of-concepts in early detection of neuronal degeneration diseases. Our work will contribute significantly to the field of quantum biosensing, fostering collaboration across physics, biology, and medical research, and providing a rich, interdisciplinary training ground for the next generation of scientific leaders.

Objectives: Arts-based programs for neurodegenerative conditions such as Parkinson’s Disease (PD) provide holistic person-centered care that can reduce motor symptoms and improve quality of life (QoL). There is a need for appropriately powered studies to demonstrate efficacy and accelerate inclusion in medical treatment models (Gros et al., 2024). Integrating aspects of social prescribing with personalised treatment goals, our study aims to 1) Establish feasibility and acceptability of arts-based programs to support recently diagnosed persons with PD (PwPD); 2) Map program access accounting for referral pathway and personalization; 3) Implement recommended outcome measures increasing applicability of findings while investigating biological changes to help determine mechanisms; 4) Use a research creation process to amplify the voices and experiences of PwPD and carepartners.

Research approach: Phase 1- Newly diagnosed/suspected PwPD and carepartners recruited through support groups, specialist clinics, and community-based networks are directed to a portal for arts-based programs. Choice of program(s) and frequency of attendance will be at the discretion of participants. Psychosocial outcomes including Quality of Life Index for PwPD (QoL), loneliness, and emotional well-being will be collected online via Qualtrics; clinical outcomes will be collected in partnership with Canadian Open Parkinson’s Network (COPN). Phase 2 - Pre/post biometric assessments will explore biological mechanisms and outcomes. Phase 3 - A research creation process will explore arts participation and agency in relation to health outcomes and QoL. Other outcomes of interest include impacts of social prescribing and program type on health behaviour, and differences between online versus in-person formats.

Novelty and Significance: PD is the fastest growing neurodegenerative disease globally, and incidences are forecast to double current rates by 2040. Newly diagnosed and suspected PwPD often wait months or years to see a specialist. Changes in health behaviour associated with physical and social activities have been shown to improve outcomes and quality of life, and arts-based programs may offer another form of direct support for PwPD during this critical period. Our study aims to demonstrate efficacy, feasibility, and acceptability for arts-based programs supporting PwPD to facilitate uptake into medical models, improve patient care, and relieve burden on the healthcare system.

Purpose: Climate change poses challenges to interconnected Water, Energy, Food, Health, and Economy (WEFHE) systems. In regions like the Eastern Nile, where these systems are deeply intertwined, climate change impacts can be complex and far-reaching. Current approaches often fail to capture the full spectrum of these intertwined impacts, notably in the context of river basins, limiting our ability to develop effective adaptation strategies. This research seeks to develop an innovative interdisciplinary adaptive planning approach that encompasses the impacts of climate change across multiple sectors and systems, with the Eastern Nile Basin serving as an illustrative case study.

Objective: This project aims to create a cutting-edge integrated modeling approach that simulates climate change impacts on interlinked WEFHE systems. The Eastern Nile Basin has a unique combination of resource interdependencies, geopolitical complexities, and environmental vulnerabilities. It offers an ideal context for demonstrating how such an approach can be applied. To ensure that outputs will be used during and beyond the project by government and academia, the project will develop a Climate Awareness Dashboard (CAD), which will host the outputs of the interdisciplinary simulations.

Approach: The modeling framework will be driven by climate change projections and combine multiple interdisciplinary modeling tools. A hydrological model will capture changes in naturalized streamflow and other parameters of the hydrological cycle, while a river system infrastructure model will simulate hydropower generation and irrigation water supply. A hydrodynamic model will simulate flood inundation, and a public health model will capture associated changes in waterborne and related diseases such as malaria and cholera. A computable general equilibrium model will be used to assess economy-wide and distributional impacts across income groups. These components will be interlinked to capture the interactions and feedback loops across systems. AI-based search algorithms will be used to search for efficient climate adaptation plans.

Novelty/significance: The project’s novelty lies in its interdisciplinary modeling framework, which integrates sector-specific models to simulate a wide range of impacts and their interdependencies. The project will engage users in the region in the development of adaptation scenarios, and the CAD will enable seamless exploration of adaptation options.

Notre « cécité botanique » combinée à la perte de biodiversité par le réchauffement climatique est à l’origine d’une écoanxiété qui peut affecter 70% des jeunes québécois et expose aux maladies mentales. L’objectif du projet est de codévelopper et examiner les effets de deux interventions mettant en relation les végétaux et les humains, en faisant interagir les sciences fondamentales avec la philosophie, les cultures autochtone et allochtone et les arts. Ces interventions visent à réduire l’écoanxiété et à favoriser l’écorésilience et l’écoresponsabilité chez des adultes. Elles auront pour cadre (1) un parcours musical au Jardin botanique de Montréal (Québec) où les musiques diffusées seront composées à partir des signaux électriques des végétaux sonifiés et déclencheront des émotions positives; (2) un spectacle immersif audio-visuel, présenté au Biodôme de Montréal qui, en plus des émotions positives, portera un message éducatif sur l’écorésilience et l’écoresponsabilité.

Des méthodes de mesures qualitatives et quantitatives seront utilisées. Les caractéristiques des interventions seront définies via des groupes de discussion transdisciplinaires et intersectoriels. La cocréation des musiques et images, du contenu et de la forme du message éducatif, se fera à partir des recommandations de ces groupes. La mesure des effets des interventions sera faite via un essai clinique randomisé en groupes parallèles chez des adultes écoanxieux·ses et non-écoanxieux·ses. L’organisation et la structuration des activités reposeront sur un écosystème apprenant de type laboratoire vivant réunissant les sciences fondamentales (botanique, neurosciences, génie), humaines (philosophie, musique, art écologique) et quatre acteurs de terrain qui sont le Jardin botanique et le Biodôme de Montréal, des artistes et des usagers.

Cette convergence transdisciplinaire et intersectorielle de connaissances et savoir-faire pour en créer de nouveaux sur un mode d'écosystème apprenant, afin de repenser notre rapport aux végétaux et au monde du vivant, en agissant au niveau individuel et avec pour toile de fond deux enjeux de la société canadienne que sont le réchauffement climatique et l’écoanxiété, est un pari osé mais c’est aussi là que réside le caractère novateur du projet et son haut potentiel. Elle permet, par ailleurs, de proposer des actions concrètes de terrain destinées à un large public (visiteurs du Jardin botanique et du Biodôme - Montréal) et réplicables à plus grande échelle.

Most drugs target cell surface expressed receptors of the G protein coupled family. Drugs targeting receptors are rarely fully selective and their development is costly. These molecules also affect receptors all over the body, thus causing side effects by acting on receptors that are not expressed in therapeutically relevant tissues. These issues not-only prevent the exploitation of valid therapeutic targets, but also lead to costly postapproval discontinuation of otherwise effective therapies. A way to circumvent these limitations is to use gene therapy to target the expression of a given receptor in therapeutically relevant cells. However, gene therapies are also problematic as they cause permanent scars in the genome and do not provide ways to reverse treatment in face of negative outcomes.

We will use a combination of computational science, synthetic biology, and nutritional science to develop and validate a pipeline for the development of reversible gene therapies. As a proof of concept, we will apply this strategy to metabolic syndrome; a complex disorder combining both diabetes and obesity that represent a huge public health burden in Canada.

We have generated a truncated version of the mRNA-targeting Cas7-11 protein that is compatible with adenoassociated virus (AAV)-driven expression. We also used computational approaches to identify about 500 blood brain barrier permeable putative Cas7-11 inhibitors that could be used as a "kill switch" in face of unexpected negative outcomes. This strategy will leave the genome intact and open the door for the development of safe and reversible precision gene therapies to replace existing drugs.

Using computational approaches in humans and mice, we identified the cannabinoid receptor (CB1) expressed in neurons of the hypothalamus as a good target for metabolic syndrome. Systemic targeting of CB1 causes severe side effects, thus providing a candidate situation for gene therapy.

AIM 1: Development of an in vivo Cas7-11 strategy targeting CB1.

AIM 2: Identification of small molecule inhibitors of Cas7-11.

AIM 3: Validation of the approach in a model of metabolic syndrome.

Results from this project will provide new ways to implement precision therapeutics that do not affect genome integrity while ensuring therapeutic safety. This can be applicable to metabolic syndrome and most diseases for which inhibition of receptor function is indicated.

Over 2 million Canadians live with osteoporosis. In clinical settings, bone mineral density (BMD) as measured by dual-energy x-ray absorptiometry (DXA) is used to identify at-risk individuals. DXA scans are commonplace and High-Resolution Peripheral Quantitative Computed Tomography (HRPQCT) is emerging as an alternative, richer clinical BMD measurement tool. BMD is a flawed metric, while it correlates with bone fragility, its predictive power is limited. For example, 60% of women with osteoporotic fractures have non-osteoporotic BMD. The reason why BMD is a poor predictor of mechanical properties is intuitive: topology matters! The strength of this porous material depends on the topology of the interconnected tissue network.

In the past decade, computational modelling, namely finite-element analysis (FEA), has helped leverage the richer information (i.e. topological properties) provided by emerging imaging tools (e.g. HRPQCT). FEA is a powerful predictive tool, but it is computationally expensive and cannot easily predict fracture behaviour. FEA requires a three dimensional (3D) mesh of elements connected by nodes which is cumbersome to create and computationally expensive to solve.

Here, our team unites a condensed matter physicist, a frailty health expert and a biomechanics researcher to adapt a lesser-known computational technique—the Lattice-element Method (LEM)—to simulate the failure of bone tissue under load. The LEM is inspired by atom-scale particle-based dynamics simulations, and chiefly used in civil engineering applications. Instead of using a mesh with 3D elements and nodes, the LEM discretizes the system as small mass points that interact through spring-like interactions. If the springs and lattice of mass points are well chosen, the LEM can model materials with a level of accuracy comparable to FEA, but with improved numerical stability when the material is deformed to failure.

Successfully adapting this method to simulate fracture in bone tissue would be a transformative advance from both a computational mechanics and health science standpoint, allowing for much cheaper, faster and more reliable diagnoses based on HRPQCT imaging. In this project, we will develop an open-source LEM simulation software, and validate it against experimental mechanical testing of 3D printed material mimicking the topology of human bone samples imaged via lab-scale and clinic-scale micro computed tomography.

Alterations in various domains of cognitive function are key findings in psychiatric disorders. Batteries of cognitive tests have been created with a view to diagnosing, predicting, and guiding treatment for different illnesses, such as psychotic disorders. Computational modeling approaches, such as Reinforcement Learning, the Hierarchical Gaussian Filter, and Active Inference models, have allowed for the estimation of latent computational parameters from task behavior, but insights from these computational methods have not been translated consistently to the clinic. Reasons for this include constraints on time, resources, and technical expertise, as well as the fact that tasks can be long, repetitive, and boring. In addition, their limited ecological validity constrains generalization to clinical settings. One approach to improving adoption and dissemination of computational tasks has been “gamification”, in the form of adding narrative framing and an engaging aesthetic to a behavioral task (e.g. Sea Hero Quest in Alzheimer’s research). There are limitations to this approach: resulting tasks have fundamentally similar dynamics and lengths as the original tasks, with only marginal improvement to their ecological validity. We aim to develop a novel research tool to bridge this gap by making use of the expertise available in modern video game design to create a game that integrates behavioral tasks into its structure. It differentiates itself from other efforts by 1) improving ecological validity by maintaining player agency and utilizing fewer abstract scenarios and goals and 2) leveraging overlapping game elements to maximize the efficiency of latent parameter estimation. This would be an open-world game in which players are tasked with reaching a goal and must overcome challenges on their way, incorporating elements of existing tasks. The tool would use built-in computational modeling to generate insights from player choices and behaviors. It is intended to be customizable, allowing users to alter parameters based on their needs. This tool would be useful as a digital biomarker, research tool, or screening program for mental illnesses such as psychosis. This is the first project aimed at creating an integrated game environment for mental health assessment aimed at providing actionable insights to clinicians and patients and as such is a high-risk endeavor, requiring innovations in game and task design and computational modeling in order to be successful.

Advancements in AI in healthcare have raised significant ethical and philosophical challenges, particularly related to unequal access to quality healthcare and the risk of misdiagnosis, especially among marginalized populations, lacking a clear definition of AI responsibility. Progress has been made in understanding the impact of AI on health inequality, such as technology-based studies aiming to define and mitigate algorithmic bias and evaluate fairness in healthcare. However, health inequality remains a complex issue. We propose a transdisciplinary approach to consolidate foundational knowledge, advancing towards Responsible AI and health equity. We aim to define key concepts/terminology (ethics and philosophy), prioritize human rights (law), revise and propose regulations (public policies), identify current technological limitations (engineering and sciences), and propose an educational tool to promote public trust (sciences and healthcare practitioners).

The three components of this proposal will be performed by our transdisciplinary research team to: 1) Analyze human data with focus on safety, privacy, ethical usage, and human rights, along with best practices for human data collection, sharing, usability. This will be led by our team of experts in ethics, engineering, law, and philosophy; 2) Aggregate and curate current healthcare frameworks, identifying standards and regulations for the responsible deployment of AI in healthcare, including a comprehensive technology risk assessment. This will be guided by our experts in science, public policies, and engineering; 3) Develop an educational tool to promote public trust, led by experts in education, medicine, science, and engineering.

This is a high-risk, high-reward proposal given the complexity of healthcare inequality, a problem that transcends a specific area of study and impacts everyone, especially marginalized communities. Health inequalities cost Canada's healthcare system approximately $6.2 billion/year, and the usage of AI in healthcare could exacerbate these disparities. Healthcare is a vast area of study, comprising diverse data types, various pathologies, multiple demographics, and different objectives (diagnosis, prognosis, follow-ups, etc.). For feasibility, our focus will be on publicly available healthcare challenges (open science). Our transdisciplinary project will define, discuss, and propose solutions related to responsible AI in healthcare, advancing towards health equity.

Breast cancer is currently the most prevalent cancer in Canada. Despite remarkable progress in the past three decades, it remains the second leading cause of death from cancer in Canadian women. Tumors expressing the estrogen receptor alpha (ER+) form the most frequent subtype of breast cancer. While these tumors usually respond well to hormonal therapies, many patients eventually develop treatment resistance, leading to cancer progression. Different genetic mutations have been associated with such resistance, with variable impacts on the different types of hormonal therapies. Screening for these mutations could thus inform therapeutic choice, but is not currently done routinely in Canada, due in part to technological availability. Detecting these emerging mutations in a clinical setting remains challenging: they appear in very few tumor cells, differ only by a single-nucleotide variation (SNV), and multiple possible versions of these mutations need to be screened for.

Our team proposes to create a novel multiplexed bioassay based on graphene field-effect transistors (GFETs), to enable testing for multiple SNVs in liquid biopsies of breast cancer patients. An emerging class of bioanalytical sensors, GFETs are miniature electrical circuits made of functionalized graphene, a thin nanomaterial whose electrical conductance responds directly to the capture of a specific biomarker, without previous amplification or labeling of the sample. In this project, our original approach will be to exploit specific modulations of the gate potential, applied electrolytically, to develop (1) a chip design enabling multiplexed SNV discrimination and (2) a novel machine learning approach to signal recognition in clinical samples. Our transdisciplinary approach will uniquely combine device engineering, molecular simulations, artificial intelligence and bioassay design for translation to clinical oncology.

The development of a new test modality for multiplexed SNV profiling in clinical samples will have a transformative impact not only in breast cancer, but more generally in medical oncology and in other areas of diagnostics such as epidemics, toxicology or other biological threats. The all-electronic format offers promising avenues for low-cost manufacturing, portability, IoT connectivity and ease-of-use at the point of care, all supporting a much-needed improvement in the accessibility of clinical tests on a global scale.

Currently contaminant surveillance and policy rely heavily on the field of ecotoxicology to establish critical loads and acceptable limits of impact from anthropogenic activities. Ecotoxicology, is predicated on Western science and typically excludes Indigenous knowledge and lived experience. Yet, Indigenous Peoples, both in urban and more remote settings, face disproportionate levels of environmental contamination resulting from settler colonialism and resource extraction. Decision making without Indigenous voices and knowledge of the land and waters reiterates colonial narratives of knowledge validation and privileging Western science. This approach leads to imposed risk mitigation strategies that amplify injustice, such as the reduction of harvesting/consumption of traditional food sources and also fails to equip Indigenous Nations (INs) with tools to understand contaminant risks.

This project asks a vital question: how can we create an anti-colonial toxicological practice that works for rather than against INs?

Led by a transdisciplinary research team (ecotoxicology, Indigenous knowledge, environmental justice) and five IN collaborators, this research addresses concerns about the major gaps in understanding of contaminant burdens in Indigenous territories, and their impacts on food security, exposure, and reduced social-environmental health. We will develop accessible environmental forensic tools that will be used to understand sources and fate of contaminants and establish risk mitigation strategies that prioritize protecting and restoring humans' relationship with land and waters, while providing INs with data that can be used to hold polluters accountable and inform land use decision making. Each phase of the project will centre Indigenous place-based knowledge, community engagement, harm reduction, and environmental justice. By drawing on commonalities from the local programs, we aim put forward an adaptable framework for informing national contaminant surveillance.

Our research has the potential to revolutionize how we assess contaminant burdens in Indigenous territories through transdisciplinary approach that integrates community-based science, decolonial perspectives, and co-designed surveillance programs. If successful, this project has can transform ecotoxicology practices, improve Indigenous health, address historical injustices, and influence environmental policies governing chemical impacts.

Context: Disaster recovery is the process of rebuilding the physical and social fabric of a disaster-struck community. Due to the increasing costs of disasters, the Canadian Federal government is revising its disaster recovery programs to improve their effectiveness. Longitudinal disaster recovery data are ideal to inform these decisions. However, because longitudinal studies are resource-intensive, we lack data to improve Federal policies.

Paradigm shift: Previous disasters for which longitudinal recovery data could not be collected are accepted as missed opportunities to learn what interventions (e.g., Federal aid) effectively mitigate recovery inequalities. This project asks: Can data collected for other purposes (i.e., secondary data) be used as a surrogate for in-situ, longitudinal disaster recovery data collection from previous disasters? This is a paradigm shift where disasters for which primary data were not collected will become learning opportunities.

Methodology: We will develop a novel computer vision algorithm to use satellite and Street View images from multiple years to develop a longitudinal understanding of recovery from one past event (e.g., Fort McMurray fires). Knowledge from the literature in urban planning, social sciences, and disaster financing will be integrated to inform the metrics collected by the algorithm. We anticipate collecting data on physical recovery (e.g., counts of homes rebuilt), quality of living changes (e.g., availability of green spaces), gentrification (e.g., home size and build quality), among others, to develop a holistic understanding of recovery that reflects physical and social aspects.

High risk: Computer vision algorithms excel in extracting high-level understanding from a set of related images. Conversely, our approach requires an algorithm that extracts detailed insights into a complex, spatially and temporally correlated societal problem. Without a deep interdisciplinary integration for the algorithm's development, it may provide shortsighted results about people’s experiences, which would be inadequate to inform policies.

High reward: Our project can produce a standardized and versatile approach to learning from previous disasters. This would allow an improved understanding of what interventions (e.g., Federal aid) effectively mitigate recovery inequalities, generating economic (e.g., improved Federal aid policies) and societal (e.g., improved disaster recovery) benefits for Canadians.

The objectives of this project are to: 1) Explore the use of automated agents or "robots" to negotiate agreements for energy trading, labour contracts, and diplomatic accords; 2) Deliver case study examples of robot-negotiated energy, labour, and diplomatic agreements; 3) Develop quantitative metrics to compare the differences between robot and human negotiated agreements; 4) Identify the most influential factors for reaching successful agreements; 5) Communicate the advantages and disadvantages of using robots to negotiate for humans. Investigators will leverage their research and professional experience in power procurement and labour negotiations to extend their automated agent energy trading platform for broader utilization. The research team will develop defined bargaining environments for each negotiation application. A bargaining environment includes simplified models of external variables (ex: wind conditions, union position, public opinion poll, etc.) that will influence negotiations. The same bargaining rules will be extended to both human and robot negotiators. The outcomes of these bargaining experiments will be measured by a set of new quantitative metrics to help evaluate the degree of each party's success in the negotiations. Outcome metrics and bargaining parameters will be varied to conduct sensitivity analyses to illustrate the variables most influential in successful deal making. While robots have been shown capable of negotiations, the literature is lacking a quantified spectrum of comparison across negotiation environments and against human benchmarks. The investigating team anticipates significant strategic value for energy, labour, and political sectors arising from this project. Negotiation between humans can be confrontational, emotional, time consuming, and resource intensive. Human-conditioned automated agents have the capacity to negotiate rapidly, tirelessly, and objectively. Creativity is a well-cited critical element to bargaining success, there is much debate about how creativity can be represented in automated negotiations, something this study will help to reveal. Automated bargaining can reach multiple deals in seconds and report them back to human evaluators for review. It is expected that the increase in objectivity, bargaining iterations, and adaptive intelligence by robot negotiators should improve the value and equity of deals for human energy, labour, and diplomacy.

Canada’s 93,000 non-speaking children with profound disabilities are 3x more likely to be underweight vs. non-disabled children and face elevated risk of malnutrition due to swallowing difficulties, gastrointestinal disorders, and inability to communicate. 1 in 5 hospitalized children are malnourished, and only 50% are then seen by a dietitian. Due in part to medical complexities, children with disabilities experience malnutrition at a much higher rate than their non-disabled counterparts. While much research has focused on medical conditions contributing to pediatric inadequacy to take food, psychosocial factors, especially caregiver feeding practices, have been largely overlooked in the nutritional care of children with disabilities.

Feeding is a complex psychosocial experience shaped by how and by whom a child is fed, their preferences, and caregiver anxiety. This complexity is evidenced by the spontaneously arising neural and behavioral synchrony between caregiver and child during responsive feeding. Nonetheless, feeding goals are narrowly informed by estimates of the child’s energy needs at rest. These episodic estimates are influenced by many factors including the child’s stress levels, and do not afford any insight into how to achieve energy balance. Exacerbating this challenge, nonverbal children with disabilities cannot communicate hunger and satiety.

We aim to elucidate the neurobiological and social factors underpinning responsive feeding of hospitalized children with disabilities as a first step towards developing a novel, dyadic feeding quality monitor. We will measure the interplay between child and parent brain activities and describe child-parent social dynamics to understand how these factors relate to feeding outcomes. This interdisciplinary approach will adaptively integrate methods from neural engineering and behavioral nutrition.

We will recruit 20 parent-child dyads with non-verbal, orally fed children (age 3-19) with disabilities admitted to rehab care for 5 or more weeks. Data collected will include egocentric videos of mealtime interactions, in situ brain signals, and nutritional biomarkers. The findings will inform the subsequent development of a medical device that provides real-time feedback about the quality of the feeding interaction. If successful, the tool will have broad potential to improve the nutritional care of individuals of any age with communication and mobility impairments, in any setting.

Objectives: The goal of this project is to develop a high-throughput method for screening cells based on the biomolecule they secrete into the environment. Most screening methods available to-date require the target molecules to be physically tethered to the target cell, but this excludes the secretome which contains many potential drug candidates. Although there exists a handful of methods for screening cells based on their secreted proteins, these methods suffer limitations in sensitivity, throughput, and the need for complex instrumentation. We will establish a method that overcomes these limitations. The outcome will enable accelerated discovery of valuable biomolecules including therapeutically relevant enzymes, cytokines, and antibodies.

Background: Biologics are increasingly used as medicines to treat diseases, possessing advantages over small molecule drugs including pharmacokinetics and binding biophysics. A common challenge in biologics discovery is to identify and isolate a handful of variants from a large pool of candidates, i.e. screening, based on a desirable phenotype such as target binding. Current methods to screen cells require physical linkage between the target molecule (phenotype) and the target cell (genotype), which is not compatible for screening secreted biomolecules and their interactions with the environment.

Approach: We combine droplet microfluidics, synthetic biology, and hydrogels to establish a novel method for single-cell screening. Droplet microfluidics will enable us to trap individual cells within hydrogels displaying baits against the target molecule. We will design the hydrogel to contain DNA crosslinks, making it susceptible to destruction by programmable nucleases such as Cas12a, a type of CRISPR-Cas nuclease. Activation of this nuclease will be placed under the control of a gene circuit, triggered by secretion of the target molecule – thus linking hydrogel destruction to the desirable phenotype. The desired cells will be liberated from the destroyed gels while non-desired cells will remain trapped in the gel. This approach is unique because it enables volumetric, massive-parallel screening of single cells which has not yet been reported.

Sex and gender must be taken into account in health research and the provision of care. While sex refers to biological differences and gender refers to sociocultural effects, these terms are often used interchangeably, and most studies only consider their combined effects in a binary framework comparing two sexes/genders. We will develop a tool to assess variation in multiple types of gendered exposures (e.g., time spent providing care), distinct from sex. This will bring together researchers in basic sciences (sex differences and epigenetics), clinical sciences (how sex and gender affect health outcomes) and social sciences (what variables describe different aspects of gender, and how these intersect with other traits). We will also engage with people with lived experience of gender-based inequities in health care. We hypothesize that there are epigenetic markers, e.g., differences in DNA methylation, that are associated with gendered exposures, and others that are associated with sex. To identify such associations, we will first use publicly-available datasets to identify DNA methylation differences between men and women in blood. These differences may be due to sex and/or gender. However, methylation differences that are present in many different populations and/or present in newborns are likely biologically-based, whereas other methylation differences are candidate markers of gendered exposures. We will then examine whether these candidate markers are associated with gender-related variables reflecting different types of gendered exposures. This research defies paradigms by considering the potential imprint of gender on biology, and by viewing gender as non-binary and made up of multiple domains. The ability to distinguish between biological sex and sociocultural gender would revolutionize health research and clinical practice by allowing us to move beyond a binary view of gender completely confounded by sex. Our index could be used for retrospective samples, where blood samples may be available, but gender-related variables are not. Prospectively, blood samples may be less invasive than surveys for study participants, and less labour-intensive to collect, making it easier to incorporate both sex and gender and to consider multiple gendered dimensions in studies. This in turn would improve understanding of the social determinants of health, identify factors likely to generalize between populations, and improve care for everyone, cis and trans.

The sustainability of the Canadian swine industry (>15M hogs, 1.2M sows; total economy activity >$24B) is threatened by the high incidence of post-weaning diarrhea in pigs. This condition results in the death of at least 1-2 piglets per litter (average mortality 10%) and costs up to $200 per sow annually. This condition, typically caused by pathogenic bacteria, impedes growth, increases mortality and raises animal welfare concerns. The situation is worsened by the (mis)use of antibiotics and an alarming rise in antimicrobial resistance.

In a quest for new treatments, we propose utilizing smart capsule technology for the intestinal delivery of de novo immunomodulatory peptides. These peptides, derived from encrypted sequences in nature mined for hidden antimicrobial and immune-boosting functions, are identified using artificial intelligence (AI) and deep learning from human, microbiome, and ancient animal proteome libraries. The peptides will be incorporated into ingestible capsules featuring thermo- and pH-responsive inlets, enabling controlled release into the intestinal lumen.

To test it, weaned pigs challenged with enteropathogens responsible for late and early post-weaning diarrhea (i.e., Brachyspira spp and enterotoxigenic Escherichia coli (ETEC) pathotypes with varying antimicrobial resistance levels) will orally receive the capsule-conjugated peptides. We will assess the clinical efficacy of treatments by mitigating intestinal protein-cleaving enzymes related to "leaky" gut syndrome and by promoting the mucus layer thickening and mucin glycoproteins that increase shielding and discourage bacterial colonization. Shifts in intestinal microbiome communities will determine specific microbiota that either regulate protease secretion or mucin glycans in reducing inflammation. We will ensure animal welfare throughout the treatment, assessing disease-related behaviour traits. The mechanisms by which encapsulated peptides enhance gut health will be deciphered in porcine intestinal cells, organoids, and murine colitis models.

This project integrates distinctively veterinary immunology related to animal production with bioengineering and peptide chemistry to develop swine-specific treatments to combat gut diseases and reduce post-weaning mortality. Lowering mortality rates in wean-to-finish production provides substantial economic benefits in the swine industry (e.g., a 1%-point reduction increases income to up to $25 per sow).

Astrochemical models are used to explain the composition of the interstellar medium and star-forming regions based on detailed chemical processes, even though their predictions and the observations against which they must be validated are often highly uncertain. Such an exceptional epistemic situation represents an opportunity for philosophers and statisticians to refine their understanding of model evaluation and to contribute to developing new methods for assessing the adequacy of a model.

Objectives: Astrochemistry represents a domain where observations are often sparse or uncertain. A major bottleneck is the requirement of costly and challenging experiments or quantum calculations to accurately predict chemistry under the extreme conditions of interstellar space. Our objectives are to: (1) identify key kinetic parameters that increase the model's predictivity by reducing uncertainty (2) develop an efficient statistical ML pipeline that predicts new kinetic parameters; (2) explore epistemological tools to characterize the success of the ML predictions; (3) identify data sparse targets for detailed lab studies.

Approach: Our ML approach will be based on emerging advances in (1) multimodal probabilistic spatiotemporal models, which fuze multiple data sources with informed priors based on physical laws; and (2) high-dimensional Bayesian optimization for guiding the use of experimental resources. The success of the pipeline for producing kinetic data for astrochemical models will be explored using novel epistemological tools. We have assembled a team with diverse expertise in the areas of astrochemistry, ML, statistical theory and epistemology to develop this multidisciplinary approach.

Novelty and significance: This project is high-risk. Our multidisciplinary approach, applying cutting-edge ML and epistemological techniques in astrochemistry, is completely novel. Despite the risk, this project is high-reward. In addition to advancing statistical and philosophical methods, our ML pipeline will enable accurate predictions of kinetic parameters at orders of magnitude faster speeds than traditional methods, while also automatically determining which systems require further resources. Predictions will be made for molecules that cannot be isolated in the lab. Laboratory resources will then be directed to key systems and their impact on the environment will be reduced.

In a society where wellbeing issues are increasingly common, how can we sustainably enhance wellbeing? While various wellbeing (WB) programs are available across different contexts (e.g., workplaces, communities), their effects tend to be modest and short-lived. A new theory suggests that the limited impact of these programs may stem from insufficient focus on fostering wellbeing literacy (WBL)—the ability to understand and use WB language effectively in different contexts to improve one’s or others’ WB. WBL has the potential to enhance the effectiveness of WB programs. However, WLB remains understudied.

Developing WBL is an interdisciplinary challenge, requiring insights from diverse fields, e.g., education, psychology, linguistics, sociology, and computational science. Given WBL’s focus on language, recent advancements in computational science, like natural language processing, could be groundbreaking for measuring and teaching WBL.

This project aims to enhance WB programs’ effectiveness by developing a free, AI-based intervention tool that assesses and fosters WBL interactively. To be effective, the tool must adapt to diverse users’ (e.g., women, LGBTQ2S+, immigrants, neurodivergent individuals) specific contexts and needs. AI would allow to measure WBL in real time and provide content (e.g., interactive text, visual activities) tailored to users’ WBL levels.

The goals are to: 1) Develop an AI model to assess and improve WBL in diverse adult users by training it on WBL literature and codesigning prompts with diverse users; 2) Integrate this model into an online tool and evaluate the tool’s effectiveness; 3) Test whether using this tool before a WB program (e.g., mindfulness, life crafting) enhances the program's outcomes.

An interdisciplinary research team from physical, health and social sciences will co-design the WBL models and intervention tool, also involving diverse users. A mixed-methods approach will evaluate the tool's effectiveness, including quantitative assessments and qualitative feedback from users. A trial will test if the AI tool boosts WB programs’ efficacy.

The project will offer a proof-of-concept of the first ever created AI-based tool focused on WBL. The integration of WBL, AI and state-of-the-art human computer interaction features (i.e., engagement, personalization) will have game-changing impacts for WB promotion, helping people of different backgrounds be more competent and prepared to undertake WB program.

Les premiers épisodes de psychoses (PEP) débutent précocement, se chronicisent, et entraînent des déficits fonctionnels et des coûts élevés pour le patient et pour la société. Les antipsychotiques (AP), pierre angulaire du traitement, n'améliorent pas les symptômes clés ni la cognition. Ils ont des effets indésirables métaboliques graves, particulièrement la résistance à l'insuline qui touche directement le cerveau. Nous proposons de reconceptualiser les PEP en intégrant le métabolisme énergétique cérébral et mitochondrial. Les grands besoins énergétiques du cerveau sont fournis principalement par le glucose, mais les cétones (beta-hydroxybutyrate, acétoacétate) sont également des carburants cérébraux importants. Selon nos études, l'hypométabolisme cérébral du glucose dans la maladie d'Alzheimer peut être compensé par des cétones exogènes, améliorant ainsi le fonctionnement cognitif. La perspective innovante du projet est que les PEP impliqueraient un problème d'hypométabolisme cérébral spécifique au glucose mais, comme dans la maladie d'Alzheimer, un métabolisme cérébral des cétones toujours intact. Nous sommes une nouvelle équipe interdisciplinaire avec une expertise complémentaire en clinique et biologie cellulaire. Le métabolisme énergétique cérébral serait évalué chez des PEP avant et après 8 semaines de traitement avec AP. Le fonctionnement mitochondrial mesuré par le profil d'expression de gènes de la chaine de transport des électrons, les symptômes cliniques, le métabolisme glucidique périphérique, les marqueurs d'inflammation (cytokines) et la cognition seront mesurés et comparés aux témoins sains. L'originalité de ce projet réside dans le fait que ce serait la première fois que l'utilisation des carburants cérébraux et la santé mitochondriale seront évaluées chez les mêmes patients PEP, surtout avant et après la prise d'un AP. Bénéfices potentiels attendus: (i) une meilleure compréhension du lien entre les symptômes psychotiques et le dysfonctionnement énergétique au cerveau, (ii) une stratégie thérapeutique éventuelle réaliste pour contrer les symptômes réfractaires aux AP, et (iii) réduire la chronicisation des psychoses vers la schizophrénie, ce qui changerait complètement la vie des patients. Risques possibles: ces patients sont vulnérables et désorganisés, ce qui pourrait poser un défi en termes de tolérance aux examens TEP et aux procédures de l'étude. Avec notre expérience collective, nous ferons des suivis attentifs et personnalisés.

Humanity’s success is founded on our unique capacity to cooperate in groups of up to millions of strangers. But cooperation does not persist indefinitely; forces of destabilization can lead to social fissures and violence, sometimes precipitating complete societal collapse. Research into how cooperation develops and breaks down, however, has been siloed and partial. Some start with individuals to understand the behaviours and choices that people make, while others focus on broad societal-level institutions and norms that set incentives and constraints. Each approach has produced important insights, but each on its own provides only one part of the story. The complex feedback between societal-level systems and the underlying cognitive processes of individuals have yet to be fully understood. This leaves many crucial questions about how cooperation and cohesion can be generated and made resilient against fracturing and internecine hostility, two of the major drivers of global instability and conflict around the world today. Here, we propose a novel collaboration bringing experts from divergent domains together to track, for the first time, the complex dynamics that result in sustained intra-group cooperation across multiple levels – the individual, the groups with which they identify, and the societies that structure the relationships between these groups – while accounting for how dynamic feedback shapes and shifts these interactions through time. Exploring the connections between these levels, we aim to uncover how individual behaviour changes as group identities form, expand, and shift, norms are established and enforced to differing degrees, and groups either compete or cooperate with others. This work is extremely high-risk due to the novel intention of bridging work both across disciplines and levels of analysis, but the rewards for a growing, diverse country like Canada are significant; our work holds the potential to help us understand – and so head off – the key drivers of social unrest and conflict before they become unavoidable. To do this, we will leverage cutting-edge reinforcement learning models, utilize proven social-psychological experimentation techniques, and combine these with models and theoretical insights taken from fields as diverse as psychology, sociology, history, and evolutionary biology. Our team is uniquely qualified to accomplish these goals, offering the range of expertise required.

Urogenital conditions pose a serious burden on public health and the economy in Canada. Kidney disease alone affects over 4 million Canadians and its annual economic burden exceeds $40 billion. Urogenital conditions are often 'silent', progressing for years before symptoms appear and a diagnosis is made. More effective treatment and management of urogenital conditions therefore depends strongly on our ability to detect them early-on. The urinary system is a fluid system. Urine is transported from the kidneys to the bladder which fills through intermittent jets emanating from the two ureters (ureteral jets). From a fluid mechanics perspective, a progressive change affecting the urinary system will drive a progressive change in urine flow patterns within the bladder. For example, changes in the flow patterns due to kidney disease or stones will arise from differences in ureteral jet intensity and frequency, though due to prostate cancer they will arise from differences in bladder geometry. In this novel research project, we investigate the urine flow patterns within the bladder and aim to develop ultrasound-compatible flow-based indicators to signal abnormal urogenital conditions in their early stages. We use our unique, custom-made in vitro bladder flow simulator to model bladder flow patterns arising from natural variations as well as from an array of urogenital conditions. This research project represents the first time these flow patterns are captured and examined, and we do so at high spatiotemporal resolution using the state-of-the-art particle image velocimetry technique. We plan to distinguish 'healthy' from 'unhealthy' flow patterns through modern fluid dynamics analyses (e.g., energy budgets, coherent structures) and data-driven modelling techniques (e.g., machine learning). Colour flow Doppler measurements of the ureteral jets in real patients will be used to validate the in vitro model and indicators. Conversely, a sparse transformation will be devised to predict in vivo bladder flow patterns from ureteral jet information. The proposed fluid dynamics approach has already shown promise for the detection of cardiovascular conditions. With point-of-care ultrasound making its way into routine physical examination and the movement toward personalized medicine, this research can have a profound impact on our ability to diagnose urogenital conditions in their earliest stages and can serve as a stepping-stone for the development of digital twins.

Context: “Food is medicine” approaches offer exciting potential to improve food access and diet-related health outcomes. One such approach involves leveraging healthcare settings to enable dietary changes by providing a financial subsidy or incentive to access fresh produce through food boxes or redemption sites such as farmers’ markets. While the potential benefits of food prescribing interventions are well-established, program utilization and integration within healthcare settings remains an ongoing challenge, particularly in Indigenous and other underserved communities. Innovative strategies to incorporate nutrition and cultural food preferences in healthcare settings are critically needed to enhance program reach and impact, and to ensure that programs are culturally inclusive and address barriers to equitable access.

Approach: Building on existing community partnerships and priorities, our team will explore the acceptability and feasibility of food prescribing models and their practice implications for promoting healthy food access and chronic disease prevention with a focus on rural and urban Indigenous contexts. Grounded in community-based participatory research (CBPR) and guided by Etuaptmumk (Two-Eyed Seeing), we will use Indigenous methodologies and implementation science to co-design a novel food is medicine program. _x000B_ Research objectives include: 1) To understand perspectives of wellness related to the role of traditional foods in fostering connections to identity, food security, and nutrition; 2) To describe community-informed components of food prescribing that can support healthy food access; 3) To identify culturally responsive practices to promote equitable access and safe delivery of a food is medicine program utilizing a community-informed food prescribing model.

Novelty: Our interdisciplinary inquiry draws on expertise in health sciences, nutrition, medicine, policy, environmental sciences, and Indigenous knowledges to generate insights into the challenges and opportunities of food prescribing as an innovative model that bridges the gap between medical and social care. The novelty of our research also lies in its community-engaged approach to advance community-led actions to improve food access and diet-related health outcomes. The work has the potential to expand Western approaches to care by creating opportunities for Indigenous knowledges and worldviews to inform wise practices for food prescribing.

The emergency department (ED) is a critical setting where timely and accurate patient care is essential. Often, patients present to the ED are in crisis and cannot provide information about their medical history, which is vital to safe patient care. This requires extensive data synthesis and analysis from the patient’s electronic health records (EHRs), potentially detracting physicians from direct patient care and contributing to provider burnout. While the need for complex knowledge synthesis on patient’s medical histories exists in other areas of the acute care environment, the time-sensitive nature of treating patients in the ED emphasizes the need for innovative solutions to streamline this process. Large language models (LLMS) can take in and synthesize complex information very quickly. However, the accuracy of LLMs in a high-stakes environment such as the ED is paramount. Collaboration between multiple LLM agents has emerged as a promising approach to enhancing the capabilities of individual LLMs. Recent studies have developed different interaction architectures and assigned agents in static patterns to enable collaborations between agents, including multiple LLM instances debating, distributing LLMs in parallel, and concatenating their answers to produce better results. In this research, we combine the expertise of a research team with computer scientists, physicians, ED operations researchers, and engineers to jointly propose a simulation framework based on LLM agents for optimizing ED operations. Specifically, the clinicians and ED operations researchers will contribute their domain knowledge, design simulated ED workflows, and evaluate the summarization outcomes from LLMs. The computer scientists will design metrics to evaluate the performance of LLM agents automatically. The engineers will develop a simulation framework to mimic ED workflows using LLM agents. Our method aims to be consistent with healthcare professionals' evaluation criteria. Our goal is to allow the framework to mimic the effects of existing and new roles and eventually to freely evaluate these roles, potentially beneficial to patients without incurring costs and headcounts to the healthcare system. This project is high-risk because it remains uncertain if LLMs, as an emerging technology, can act as effective agents for roles in ED. Yet it is high-reward for its great potential to reduce burnout among healthcare practitioners and shorten patient wait times in ED.

L’autisme est historiquement conçu selon une perception déficitaire, c’est-à-dire comme un ensemble de déficits à compenser, notamment en matière de communication. Pourtant, des recherches récentes montrent que les personnes autistes peuvent communiquer efficacement avec d’autres personnes autistes, tandis que les non-autistes ont autant de mal à déchiffrer les personnes autistes que l’inverse. C’est ce qu’on appelle le problème de la double empathie : les problèmes de communication entre neurotypes ne sont pas liés à un déficit mais à une rupture mutuelle de l’empathie. Concevoir l’autisme comme un déficit peut nuire au bien-être et à la santé mentale des personnes autistes. Le concept de double empathie permet d’approcher le problème de la communication entre neurotypes d’une façon équitable.

Les progrès des dernières décennies dans le domaine de l’intelligence artificielle ont permis de raffiner l’étude du langage naturel par des méthodes informatiques. Le Traitement Automatique des Langues Naturelles (TALN) permet le traitement de larges volumes de données textuelles pour identifier des motifs récurrents. Les applications fréquentes permettent par exemple de détecter des émotions ou de catégoriser des textes. Ainsi, la recherche en psychologie fait déjà appel à des techniques de TALN pour identifier certaines conditions dans les productions écrites, telles que la dépression. L’autisme demeure toutefois peu étudié par ces moyens, et les travaux existants comparent encore les productions des personnes autistes à des standards neurotypiques, ce qui maintient l’autisme dans une perception déficitaire. Il n’existe pas, à ce jour, d’étude mêlant le TALN et l'autisme sous l’angle de la double empathie.

En ce sens, nous proposons d’utiliser des outils de TALN et d’intelligence artificielle pour analyser les communications entre neurotypes. Nos objectifs spécifiques sont de constituer un corpus d’échanges entre personnes autistes et non autistes, et de modéliser concrètement la façon dont les différents neurotypes communiquent. L’atteinte de ces objectifs permettra de démontrer empiriquement que l’approche portée par la double empathie, par opposition à une perception déficitaire, est une façon valide de concevoir la communication des personnes autistes. À long terme, cette conception qui repose sur un traitement équitable des modes de communication autistes et non autistes a le potentiel d’améliorer grandement les échanges entre neurotypes.

The goal of the proposed research is the creation of human digital twin predictive simulation driven by machine learning to guide clinical decision-making for geriatric rehabilitation.

In Canada, more than 20% of the total population is 65 years of age and older, and this proportion will only increase in the coming decades. This was not followed by increased autonomy and healthy living, creating important care needs for the senior population. One of the important issues faced by this population is the loss of balance, creating an important risk of falls that can lead to hospitalization or even death. This is putting stress on our healthcare system and on rehabilitation professionals such as geriatricians, physiotherapists (PT), and occupational therapists (OT), who attend to the healthcare needs of this population. Most rehabilitation strategies are driven mainly by experience and patient observation. This limits the quality of care, which can vary widely between professionals.

This project combines engineering and geriatric science expertise and aims to increase senior balance rehabilitation outcomes by guiding rehabilitation strategy. For this, we will combine our expertise in mathematical modelling of biological systems and machine learning with dynamics simulations of multibody systems and geriatric rehabilitation. This project will:

1- Create human digital twins of seniors driven by machine learning for balance simulation

2- Create a predictive algorithm for balance rehabilitation therapy

3- Validate the prediction on a group of senior participants

Human digital twins or neuromusculoskeletal modelling is a mathematical representation of bone, muscle, and neural control. We will first create a senior neuromusculoskeletal model that reproduces balance under perturbation using a reinforcement learning paradigm. We will then create a predictive simulation of this model undergoing virtual balance rehabilitation therapy to find the optimal therapy that presents the best outcomes. Finally, we will validate this simulation against real data from senior participants.

Mathematical modelling of biological systems and machine learning have been applied mainly in the health domain for drug discovery but never for geriatric rehabilitation. This research could have a significant impact on increasing the quality of care for balance rehabilitation and this methodology could also be used in other diseases such as Parkinson's, stroke or multiple sclerosis.

Emerging viral diseases, such as those caused by SARS-CoV-2, influenza A, Ebola, and Zika viruses, represent a significant threat to global health. Developing broad-spectrum antiviral agents to address these and future viral threats is crucial; however, traditional approaches focus on narrow-spectrum, direct-acting antivirals, which require extensive virus-specific characterization and lack the agility to rapidly address a new outbreak. Advancing broad-spectrum antiviral drug candidates is critical to safeguard global human health.

Host-targeting agents (HTAs) that leverage conserved virus-host interactions hold great promise for development as broad-spectrum antivirals. Nevertheless, the potential for off-target effects on healthy cells currently limits the therapeutic utility of such molecules. We propose an innovative solution to address this challenge: developing HTA prodrugs activated exclusively in infected cells. With its potential to revolutionize antiviral therapy, this strategy aims to deliver potent antiviral activity while mitigating harmful effects on uninfected bystander cells. Enhancing specificity for virally infected cells and minimizing side effects will improve patient outcomes.

We will harness thapsigargin (Tg), a potent inducer of endoplasmic reticulum stress with broad antiviral activity, to establish proof-of-concept against coronavirus (CoV) infection. We plan to leverage our novel total synthesis of Tg to prepare prodrugs activated only in CoV-infected cells following cleavage by the conserved CoV main protease (Nsp5). Specifically, we will:

1. Design and synthesize a series of Tg prodrugs containing the Nsp5 recognition sequence.

2. Test their specificity for virus-infected cells.

3. Assess their antiviral efficacy against CoV infection.

While prodrug strategies have traditionally enhanced the bioavailability of direct-acting antiviral nucleosides, our approach to developing HTAs that are only activated in infected cells is novel and unprecedented. This high-risk research program promises to introduce a new antiviral paradigm that can be readily adapted for unrelated viruses. Our interdisciplinary approach and synergistic expertise in medicinal chemistry and molecular virology uniquely position us to develop and validate a Tg-derived antiviral prodrug with real clinical potential. Such innovative molecules are essential for preparedness against inevitable future viral outbreaks.

Microbial communities are an immense and powerful frontier in science, shaping everything from our health to global geochemical cycles. Our natural human urge to understand, predict, and harness that power is stymied by the size (>billions) and complexity (>100s of taxa) of most microbial communities. However, in reality, individual microbes live in a tiny world governed by fine-scale spatial structure that microbiome science usually ignores. At this scale, microbial communities become tractable. The stakes are combinatorial: if microbes directly interact with only the few taxa in their vicinity, then there will be many fewer interactions in plant and animal microbiomes (with >100s of taxa) than would be predicted by the curse of dimensionality. This is a crucial simplification for microbiome science.

The proposed research aims to test the hypothesis that interactions occur between species that share physical space, constraining the number of microbe-microbe interactions and thus solving the curse of dimensionality in microbiome modeling and experimentation. To test this hypothesis, we will use fluorescence in situ hybridization (FISH) and spatial metatranscriptomics of a tiny plant (duckweed) to correlate the biogeography of microbiota with host gene expression. The small size of Lemna minor (duckweed) means that the plant’s entire microbiome and transcriptome can be spatially mapped by capturing microbial 16S sequences and host mRNAs at ~5000 positions on glass slides with spatially barcoded probes, then sequencing samples. We will use these data to build distributions of colony locations and colony sizes for each microbial taxon in the microbiome and to predict the physical space overlap between microbial species. We will then test whether the strength of microbe-microbe interactions scales with the extent of their spatial overlap using in vivo empirical plant-microbe experiments and in silico individual-based models.

Until now, we have lacked the tools to study how microbes interact on and with hosts at the right spatial scale. This grant will address that strategically important gap and integrate highly resolved spatial models of the microbial world with empirical data from a real plant microbiome collected at the same sub-millimeter physical scale, using new lab automation and computer vision tools to handle and phenotype tiny plants (duckweeds) and their microbes. We will, in short, characterize the size of a microbe’s world.

Total knee arthroplasty (TKA) is a life changing treatment for most Canadians, with over 75,000 procedures performed annually. Unfortunately, a considerable proportion (up to 20%) report dissatisfaction following their TKA procedure. In most cases, such negative outcomes cannot be explained by functional tests or standard radiological findings. We propose that dissatisfaction following TKA should be investigated as a more complex process involving the central processing of the brain. We will use an advanced neuroimaging technique called functional magnetic resonance imaging (fMRI) to establish relationships between individual differences in brain circuitry and dissatisfaction following TKA. Our proposal has two objectives. First, we will define pre-operative networks predictive of postoperative dissatisfaction. Second, we will define network changes occurring postoperatively to delineate pathophysiology associated with dissatisfaction and inferior outcomes following TKA. We hypothesize that patient selection can be informed with a network model derived from fMRI which can preoperatively predict treatment response, and that longitudinal network changes will be related to clinical outcomes. Patients will undergo resting state fMRI (rsfMRI) and complete patient-reported outcome measures (PROMs) focusing on satisfaction prior to surgery and at 1-year post-surgery. Using PROMs and pre-operative rsfMRI, we will build a machine learning model predicting post-operative patient satisfaction. This project has significance for the healthcare system as, to our knowledge, this is the first proposed study to evaluate the brain circuitry of patients following TKA who report persistent dissatisfaction. Given that up to 15,000 patients per year in Canada report dissatisfaction following TKA, there is a clear need for further investigation. Additionally, the estimated annual cost of dissatisfaction following TKA to the Canadian healthcare system ranges from 82 to 120 million CAD. Ultimately, this project would (1) inform patient selection; (2) further our understanding of the brain circuitry of dissatisfaction; and (3) potentially provide new network targets amenable to neuromodulation treatments. Lastly, the results of this study may allow for knowledge translation and application to a multitude of other clinical fields and procedures, while providing an avenue for significant future cost-savings to the healthcare system.

Intracellular polymerization is an innovative novel technique that has the potential to generate enormous possibilities in medicine, especially for bioimaging and cancer treatment. However, significant challenges remain to be solved such as the precise control of the optical properties of the polymers formed within cells as well as achieving polymerization in a crowded macromolecular environment under unfavourable conditions.

To meet this challenge, this project proposes, as a main objective, the controlled intracellular polymerization of conjugated polymers tailored to bioimaging and early-stage cancer treatment. More specifically, the subsidiary objectives will include the study of in-cellulo effect of the direct-heteroarylation polymerization (DArP) approach on cell biophysics and metabolism as well as the in-depth analysis of the polymerization kinetics. The method will be developed for study on colorectal cancer cells as it is the most occurring cancer within the First Nations in Canada.

The research approach retained for this project will first involves the synthesis of monomers and corresponding conjugated polymers based on a diketopyrrolopyrrole moiety possessing high fluorescence quantum yield (between 560 to 640 nm allowing to avoid autofluorescence of cells), tunable optical properties (by modification of the chemical structure) and optimized with intended functional groups to enhance cellular uptake (bioavailability).

The intracellular polymerization progress will be monitored by nanorheology combined with advanced single-molecule fluorescence methods such as Fluorescence Correlation Spectroscopy (FCS), Förster Resonance Energy Transfer (FRET), Fluorescence Lifetime Imaging Microscopy (FLIM) or brightness analysis. Among others, FCS enables non-invasive, high statistics measurements of concentrations to validate the cell culture uptake and diffusion coefficients to determine molecular interactions at nanoscale in one short experiment (60-180s per one measurement).

This project is highly innovative as it leverages the unique optoelectronic properties of conjugated polymers for bioimaging. Furthermore, the proposed polymerization approach (DArP) has never been attempted in living cells. The significance of this work lies in its potential to transform cancer diagnostics by providing a more effective and sensitive imaging platform for First Nations communities.

Our team, comprised of a low temperature quantum physicist and an electrical engineer, propose to fabricate a groundbreaking electronic device. The device, whose properties and functionalities are fully quantum at room temperature, would be a world-first achievement. While numerous electronic devices currently exist that can be quantum at cryogenic temperatures and can generate electronic flow defined by fundamental units, including the ubiquitous Plank constant (h) of quantum mechanics and the electric charge (e), none of these devices can operate at room temperature. For this reason, the demonstration of such a true quantum device capable of functioning at everyday temperatures would be revolutionary. If such a device could demonstrate electronic properties directly related to the quantum resistance Rk=h/(e^2), it would prompt a significant revaluation of what can be achieved by modern electronics.

Our proposal is based on an unexpected discovery made recently in the applicants’ laboratories. In this groundwork, thin electronic devices of pure bismuth were fabricated and studied. Bismuth is a wonder material, and it was in the context of this element that Faraday’s study of magnetism led to the discovery of diamagnetism. It is also in bismuth that the first quantum oscillations were observed in a magnetic field by Shubnikov and de Haas, a phenomenon that is precursor to the integer quantum Hall and fractional quantum Hall effects for which two separate Nobel prizes were granted.

In our recent studies, we discovered an Anomalous Hall Effect (AHE) in bismuth devices that was entirely temperature independent from near absolute zero to 300K (room temperature), a first of its kind. In his seminal 1988 work, 2016 Physics Nobel laureate Duncan Haldane theorized that the AHE could lead to a quantized version, whereby electronic properties become fully quantum and are related to fundamental units previously mentioned. We firmly believe that upon improvements in the fabrication of bismuth devices, we will be able to reveal the first observable Quantum Anomalous Hall Effect at room temperature. In doing so, we will provide Canada with an unprecedented transformative scientific and engineering advancement that will likely blossom into a solid portfolio of patents. Importantly, via the great talent of our HQPs and their entrepreneurial spirit, we aim to position Canada as a global leader in the foreseeable future of quantum electronics at room temperature.

Significance and potential impacts: Dementia is one of the main causes of dependency among older adults, leading to cognitive changes that can negatively affect social interactions. Alzheimer Society of Canada has identified “caregiver support” and “emotional well-being and maintaining sense of dignity” for Persons with Dementia (PwD) as priorities for research. Affective memory — the sense of who one has been in their life (e.g., a mother, a manager) — is shown to stay intact despite cognitive decline. This sense changes in each social situation and is crucial for success of interactions with PwD. For example, if a PwD feels that they are still in the role of a manager, when a caregiver provides instructions in a specific way (e.g., order them to do a specific task), it can negatively affect the interaction of the PwD and caregiver, and ultimately the wellbeing of the caregiver and PwD. Support and training on how to improve this affective connection is expressed by the caregivers to be missing yet crucial.

Approach: This research will follow user-centred development methods and Affect Control Theory (a social theory of humans’ affective interactions) to create and evaluate a novel training system comprised of a mobile application and a human-like social robot (filling in the existing gaps in available resources for training) to provide this training and support for caregivers. The project is highly interdisciplinary, requiring complementary expertise in social robotics, human-computer interaction, data modelling, affect and cognition, and social theories of human interactions to design and evaluate the training approach and dynamic interaction styles through a user-friendly interface.

High risk/high reward: This project addresses a unique and complex challenge and is high risk as predicting humans’ perceived identity at each interaction and generating suggestions for optimal affective interactions accordingly is a complex topic and requires highly interdisciplinary research. The outcome is high reward as it will provide the support needed, yet missing, for improving success and quality of daily interactions between PwD and caregivers. With dementia being one of the main causes of dependency among older adults, this training addresses a long-standing issue affecting quality of life of PwD and caregivers and will have significant social and health impacts.

The rising prevalence of mental illness among young Canadians is alarming and significant, with approximately 20% of Canadian children and youth experiencing mental illness in their lives. Extensive research indicates that exposure to multiple early-life adverse events such as stress and trauma significantly increase the risk of developing mental illnesses by altering the body’s stress response. The impact of such events on mental health is profound; suggesting the early detection and mitigation of these adverse events could potentially reduce the prevalence of mental illness by as much as 44%.

Despite the well-established association of early-life adverse events and various psychiatric disorders later in life, there is a lack of reliable and valid biosignature tools to objectively identify the presence of early-life adverse events. Given that primary teeth start to develop in utero and continue to form after birth, analysis of primary teeth could serve to identify children who experienced adverse events during this time. It has been shown that exposure to stressors during teeth formations can be recorded as changes in the structure of enamel including neonatal dental lines. However, it is not clear to what extent teeth can imprint early-life adverse events and how dental biomarkers can be used to detect the potential risks of mental illness.

The objective of this project is to perform image analysis guided by machine learning to process large amounts of histological examinations of dental images in primary teeth.

This is a two-phase project which involves a) a microscopic analysis of primary teeth of two groups of children with and without known mental disorders to identify differences in their teeth microstructure and its relations with early-life adverse events, and b) classifications of the common childhood mental conditions based on the histological examination of teeth.

This highly interdisciplinary project brings together a non-conventional diverse team of researchers in Mental Health, Dentistry, and Computer Sciences to address the complex problem of early detection of children at risk of mental illness.

If our approach proves to be successful, it introduces a novel cost-effective measure with low burden on families using a non-invasive technique for earlier detection of those at risk of mental health conditions which is essential for effective preventions and treatment plannings.

Dysphagia, a prevalent swallowing disorder, affects over 590 million people globally, including 80% of Parkinson’s disease patients and 50% of stroke survivors. This condition, marked by difficulty swallowing, can lead to chronic malnutrition, compromised immune function, and fatal complications such as pneumonia, the 6th leading cause of death in the US. Dysphagia diagnosis relies initially on subjective questionnaires, followed by clinical instrumental examinations. Accurate detection involves invasive methods such as fiberoptic endoscopic evaluation of swallowing or video-fluoroscopy, which increase costs, delay diagnosis, and deter patients, leading to a 67% rate of missed diagnoses when relying on questionnaires alone. The act of swallowing involves a complex sequence of coordinated phases, beginning with the tongue’s crucial role in propelling food or liquid into the throat. Tongue strength is a pivotal early indicator of dysphagia. Assessing tongue strength offers a less invasive and potentially more accessible diagnostic approach. Existing market solutions, such as the Iowa Oral Performance Instrument and SwallowSTRONG (SS), are designed to evaluate tongue strength, providing valuable insights into dysphagia risk. However, these devices are often costly, with limited spatial resolution, and SS demands time-consuming and expensive customization. To address these limitations, we propose the development of an innovative, cost-effective, and user-friendly sensing platform designed to detect the early stages of dysphagia by precisely measuring tongue strength. Our platform features a thin-sheet pressure sensing array that can be readily placed on the palate, enabling accurate measurement of tongue strength. This sensor is constructed from biocompatible and low-cost materials. The array of sensing points delivers a detailed time sequence of tongue-palate contact and force during swallowing, enabling rapid diagnosis by a speech therapist, and guiding treatment, improving the likelihood of restoring normal swallowing. The accompanying visualization software will facilitate real-time data presentation and analysis, while the collection of extensive tongue strength data during swallowing provides a foundation for machine learning algorithms. These algorithms can offer deeper insights into tongue muscle activity, potentially transforming our understanding of the swallowing process and dysphagia.

Through the food we ingest, our intestine is exposed to multiple threats that include toxins and pathogens that cause intestinal damage, ultimately leading to increased risk to develop colorectal cancer (CRC). The first line of defense against these threats is the mucus barrier, a thick mucin-rich layer that coats the intestinal epithelium, thereby providing a physical protection for the gut. In the colon, the mucus layer is impermeable to the luminal intestinal microbiota, but it serves as an energy source for specific microbes evolved to metabolize the glycan rich mucus layer. Because of its critical importance for intestinal protection, the mucus barrier is attracting increasing attention, and identifying ways to increase the protection afforded by this barrier is a major long-term objective with exceptional biomedical and translational potential. However, research is severely hampered by the lack of robust tools to assess the mucus barrier in vivo, as it relies on poorly sensitive and indirect assays. In this proposal, we will explore a novel paradigm by leveraging transcriptional information provided by specific microbes that closely interact with and metabolize mucus. In Aim1, we will first precisely map the dynamic landscape of Polysaccharide Utilization Loci (PULs) which, in Bacteroides bacteria, represent gene clusters that are responsible for the metabolism of complex carbohydrates, including intestinal mucins. For this, we will : (i) colonize mice harbouring a defined synthetic bacterial community with Bacteroides massiliensis, a bacterium that is evolved to utilize host mucins, (ii) apply interventions (diet changes, infection) that directly modulate the mucus barrier and, (iii) identify using metatranscriptomics the genetic elements within B. massiliensis PULs whose regulation informs on the host’s mucus barrier dynamic, in vivo and in situ. With this information, in Aim2 we will next engineer B. massiliensis to incorporate a reporter system under the control of the genomic elements identified, to generate a “mucus growth biosensor” bacterium. This biosensor will be used to screen for prebiotics and postbiotics that promote mucus growth and protect against CRC in vivo. Together, this project will use an interdisciplinary approach to provide a novel paradigm for mucus research, which will have immediate translational potential and will help in the rational development of pro-health diets that can protect against infection and CRC.

Single-cell manipulation is a cornerstone of advanced biological research, essential for applications such as cellular surgery, targeted drug delivery, and diagnostics. This precise control over individual cells allows researchers to investigate cellular behaviour and properties, enabling breakthroughs in fields like personalized medicine and tissue engineering.

This project proposes a novel approach to single-cell manipulation by developing advanced micromachines specifically designed to operate in fluidic environments. By merging expertise in material science, nanomanufacturing, fluid dynamics, and biomedical research, we aim to create precision tools capable of interacting with individual cells in ways that existing technologies have struggled to achieve.

Our approach involves integrating nanoarchitected microstructures to enhance the performance of these micromachines, enabling them to perform complex tasks on living cells. By harnessing chemical reactions for actuation, we aim to develop innovative tools for cellular manipulation, surgery, and advanced diagnostics. Specialized materials with tailored properties will be used to gently and precisely manipulate single cells. These micromachines, fabricated using state-of-the-art nanomanufacturing techniques and scalable batch silicon processes, are engineered to function within environments similar to those found in the human body.

This research is characterized by its high-risk, high-reward nature. The high risk arises from the ambitious integration of diverse fields—nanomanufacturing, material science, fluid dynamics, and biomedical research—into a novel approach not previously attempted. However, the potential rewards are substantial, offering the possibility of revolutionizing how we manipulate and study individual cells, which could lead to significant advancements in personalized medicine, cellular surgery, and targeted drug delivery.

The feasibility of this project is supported by a robust interdisciplinary framework and the team’s collective expertise, ensuring that we can navigate the inherent challenges and deliver impactful results. By blending engineering and biomedical sciences, this project exemplifies the power of interdisciplinary research to create innovative solutions for complex biological challenges. If successful, our work could set new standards in single-cell analysis and manipulation, with wide-reaching implications for both scientific research and healthcare.

Background In 1974, Christopher Stone, a legal scholar, argued for granting legal rights to environmental entities, such as trees. Since then, many countries have enacted laws that recognize the rights of bodies of water, shorelines, and forests, granting them a status that could be protected by governments and upheld by courts. However, discussions on environmental protections have largely overlooked microbes, despite their foundational role as the first and most essential organisms on Earth, upon which all other life depends. Microbes form complex, interdependent communities known as microbiota, which are crucial to the survival and well-being of plants, animals, and ecosystems. The common perception of microbes as mere pathogens is misleading, as this applies to only a small fraction of microbial life.

Like many species, microbes face the threat of extinction. For decades, microbiologists have warned of the declining diversity of human microbiota in industrialized societies. Scientists link these changes to antibiotic use, urbanization, sedentary lifestyles, Caesarean births, modern diets, food production methods, and other practices that affect microbes and our exposure to them. This loss of diversity, known as dysbiosis, has been linked to various illnesses, including diabetes, asthma, autism, and Parkinson’s disease. Similar disruptions in microbiota have been observed in animals, insects, plants, and other organisms, leading to significant impacts on their health and ecosystems.

Objectives and Research Approach The primary goal is to develop innovative legal frameworks that draw on existing environmental laws with the latest microbiological knowledge. These new legal frameworks, created collaboratively by our research team, aim to both complement and, in some cases, replace current ones. As secondary objectives, we also plan to apply these novel frameworks to specific fields of microbiology, such as medical microbiology and microbial ecology. To broaden the impact of our findings, we are collaborating with relevant stakeholders.

Novelty and Expected Significance of the Work Recognizing that life and the well-being of ecosystems are inseparable from microbes, we must rethink our legal frameworks from the ground up, starting with microbes. This approach challenges long-standing legal precedents and could fundamentally change human activities that impact microbes, including agriculture, medicine, and resource extraction.

This trans-disciplinary project bridging neurobiology, cell biology, biochemistry, physical chemistry, and materials science aims to answer a crucial open question in biology – namely, how do living organisms sustainably fabricate biopolymeric materials with exceptional properties. Certain biological organisms such as spiders and mussels fabricate high-performance polymeric materials in an environmentally friendly way. Understanding these materials and their assembly processes has proven potential to improve the sustainability and materials properties of human-made polymeric materials (e.g., plastics, elastomers), but also for aiding design of new biomedical materials for tissue engineering and bioimplants.

The PI’s group has made significant recent progress in understanding the assembly processes underlying fabrication of mussel byssus – a high-performance self-healing, fibrous adhesive used by mussels to attach on rocky seashore habitats. Recent findings highlight the crucial role of stimuli-responsive fluid protein condensates and carefully controlled physicochemical inputs. However, there are two crucial missing elements in our understanding 1) the multiscale anatomy and 2) neurological control of the biological machinery used to produce the byssus. We lack an understanding of the byssus “factory” and the means by which the stimuli required for material fabrication are regulated.

The mussel foot organ is a biological factory for making high performance byssus fibers. Like any production facility, it consists of the machinery that produces that product and a control center that directs a well-orchestrated production process. However, in this case this machinery comprises living tissue and neurological networks. The goals of the current proposal are to 1) Map the anatomical machinery for fabricating the byssus at the molecular, cellular, and tissue-levels, 2) Elucidate the neurobiological control system that regulates the formation and subsequent controlled release of the byssus, and 3) Leverage the new insights gained from these investigations to facilitate previously unattainable in vitro thread assembly studies.

Combining the PI’s expertise in multiscale structural and compositional analysis of biological tissues and materials with the Co-PI’s expertise in characterizing and mapping neurological networks, we will embark on a high risk, high reward cross-disciplinary exploration of the mussel foot tissue fabrication of byssus.

Dental implantology is a non-specialist procedure, but adverse events are exponentially increasing. It is still unclear if this is due to increased use or a drop in success rate or both. In contrast to orthopedic surgery, Canada lacks a dental implant retrieval program, which would allow retrospective analyses including implant factors. Common factors for failure include general patient health, surgical factors, implant factors, and bone quality. Surgical protocols, including maximum torque, vary tremendously with implant type. Mechanical, chemical and biological processes are operational at the interface between the titanium alloy implant, bone and soft tissues. We hypothesize that corrosion and some surgical factors are instrumental in mechanical failure and the generation of detrimental wear particles. Titanium alloy corrosion processes result in the absorption of hydrogen atoms and the formation of oxides, which increases the risk of fractures and particle generation. The research objectives are to determine 1) whether corrosion is an important factor in dental and orthodontic titanium implant failure and 2) mitigation strategies to improve patient outcomes. Our experts in dentistry and orthodontics are complemented by a panel in materials science, leaders for orthopedic implant retrieval studies, and radiologists. Using our optical coordination machine, optical microscopy, and scanning electron microscopy, we will estimate the amount of lost wear volume from retrieved implants. We will also determine the amount and type of corrosion products formed, such as titanium hydride and oxides, by means of X-ray diffraction and electron microscopy. The duration of implantation, clinical complications (infection /inflammation), mechanical factors (angle, torque, loosening) will be evaluated. In-vitro tests, combining chemical and mechanical degradation modes of titanium alloy, will attempt to resemble the corrosion products and mechanisms found to prevail in vivo, so that fundamental insights into the dominant causes of the failure could be understood. Further investigation will explore the effect of corrosion, using an electrochemical pre-aging approach, on the mechanical stability of dental screws in bone. Research findings on dental/orthodontic implant corrosion will be published scientifically, communicated to the public, will inform digital planning tools as a surgical guide for dental surgeons and be included in undergraduate and graduate courses.

RELEVANCE Delivering medical supplies to rural Indigenous communities in Canada presents significant challenges, leading to health inequities. Drones offer a promising solution by transporting medical supplies, potentially bridging these gaps in healthcare access. By doing so, this project addresses these inequities and supports the Truth and Reconciliation Commission’s Calls to Action.

HIGH RISK The success of this initiative requires collective action and collaboration across multiple disciplines and communities. To tackle this issue, an implementation science approach is essential for identifying how partners can effectively work together. This project focuses on establishing relationships and common goals to enable interdisciplinary cooperation, addressing the unique needs of rural Indigenous communities.

OBJECTIVES The objectives of this project are to: 1) Co-develop an innovative model for using drones to deliver medical supplies in rural Indigenous communities 2) Co-implement the use of drone technology in three rural Indigenous communities 3) Co-evaluate partners’ experiences, outcomes and impacts.

APPROACH This project will be conducted in partnership with an interdisciplinary team of researchers, rural Indigenous community members, health administrators, health professionals, decision makers, and industry partners. Insights from our initial demonstration phase, environmental scans, and community needs assessment will inform our drone delivery model. Building relationships with the three rural Indigenous communities will be crucial for success, as their views and needs will guide the work. Semi-structured interviews will be used to identify barriers and facilitators, and mixed methods will assess implementation processes, outcomes, and impacts from different perspectives.

HIGH REWARD We will engage diverse partners to integrate drone technology into rural Indigenous healthcare systems. By building and strengthening relationships, this project will enhance healthcare access and delivery. We will establish best practices for future implementation, setting a new standard for rural Indigenous healthcare. Beyond healthcare, it offers capacity-building opportunities, contributing to economic development and innovation. Ultimately, our vision is for communities to own drones, enabling various applications, such as disaster management and wildfire monitoring, thus enhancing the overall impact.

The objective of this research is to develop a scalable methodology for creating architecturally scaled living building materials (LBMs), aiming to revolutionize sustainable design in the architecture, engineering, and construction (AEC) industry by integrating cyanobacteria biomineralization with advanced 3D bioprinting technologies. This collaboration between geomicrobiology and architecture holds the potential to permanently sequester carbon from the atmosphere.

Research Approach: Cyanobacteria Cultivation & Biomineralization: Utilize the biomineralizing properties of cyanobacteria, particularly their ability to induce calcium carbonate precipitation, to create bio-cements and other biogenic materials. 3D Bioprinting of Living Materials: Employ 3D bioprinting techniques, such as extrusion and inkjet printing, to deposit cyanobacteria-laden bio-inks in a controlled, layer-by-layer process, forming complex, customized structures with self-healing, adaptive capabilities, and carbon sequestration potential. Optimization and Scaling: Address technical challenges related to microorganism viability and distribution within printed structures, optimize the mechanical properties of the materials, and scale the technology for architectural applications. Environmental and Regulatory Considerations: Assess the environmental impact and regulatory challenges of using living materials in construction, ensuring that they meet safety and sustainability standards.

Novelty & Significance: This research represents a groundbreaking innovation in combining geomicrobiology with architecture. Integrating mineralizing cyanobacteria into 3D-printed structures offers several significant advantages: Sustainability: Sequesters carbon during construction, addressing the AEC industry's contribution to global carbon emissions and aligning with regenerative design principles. Self-Healing and Adaptation: LBMs can repair themselves and adapt to environmental changes, reducing maintenance needs and extending building lifespans. Customizability: 3D bioprinting enables the creation of highly customized architectural components, unlocking new design possibilities.

The successful development of scalable LBMs has the potential to transform building design and construction, making them more sustainable, resilient, and aligned with circular economy principles, thus contributing to a more sustainable future for the AEC industry.

One of the most common diagnoses in the intensive care unit (ICU), acute respiratory distress syndrome (ARDS) has an exceedingly high mortality rate (>40%). ARDS is the culmination of bacterial and viral infections like influenza, SARS, MERS and more recently, COVID-19. The diagnosis of ARDS is relatively easy: The presence of “infiltrates” on chest X-ray, reflecting the pulmonary edema that occurs due to leakage of the alveolar-capillary membrane. This diffuse interstitial edema makes the lung “heavy” and results in lung collapse in the dependent regions.

The continuous monitoring of this edema and its responses to therapies (drugs and/or mechanical ventilation) is however challenging, if not impossible. Patients require daily exposure to ionizing radiation through either 2D chest X-rays or 3D computed tomography (CT). The latter delivers the equivalent radiation dose to that of 70 chest X-rays, substantially increasing the risk of radiation-induced malignancies. To make matters worse, transport of intubated ICU patients to the CT radiology suites is inherently risky given how unstable critically ill patients can be.

To address this challenge, we will develop non-invasive, non-radiation-based imaging tools that revolve around combined ultrasound (US) and photoacoustic (PA) imaging. US is clinically available and widely accessible, but it does not get used for lung assessments due to the presence of air which blocks the US signal. However, in the lungs of ARDS patients, the edema generates fluid-filled regions of the lung that offer a conduit for the sound waves to travel. PA imaging is very similar to US, but instead uses non-invasive light pulse illumination to generate acoustic waves. These acoustic waves will travel only in the fluid-filled, injured lung, offering a novel path of access for examining a previously inaccessible organ. The addition of multi-wavelength laser illuminations permits the examination of functional information on the lung such as its oxygenation, perfusion, and metabolic rate of oxygen release. Furthermore, both US and PA can be performed at the bedside, obviating transport of the patient. We will develop ARDS US/PA biomarkers that will first be validated in preclinical models, before demonstrating proof of principle in relevant patient cohorts. Taken together, US and PA imaging offer an unprecedented level of non-invasive, radiation-free structural and functional monitoring of ARDS, an unsolved clinical problem.

During hemodialysis (HD), blood is pumped through a dialyzer consisting of a bundle of microfibers, which filters waste products from the blood. Despite improvements, most patients still experience side effects during HD. These side effects include reduced blood flow to the heart, low blood pressure during HD, a disturbance in capillary blood flow in vital tissues, disruption of immunological responses (T-cell exhaustion), and blood coagulation throughout the body. Moreover, due to the repetitive nature of the HD procedure, patients are subjected to these side effects multiple times per week, eventually leading to irreversible tissue and organ damage and increased patient mortality.

During HD, direct blood contact with the dialyzer membrane results in protein adsorption on the inside of the fibers, forming a bioactive layer on the fiber's interior wall, which can affect the filtering capacity and result in activating complement and coagulation pathways inside the dialyzer. Also, it generates bioactive compounds, which are returned to the central blood flow and lead to systemic activation of complement, coagulation, and disruption of immune responses (T-cell exhaustion). By chilling blood inside the dialyzer, we aim to inhibit this bioactivation, reduce the observed side effects, and improve the procedure's biocompatibility.

This project focuses on a novel approach using a newly designed hemodialysis method to reduce these adverse effects and improve the procedure's biological tolerability (i.e., biocompatibility) by chilling the blood flowing inside the dialyzer to 5 ̊C and rewarming it to 37 ̊C before returning it to the body. We recently developed an animal model that allows dialyzing small laboratory rats, while observing surgically exposed muscle tissue through intravital microscopy. This model allows us to investigate how microvascular blood flow is affected by HD, while blood sampling enables us to investigate which bioactive molecules and immune responses are involved. Since these miniaturized dialyzers are built in-house, we can easily change the design or procedure to investigate the physiological effects of the procedure.

Understanding which factors are essential for dialyzer biocompatibility is of great importance. The ability to improve HD biocompatibility will not only enhance the effectiveness of the procedure; it will also reduce the irreversible side effects that lead to cardiovascular complications and patient mortality.

Background: Uropathogenic Escherichia coli (UPEC) cause most urinary tract infections (UTIs), with over 300 million cases globally per year primarily in female patients. Treatment consists of antibiotic therapy, but reinfection in >25% of cases leads to recurrent UTI (rUTI), often seeded by UPEC persisting in the gut microbiome. rUTIs are a significant unmet clinical need, with standard of care consisting of chronic antibiotic treatment. Besides leading to resistance, this ironically can increase infection susceptibility over time by damaging commensal microbes, highlighting the need for new treatment approaches.

Objectives/Approach: We will combine approaches from synthetic biology and infectious disease research to pioneer an antibiotic-free strategy for treating rUTI, rationally engineering synthetic anti-UPEC microbial therapeutics (synAMTs) that actively target and deplete UPEC in the gut microbiome. This involves two complementary objectives: 1. Engineer synAMT prototypes for selective UPEC depletion: We will prototype synAMTs by modifying E. coli Nissle 1917 (EcN, a gut commensal probiotic) to actively target UPEC reference strain UTI89. We will program EcN to express surface-displayed nanobodies to selectively bind UPEC-specific surface biomarkers, along with a type VI secretion system (T6SS) to inject toxic effectors and kill bound UPEC, applying modular strain engineering approaches to enable flexible redesign and iteration of synthetic nanobody/T6SS components. 2. Develop UPEC-microbiome culture platform for synAMT evaluation and optimization: We will establish microbiome culture platforms based on UPEC-positive patient fecal samples to serve as ex vivo disease models that validate synAMT performance within a realistic complex microbiome, as well as to isolate and sequence clinical UPECs for bioinformatic identification of conserved surface and T6SS immunity biomarkers. This rapid feedback and insight will inform an iterative design-build-test-learn pipeline to engineer optimized synAMTs as candidates for animal and clinical trials against rUTI.

Novelty/Significance: This high-risk interdisciplinary collaboration between synthetic bioengineers and infectious disease researchers pioneers a novel paradigm of rationally targeted clinical microbiome editing against bacterial pathogens, and delivers high rewards by offering an alternative to chronic antibiotic use for the millions of Canadians who develop rUTI in their lifetimes.

Our ongoing stream of thoughts dynamically unfolds over time. Impairment in the capacity to regulate these internal thoughts have been observed in obsessive-compulsive disorder (OCD), autism spectrum disorder (ASD), and attention-deficit/hyperactivity disorder (ADHD), among others. We propose that thought regulation may be a transdiagnostic process that underlies co-occurring internalizing and externalizing disorders.

Our knowledge of clinical disorders is largely based on the nomothetic approach, in which conclusions are derived from aggregated data in a clinical group. These conclusions often do not apply to individuals. Recent studies have implemented the idiographic approach, in which detailed information is acquired from an individual to make accurate personalized predictions. For example, ecological momentary assessment uses multiple daily surveys spanning days to capture experiences across time and context. This approach can facilitate an in-depth exploration of dysregulated thought control in an individual in the real world.

Our project aims to determine 1) if the capacity to regulate thought content is a transdiagnostic process that underlies co-occurring internalizing and externalizing disorders, and 2) if we can make person-specific predictions of symptom severity and affective well-being using the idiographic approach to capture thought regulation.

We will recruit individuals with OCD, ASD, ADHD and controls. They will complete six surveys daily for one week that assess their level of thought regulation, symptom severity and affective well-being. We will examine the relationship between thought regulation, symptom severity and affect at two levels using statistical and machine learning analyses: comparing the four groups (nomothetic) and predicting patterns within individuals (idiographic).

Our interdisciplinary approach integrates new conceptual perspectives and methodologies on the regulation of thought content as a transdiagnostic process underlying clinical disorders, by combining expertise from pediatric medicine, cognitive science and computational science. Although identifying a novel, ecologically valid transdiagnostic process with an underused approach is high-risk, these risks are worthwhile due to its anticipated far-reaching impact and potential for applications in other disorders. Our study will develop a new diagnostic approach targeting a naturalistic cognitive process that may play a critical role in patient well-being.

Climate change has significantly impacted the globe, causing biodiversity loss, water scarcity, extreme poverty, urban inequity and food insecurity. Cities, as hubs of human activities, consumption, pollution, and waste, must be the focus of climate mitigation and adaptation strategies. While innovative engineering technologies like solar panels, water recycling, smart grids, transport electrification, and green building materials enhance sustainability, their impact is effective only when paired with social innovations that drive behavioural change. Social innovations extend engineering solutions to include community retrofitting programs, citizen social science, community-building, gamification and incentives that encourage the adoption of these sustainable technologies, fostering community buy-in.

Unlike most climate “top-down” mitigation strategies, this proposal advocates a paradigm shift to a “bottom-up” citizen-centric social innovation approach. The objective is to ensure sustainable engineering solutions are widely adopted by anticipating the collective behavioural responses of citizens and communities. Modelling and analysis of citizens' responses, attitudes, perceptions, socioeconomic, demographics, income, and social inequalities are key for crafting sustainable equitable solutions, strategies and policies. By leveraging tools from social sciences and engineering—complexity theory, complex network theory, welfare economics, ethics and morality, strategic foresight, science and technology studies, environmental and design justice, behavioural modelling, game theory, computer simulation, policy analysis, actor-network theory, AI and ML, citizen social science, and agent-based modelling—we will develop holistic, socially equitable solutions that ensure sustainable technologies and policies are systematically adopted by diverse communities. This interdisciplinary approach is instrumental, as the climate crisis cannot be addressed by isolating its components. The novelty of this work lies in its unprecedented integration of cutting-edge technologies with social innovations to create adaptive, citizen-centric solutions. Engineering micro-to-macro interactions that co-create desired behaviours is high-risk due to the unpredictable outcomes inherent in "wicked problems", yet offers high rewards with the potential for transformative, systematic solutions. Success will be measured through microsimulation, surveys, and scenario and sensitivity analyses.

Canada is renowned for its natural beauty, but increasingly severe wildfires are impacting humans, wildlife, and the ecosystem. Novel, interdisciplinary solutions are required to facilitate habitat restoration, particularly for endangered species, such as the whitebark pine (Pinus albicaulis; “WBP”), which is declining due to white pine blister rust. Destruction of rust-resistant trees by fire has the potential to tip this fragile species into extinction, triggering cascading ecosystem impacts. The primary objective of our OneHealth research approach is to explore how natural ecosystem services may be leveraged to facilitate the restoration of WBP forests destroyed by severe fire.

The Clark’s nutcracker (Nucifraga columbiana; “CLNU”) may play a critical role in restoring WBP forests. WBP rely on CLNU as obligate seed dispersers, which cache more seeds than they retrieve. Seeds that remain grow into future WBP forests. CLNU preferentially cache in open areas, such as occur after severe fire, but this behavior may depend on personality traits like boldness. Tourists and local residents supplementally feed CLNU, typically attracting bolder individuals. This may provide birds with additional resources and enhance public support for conservation efforts, but may also reduce dispersal of WBP seeds.

Our specific objectives are to 1) test whether CLNU prefer to cache seeds in burned areas, initiating WBP forest regeneration; 2) understand how food supplementation and avian personality influence CLNU movement and caching, to evaluate impacts on WBP seed dispersal; and 3) evaluate how human attitudes and behaviors affect the WBP ecosystem.

Our approach will test an interdisciplinary integration of specialized technology to track CLNU seed dispersal with avian personality assessments and human dimensions data collection. Additionally, we will develop data collection methodologies permitting inclusion of individuals with mobility challenges, a limitation for many field research programs. This project has potential for high-reward, through enhancing ecosystem restoration, reducing conservation costs, and building community involvement in forest restoration. Due to the novelty of this approach, potential risks also exist. For instance, CLNU may not prefer burned areas for caching. However, other risks are minimal as our interdisciplinary team has significant expertise with the methodological approaches necessary to develop this novel OneHealth study.

The overarching goal of this project is to address two pressing health issues in Saskatchewan: Indigenous youth mental health and wellness, and underrepresentation of Indigenous health professionals. Using community-based participatory research methods, this project will co-develop culturally appropriate virtual reality (VR) resources for Indigenous youths. Specifically, the research objectives are to co-develop: 1) a culture-based brain anatomy and neuroscience VR module with linkages to mental health and wellness challenges, and culturally relevant strength-based resources identified by community members. 2) a VR orientation module with narrated 360-degree videos to introduce Indigenous youth to health profession career options.

VR is a cutting-edge educational technology with transformative potential. However, there is a critical gap in VR resources. Current VR development tools and stock images are overwhelmingly Western-centric, with few resources available to authentically represent Indigenous culture. Through co-development with Indigenous youth, this project embodies a bold departure from conventional anatomy pedagogies. Teaching anatomy, traditionally steeped in Eurocentric perspectives, has rarely been reimagined through an Indigenous lens.

Our approach represents a high-risk, high-reward innovation by breaking away from traditional VR design paradigms and prioritizing Indigenous youth knowledge and preferences for health education. However, involving youths as researchers carries the risk of fluctuating engagement, compromising project timelines and potentially jeopardizing the success of the project. Despite these risks, the rewards of collaborations with health scientists, educators, community members, and Indigenous artists will result in VR modules that are scientifically sound and transformative, fostering mutual growth and innovation while being deeply relevant and respectful of Indigenous perspectives. This pioneering effort will not only address the lack of culturally appropriate health education resources but also pave the way for future VR developments. The anticipated outcome is a profound impact on Indigenous youth health education and the establishment of a new standard for culturally inclusive VR tools. As a key component of a pipeline model from secondary to post-secondary education, the VR orientation module will be crucial in fostering interest in health professions and capacity for future Indigenous healthcare providers.

Background: Infection and inflammation are major complications associated with implantable and retrievable medical devices that come into contact with blood. Due to the lack of technological advancements, it remains a significant unsolved clinical problem, complicated by the prevalence of multi-drug resistant bacteria. Given that at least 30-40% of hospitalized patients undergo procedures involving vascular devices/implants (e.g., catheters, hemodialyzers, blood filters, stents, aortic valve, pacemaker, defibrillator etc.), this is a significant problem given the high risk for bloodstream infections and mortality. Current device technologies are not effective due to a lack of broad-spectrum antibiofilm activity and the initiation of life-threatening thromboembolism.

Mortality following vascular device implantation is driven by inflammation and thrombosis via the activation of the contact pathway, and by bacterial biofilms on their surface. Strong evidence exists that these life-threatening complications can be prevented by minimizing the initiation of these key events, however, no technologies currently exist. We hypothesize that novel active antibiofilm and coagulation proteins stabilizing device surfaces can prevent implant/device complications and premature device failure.

Objectives: 1. Identification of novel surface specific blood-compatible and broad-spectrum antibiofilm peptides (ABPs). We will screen and identify novel peptides with high activity on surfaces and stable in the blood environment. 2. Development of a novel anti-thromboinflammatory surface that stabilizes and prevents activation of coagulation factors. We will screen novel surface coating compositions with specific affinity for coagulation factor XII providing stabilization and preventing the contact activation. 3. Development of a combination device consisting of an anti-thromboinflammatory surface with optimal ABPs, and its in vivo validation. We will use the optimal combinations to test device performance in blood to prevent biofilm formation and thromboinflammation.

Significance and Novelty: The proposed durable surface-specific antibiofilm and thrombosis-resistant devices will be the first of their kind, incorporating a novel mechanism of action to prevent a challenging clinical issue. Collectively, this has significant potential to improve the health of Canadians suffering from device-associated bloodstream infection and thromboembolism.

Nonspeaking autistic people (nonspeakers) can learn to communicate by typing with a single finger, offering an effective alternative to speech. However, they face significant challenges in controlling the precise finger movements required for typing. These motor challenges, along with attentional and sensory difficulties, mean that learning to type often takes years. While human therapists can help, this training is costly, not widely available, and can create dependency on specific individuals. As a result, few nonspeakers access the training needed to learn to type. This project addresses this gap by developing an innovative, affordable, automated system providing adaptive typing training that addresses the unique motor, attentional, and sensory challenges nonspeakers face. The collision of autism science and technology in this project unique, combining for the first time the lived experiences of nonspeakers with cutting-edge AI. Through a community-based participatory research framework, we will directly collaborate with nonspeakers who have learned to type, to create a highly personalized learning system — an approach not explored before.

The proposed solution is a tablet-based system that delivers engaging multimedia lessons from a freely accessible community-contributed library. Integrated with a large language model, the system will use the lesson context to generate questions whose difficulty is dynamically adjusted based on the user’s history and real-time performance. Users will respond on a keyboard that employs AI to dynamically adjust the number and salience of keys available, guided by historical performance and real-time hand- and eye-tracking data. The system will also use this tracking information to provide audio-visual prompts to initiate and guide finger movements, offering attentional and regulatory support to maintain engagement. In Year 1, the system will be co-designed and tested with a small cohort of nonspeakers and practitioners. In Year 2, a longitudinal study will evaluate improvements in typing skills and communication.

This project is high risk because it attempts to create a complex, AI-driven system that addresses deeply individualized challenges in motor control and communication—a level of personalization that has not been achieved before. However, the potential rewards are substantial: accelerating the acquisition of typing and communication skills, thereby transforming the lives of millions of nonspeakers worldwide.

This interdisciplinary research tackles a significant challenge in non-invasive cancer therapy: the need for precise and quantitative imaging guidance during procedures such as High-Intensity Focused Ultrasound (HIFU) and Radiofrequency Ablation (RFA). While HIFU and RFA offer significant promise as alternatives to traditional cancer treatments like surgery and chemotherapy, particularly in the treatment of solid tumors, their broader adoption is hindered by the limitations of existing imaging modalities. During these procedures, clinicians must depend entirely on medical imaging and their expertise, as direct interaction with the targeted tissue is not possible. However, current imaging technologies present notable drawbacks: MRI, though providing high-resolution quantitative data, is prohibitively expensive and impractical for regular use in operating rooms, while ultrasound, despite being more accessible, lacks the precision and quantitative detail for effective treatment.

This research aims to overcome these limitations by developing an advanced imaging solution that integrates the strengths of both MRI and ultrasound, thereby enhancing the efficacy and precision of non-invasive cancer therapies. Our innovative medical imaging system will merge AI, advanced computational modeling, and ultrasound technology to deliver MRI-level accuracy and resolution in a more accessible and cost-effective format. This novel approach will enable real-time visualization of tissue ablation with detailed quantitative information during HIFU and RFA procedures, significantly improving treatment guidance and accuracy.

This high-risk, high-reward project seeks to establish a new paradigm for non-invasive cancer therapy by equipping the clinicians with detailed real-time intraoperative information and procedural guidance for optimal treatment outcomes. Our interdisciplinary research team, comprising experts in AI, biomedical engineering, and oncology, is uniquely positioned to execute this ambitious endeavor.

The project's transformative potential lies in its ability to enhance the precision and effectiveness of image-guided procedures, reduce recovery times, minimize side effects, and ultimately improve patient outcomes. By proposing a novel approach that transcends traditional disciplinary boundaries, this project aligns with the NFRF’s mandate for transformative, interdisciplinary research and has the potential to set a new standard in non-invasive cancer therapy.

Despite potential risks to clinicians and their delivery of care to patients, little attention has been devoted to the load and heat problem inherent in providing wearable x-ray protection. Many hospital procedures are performed in rooms with x-ray imaging, requiring staff to wear lead aprons. Despite its effectiveness in x-ray protection, wearing heavy lead aprons (12-23 lb) for long durations, often in poor postures, contributes to excessive spinal loading, that can result in fatigue and injury. Work-related musculoskeletal disorders are common among healthcare professionals, leading to work absences, back surgery and, for some, early termination of their careers. The addition of lead aprons on top of the surgical gown further creates challenges with respect to temperature and moisture transport properties, resulting in excessive heat exposure, sweat and discomfort. Fatigue, due to load, and discomfort, caused by heat, have been shown to decrease proficiency in activities requiring high physical and cognitive load. As such, there is a need to better address the heat and load issues of personal protective equipment (PPE) of radiation. Inspired by innovations for material handling tasks, recent efforts to redistribute the weight of lead aprons with exoskeletons raised some hope. However, despite better load distribution, exoskeleton studies in operating rooms found them to be bulky, interfere with range of motion, and without evidence of effectively reducing discomfort, load, or fatigue.

Our interdisciplinary team is composed of engineers/scientists, clinician-investigators (in orthopedics and cardiology), a designer and an x-ray imaging specialist. We propose a holistic approach to combat issues of heat, load and radiation safety simultaneously within the clinical environment through engineering a novel modular PPE as a form of customizable technical apparel. This would be accomplished as a two-layer solution: 1) an optimized outer lead apron design to provide essential radiation protection and minimize spinal loading and 2) an inner layer wearable undergarment designed with passive textile actuators (that can indicate and offset loading with poor posture), with embedded micro-fluidics channels to facilitate air circulation to maintain body temperature. The proposed novel PPE is a high-risk effort, that if successful, will improve comfort and reduce musculoskeletal injury risks for healthcare workers, leading to better patient care delivery.

Spatial transcriptomics (ST) technologies have been widely used in constructing spatial tissue atlases. ST holds a great promise to characterize the tumor microenvironments (TME) and identify therapeutic genes that can help re-shape the tumor transcriptome landscapes and restore normal homeostasis. However, the current ST platforms do not profile the entire transcriptome, are not at single-cell resolution, or do not capture large tissue areas that can be used to discover interactions between the tumor and its microenvironment that contribute to cancer progression. The full potential of ST data could be unleashed by leveraging two types of large data repositories: (1) large pathology image databases, encompassing over hundreds of thousands of high-quality histological images; (2) large single-cell transcriptome databases, encompassing over tens of millions of cells for a wide range of cell types and tissues. While artificial intelligence (AI) foundation models (FMs) have been developed to train on these big data, there is a lack of multi-modal AI framework that can bring to bear the transfer learning capability of both FMs. Inspired by the fact that histology images often come with the paired ST data especially for spatial TMEs, we propose an AI framework called AI-oncologist for integrating histology images and spatial transcriptome by jointly training two FMs pre-trained on histology image data and single-cell transcriptome, respectively (objective 1), imputing super-pixel resolution gene expression from pathology images (objective 2), and in-silico gene therapy to knockout candidate genes to disrupt the equilibrium state of TME (objective 3). By detecting the tumor driver genes for each patient, AI-oncologist can make personalized recommendation for subsequent treatment regimens. Our long-term vision towards personalized oncology is to develop an AI-assistance to aid the oncologists in designing personalized cancer treatment to individual patients based on their histopathology and gene expression profiles. Our project will lay the foundation for the national and international funding opportunities.

Agriculture has a unique position in the context of climate change. It is both a massive contributor to greenhouse gas emissions and environmental degradation, and a potential mitigator of damaging effects through transformative land-use practices and carbon sequestration. A more sustainable and resilient agricultural system is critically needed and requires the development of innovative practices that address both the unique stressors impacting crop production and agriculture's impact on climate. To this end, our project explores the role of plant "memory" in enhancing resilience and adaptive capacity in sustainable farming systems.

Plants have the ability to temporarily store information of past environmental stresses at a molecular level and potentially pass this on to new cells and even future generations through epigenetic changes. This allows plants to be more responsive to future environmental triggers and threats. Despite this adaptive capacity much remains to be learned about the exact mechanisms of epigenetic processes and the potential of epigenomic diversity to enhance plant yield and performance. To explore this unique line of inquiry we pursue the following objectives:1)Through collaborations with farmers catalogue the imprints of diverse farm practices on plant “memory”(epigenomes); 2)Analyze to what extent people-plant engagements and approaches to plant care shape a plant’s epigenome to determine if there is a direct link to plant yield and performance; 3)Establish what role epigenetics play in increasing the adaptive capacity and resiliency of cropping systems; 4)Develop a unique transdisciplinary framework that combines methods from cell and systems biology, sociocultural anthropology and agrarian citizen science to create new methodological and theoretical paradigms for understanding human-plant relations; 5)Challenge simplified narratives of climate change and sustainability to situate the agency and essential contribution of plants.

Agriculture in Canada continues to be driven by productivist models which have been targeted as particularly unsustainable due to their reliance on chemical inputs, patenting of seeds, and impacts on biodiversity. Our novel transdisciplinary approach will produce the first comprehensive collection of farming imprints on crop epigenomes and transform conventional thinking about the role of human-plant relations and the possibilities of genomics to enhance the sustainability of agriculture.

Multimorbidity refers to the coexistence of two or more chronic diseases. Children with multimorbidity (e.g., asthma, obesity, cardiovascular disease, mental health) account for a large proportion of healthcare expenditures and pose a substantial burden for their families, the healthcare system and society. Current knowledge of multi-morbidity in childhood is sparse, hampering efforts for developing effective interventions and treatments. Understanding and addressing multimorbidity in children is crucial for preventing and improving their health outcomes, quality of life, and overall well-being.

The overarching goal of this proposal is to advance the understanding of multimorbidity development in early life through adolescence and inform strategies for predicting and mitigating long-term health risks. Our innovative study will leverage extensive medical, biological, social determinants and environmental data sets from large Canadian birth cohort studies, combined with novel data science and artificial intelligence approaches.

This study has three specific aims: (a) Aim 1: To characterize the interaction and progression of multimorbidity over time from 5 to 13 years of age. (b) Aim 2: To quantify the combined effects of the exposures shaping multimorbidity through an exposome approach (c) Aim 3: To evaluate the replicability and generalizability of findings from Aims 1 and 2 in an independent Canadian cohort.

This is an ambitious and high-risk study. This is the first study to apply machine learning approaches to understand multi-morbidity in large birth cohort studies. The large sample size, high-dimensional, multi-source and longitudinal nature of the data pose substantial challenges in model specification, computational complexity and results interpretation. This project is highly rewarded for mitigating the development of chronic diseases, significantly improving the quality of life and reducing healthcare costs; this proposal could transform diagnostic and prevention approaches, leading to better health outcomes for children with multi-morbidity; It will also enhance clinicians’ decision-making processes, leading to more personalized and effective healthcare interventions. We have also assembled an interdisciplinary team with expertise in epidemiology, pediatrics, environmental science, data science, and microbiology. This project will refine our understanding of multimorbidity in children, and the exposome profiles shaping multimorbidity.

Artificial intelligence holds tremendous promise in drug discovery. We propose to apply artificial intelligence (AI)-driven computer vision methods for phenotypic characterization of the model organism C. elegans to assign gene function and identify novel drug leads. The analysis of mutants has been instrumental in furthering our understanding of developmental and cellular processes, identifying gene function, and delineating cellular signaling pathways. Phenotypic characterization is also a powerful tool used to identify drug targets. When the action of a drug phenocopies a mutant, this often identifies the molecular target enabling downstream development. While phenotypic characterization can be powerful, it is often challenging when phenotypes are subtle. We propose to use AI to meet this challenge. Computer vision can be used to recognize patterns that are not obvious to the eye. We will develop a system to phenotype animals by first generating a library of images of known mutants and drugs with known action to serve as a training set. Using recent advances in computer vision we will develop methodology to characterize the phenotypic appearance of a panel of mutant animals. Furthermore, we will perform similar AI training analysis on worms exposed to a library of FDA-approved drugs. Using the prototyping signature developed during training, we will identify specific phenotypes in a set of images of animals exposed to our unique library of natural products, thereby establishing the phenotypic impact leading to new drug candidates. Specific Aims Generate an image library using C. elegans mutants with and without reported phenotypes and worms exposed to FDA approved drugs. This library will serve as the training set to identify and characterize the visual signature of different phenotypes, or their prototype in a very dense feature space Use computer vision approaches to identify phenotypes in A) mutant animals with no known phenotypes and B) animals exposed to a library of natural products.

Significance: Our approach will enable the rapid phenotypic analysis of C. elegans. Our image library generated in Aim 1 can be expanded and used in many applications. Additionally, we will identify unreported phenotypes for some mutants and identify natural product extracts capable of producing effects that can be matched to phenotypes resulting from disruption of known pathways.

Background: Immunotherapy leverages the ability of immune cells to kill cancer cells. However, the immunotherapies currently offered are very expensive (~$500,000/patient) and are not readily available (> 2 months wait time). One way to reduce the cost and time to treatment is by developing “off the shelf” immunotherapies using cord blood derived hematopoietic stem and progenitor cells (HSPCs). The goal of our project is to develop a novel engineering platform that will enable us to manufacture “off the shelf” immune cells from cord blood and identify the key socio/economic factors across Canada that are critical for clinical translation. This will be achieved as follows:

Aim1: To understand the experience of patients undergoing immunotherapy. We will conduct interviews with patients who have received immunotherapy, to understand their knowledge of these therapies.

Aim 2: To use machine learning to develop small molecule cocktail (SMC) libraries for manufacturing “off the shelf” immunotherapies from cord blood HSPCs. We identified small molecules that promote production of natural killer (NK) cells from cord blood HSPCs. Next, we will combine these molecules and use modeling and machine learning to identify the concentrations at which they work in a synergistic manner to promote enhanced production of NK cells. Lastly, we will test the ability of the NK cells to kill cancer cells and characterize them.

Aim 3: To identify the barriers and opportunities that exist in the process of bringing new socially responsible immunotherapies to patients. We will conduct a scoping review of literature, interviews of the key social actors in this space (e.g: clinicians, scientists, patient advocates), and perform economic assessment of the commercialization pathway and identify alternative pathways.

Aim 4: How do cord blood derived immunotherapies influence cord blood donor behavior. Given that off-the-shelf immunotherapies will be developed from cord blood, we will investigate cord blood donor perspectives on uses of their donation in developing for profit immunotherapy products.

Significance: We will develop a novel engineering platform that will allow us to produce next generation immune cell therapies that are potent, safe, and readily accessible to Canadians at a low cost. Our work will also identify the key socio/economic factors across Canada that are critical (but are neglected) for clinical translation and promote sustainable innovation in the long-term.

The objective of the proposed research is to develop stimuli-responsive biomaterials that can enable improved detection and diagnosis of endometriosis. Endometriosis is estimated to affect one in ten women of childbearing age and is a condition where endometrial-like tissue develops outside of the uterus where it can cause inflammation and debilitating pain. Endometriosis is treatable through menstrual suppression or surgery, but commonly goes undiagnosed for 7-10 years. Societal norms stigmatize menstruation making it more difficult for women who may have endometriosis to receive the treatment they need without validation of a diagnosis. Yet, current diagnosis of endometriosis requires a surgical procedure. Moreover, early-stage endometriosis can be missed during surgical diagnosis which can result in both physical and mental health ramifications for patients. There is an urgent need for minimally invasive technologies to improve detection and diagnosis of endometriosis.

Biomaterial-based sensing has the potential to enable paradigm shifting technology for minimally invasive endometriosis detection and diagnosis. We aim to develop a stimuli-responsive microbubbles that can be injected into the peritoneal cavity and used to identify endometriosis using imaging. The microbubbles will produce a visible ultrasound signal only in the presence of endometriosis biomarkers in the peritoneal cavity. While there are no established blood or urine biomarkers for endometriosis, there are several markers associated with endometriosis found in the peritoneal cavity that we hypothesize will have more predictive power if measured directly from the peritoneal fluid. For example, markers of oxidative stress are upregulated in the peritoneal cavity where endometrial growths have developed. Our biomaterial sensors will enable local detection of these markers of endometriosis. Our approach combines expertise in biomaterials, gynecology, and imaging to engineer a minimally invasive technology for endometriosis diagnosis. If successful, our technology can provide an alternative to surgery for endometriosis diagnosis. Further, the sensitivity of our technology can be tuned to improve detection of early-stage endometriosis, which would result in earlier medical, surgical, or allied health interventions for improved quality of life for patients.

Low back pain (LBP) is the leading cause of years lived with disability, and in 2020, there were more than half a billion prevalent cases worldwide -an increase of 60.4% since 1990- most of which is observed in low and middle-income countries. Most cases of LBP are non-specific, characterized by biophysical, psychological, and social dimensions affecting function, societal participation, and financial well-being.

The World Health Organization’s efforts to implement health equity policies aim at ensuring the availability, accessibility, affordability, and quality of prevention strategies, treatments, and healthcare services, but challenges persist. Even in the most egalitarian societies, individuals with LBP encounter disparities in care, resulting in unequal health outcomes. Persistent inequalities related to sociodemographic determinants such as age, sex, race, ethnicity, and socioeconomic status significantly influence access to care and care pathways for patients suffering from LBP.

Despite the burden of LBP and the impressive corpus of knowledge generated in the past 30 years, specific populations remain underrepresented in LBP studies. For instance, older adults are often excluded, and individuals member of ethnic and cultural minorities are commonly underrepresented, leading to a lack of generalizability of study results to these groups and potentially resulting in disparities in health outcomes.

Addressing the underrepresentation of sociocultural and demographic diversity is essential to ensure that LBP prevention and management strategies apply to a broader range of individuals. Therefore, the aim of this study is to identify the individual and sociocultural determinants of LBP trajectories among individuals representing the diversity of the Canadian population.

To achieve this objective, a cohort study of individuals with LBP will be conducted in the province of Quebec. Partnerships with community-based organizations will help ensure that an equivalent number of individuals from underrepresented groups are recruited. For each individual, a set of sociodemographic, clinical, and health services utilization outcomes will be collected to [1] establish LBP trajectory profiles and [2] identify the determinants of each profile. Advanced machine learning techniques, including clustering and classification algorithms, and predictive modeling will be used to enhance our understanding of LBP trajectories in underrepresented populations.

Conventional ‘Western’ (or Eurocentric) Science (WS) is deeply entrenched within modern academic research institutions; as the dominant colonial knowledge system, WS has historically been viewed as the gold standard for scientific inquiry. Indeed, WS enjoys unequivocal precedence and privilege over other knowledge systems, including Indigenous Knowledge (IK). Ironically, WS has long benefited from the appropriation of IK. To address these issues, ethical co-creation practices involving Indigenous experts and communities—who hold Traditional Ecological Knowledge (TEK)—are crucial.

The proposed research aims to explore how to effectively interface IK-WS through innovative educational methods. These methods include integrative pedagogy and novel forms of cross-institutional accreditation (e.g., microcredentials, IK-WS pathway programs, accredited land-based activities). Through the development of a framework, this work will support transformative approaches to collaboratively conduct and teach science. We envision that our methodology-based framework will: 1. Create novel pathways for Indigenous youth to experience WS in ethically safe ways with central regard for community, culture, & IK; 2. Increase the level of exposure Indigenous & non-Indigenous students have to IK-TEK; 3. Improve local institutional capacity to provide IK-WS programming while expanding their autonomous research capabilities.

To improve meaningful interdisciplinary engagement with TEK, and to systematize co-creation in a way that ethically safeguards IK, local Indigenous-led and community-centered institutional research capacity is needed. Six Nations is capacity-building through the development of a Haudenosaunee-led Research & Training Institute. Key objectives of the Institute include 1. offering experiential learning & field research; 2. developing inclusive & equitable accreditation for community; and 3. training students & community members in research that centers community needs & priorities.

We emphasize that the proposed research is not simply curriculum development, but will support the pedagogical goals of the institute. Developing the framework for IK-WS knowledge interfacing and programming will be carried out through participatory engagements at Six Nations with support from an existing network of Indigenous and non-Indigenous stakeholders. The goal is to facilitate community-led resilience that is rooted in connection to land, place, culture, and reconciliation.

Aims. The overarching goal of our proposal is to measure the capacity of small intestinal (SI) microbiota to metabolize and limit the efficacy of L-dopa, pillar of medical treatment in people with Parkinson’s disease (PwPD). We will interrogate non-invasively SI metagenomes to explore taxa expressing enzymes metabolizing L-dopa, then analyse L-dopa metabolism in the SI of PwPD, correlating it to bioavailability of, and patients’ acute responsiveness to, oral L-dopa. We will develop and validate this detection platform in a representative sample of PwPD.

High risk. A pathway sustained by different SI bacteria metabolizes L-dopa reducing its absorption and brain availability. In humans, this microbiome pathway has not been explored at the SI, due to the need of invasive, endoscopic sampling. A novel alternative is the SIMBA capsule (Nimble Sci., Calgary), a pH-based sampling system that transits the SI, capturing luminal fluid for multi-omics analyses. This technology offers the unprecedented opportunity to assess in vivo the impact of SI microenvironment on the bioavailability of L-dopa released by both tablets and L-dopa-loaded SIMBA capsules.

High reward. PD is increasing in prevalence worldwide. Most patients experience fluctuating response to L-dopa, partly due to limited SI absorption. This project goes beyond standard surveys of biomarkers of response, developing and validating a direct, in vivo quantitative detection system for bacterial degradation of therapeutic L-dopa. Our findings will inform clinical practice-changing algorithms to guide timely recommendation of therapies bypassing the gut. Furthermore, we will funnel our findings into enzyme inhibition-based drug discovery platforms to provide new therapeutics or microbiota manipulation strategies to optimize bioavailability and response to L-dopa.

Interdisciplinarity. The quantification of therapeutic L-dopa metabolism in the SI microenvironment requires the development of a detection platform that will leverage the versatility of the SIMBA system. The collaboration between the selected disciplines (neurology, metabolomics, microbiome science and nanodelivery biomedical engineering) is essential to adapt the SIMBA system to measure the SI metabolic capacity of L-dopa in experimental (L-dopa-loaded capsule) and naturalistic (L-dopa tablet) settings. Leveraging expertise in the selected disciplines, our team is uniquely positioned to begin unraveling the complexities of drug response in PD.

There is a growing awareness in research and policy circles of the material requirements of low-carbon energy technologies such as wind turbines, solar photovoltaics, and electric vehicles. Material and waste management policies, often called ‘circular economy’ policies, have been proposed to curb the demand and waste outcomes of traditional decarbonization policies, particularly for critical minerals, as well as to further reduce emissions by promoting a broader life-cycle analysis perspective.

The emissions reduction potential of circular economy policies, however, remains unclear. The study of emissions-focused and material-focused policies continues to be siloed, with the development and use of policy assessment modelling tools isolated within specific economic and engineering sub-disciplines. Policymakers urgently require next-generation models to better understand interactions between circular economy and decarbonization policies, both in terms of energy use/emissions and material demand/waste. The real-world policy implications from developing such models are therefore significant.

Yet advancing this kind of interdisciplinary modelling is methodologically and logistically challenging, particularly in the compressed timelines needed to inform current policy development. It is also an inherently interdisciplinary problem, requiring innovative research at the modelling-policy nexus across fields including engineering, economics and political science. We propose to undertake this challenge, and to integrate these currently siloed lines of modelling. We will explore soft- or hard-linking existing material flow, energy-economy, energy system optimization, and lifecycle assessment models, building a new integrated modelling framework.

If successful, our project will allow for the integrated assessment of energy use, emissions, material demand, and waste in response to both decarbonization and circular economy policies. It will also provide a novel framework for the exploration of policy interactions within and between circular economy and emissions reduction policy, teasing out the potential additive or subtractive effects of different policy packages. It will inform researchers and policymakers alike, assessing the extent of policy overlap and quantifying emissions reductions from new policy instruments, while building understanding of the material, waste, and broader environmental impacts of potential decarbonization approaches in Canada.

La persévérance scolaire constitue un enjeu majeur dans les établissements d’enseignement secondaire, particulièrement dans les communautés autochtones. Les données montrent que les élèves autochtones présentent des taux de réussite variés selon les contextes, avec des défis distincts comparés à leurs pairs allochtones (Richards, 2014) et ce, tant au Québec qu’en Amérique Latine (Luisetto, 2023). Selon les chercheurs spécialisés, la réussite scolaire des autochtones nécessite une compréhension approfondie des déterminants culturels et sociaux qui influencent leur parcours académique (MELS, 2010). Les facteurs identifiés incluent les influences familiales, personnelles, scolaires et environnementales, mais ces déterminants doivent être adaptés au contexte spécifique des communautés autochtones pour être véritablement efficaces (Campeau, 2017). En particulier, la réalité des élèves autochtones, montre que les approches pédagogiques doivent intégrer des éléments culturels pour améliorer la persévérance scolaire (Sioui, 2013). À la lumière de nos expériences de recherche auprès de communautés quechuas au Pérou et innues en Basse-Côte-Nord, nous croyons que permettre aux membres de ces communautés de croiser leurs perspectives culturelles est une avenue de recherche innovante en éducation et santé autochtone. Nous avons pu constater que les membres de ces communautés, tant au nord qu'au sud, vivaient des enjeux liés à la persévérance scolaire désiraient se réapproprier des savoirs traditionnels ancestraux (STA) en botanique et connaissances écologiques traditionnelles (CET). La pertinence de mettre de l’avant un projet éducatif intergénérationnel sur les STA et CET en contexte scolaire autochtone, auprès d’élèves du secondaire, dans une perspective de sécurisation culturelle, s’est confirmée, tout comme l'intérêt de réaliser un échange interculturel entre ces communautés du nord et du sud. La prémisse du projet est donc d'approfondir les STA en botanique et CET par les arts dans 2 écoles secondaire autochtones, soit au Colegio Anilmayo (Callatiac et Urin Ccoscco, Pérou) et à l’École Pakuashipu (Pakuashipi, Canada), afin de réaliser un échange entre ceux-ci. À l’aide d’une recherche-action en partenariat avec les communautés, le projet aura pour but de développer des stratégies d'éducation basées sur les arts et la pédagogie du lieu afin de faciliter l'exploration des STA en botanique et CET au sein des communautés.

Electromagnetic (EM) inverse problems involve determining the internal properties of a region of interest from its external EM signature. EM inverse problems have various applications including medical imaging, environmental remote sensing, and industrial non-destructive evaluation. Solving these problems requires passing acquired EM measurement data to computers on which advanced mathematical algorithms are implemented to yield a reconstruction of the physical properties of interest. This conventional process is often expensive and time consuming.

The objective of this proposal is to design compact and affordable systems to solve EM inverse problems at the speed of light. These systems will consist of metasurfaces, which are thin artificial materials composed of lattices of small (subwavelength) unit cells. The metasurface system will be designed so that when EM waves emanating from the region of interest interact with it, the resulting EM fields represent the solution to the inverse problem of interest. Thus, the metasurface effectively embeds the mathematical operations required to solve the inverse problem. Consequently, these operations are performed at the speed at which the incident EM waves interact with the metasurface, which is the speed of light.

Inverse problems will be mathematically represented as Fredholm integral equations of the first kind. In particular, the metasurface system will be designed so that its EM wave transformation functionality mimics matrix-vector multiplication, with the matrix being related to the kernel of the integral equation. This will be accomplished by designing metasurfaces that exhibit a property known as strong spatial dispersion, which is challenging to achieve. To this end, cascaded metasurfaces will be designed to collectively create strong spatial dispersion via mutual interactions between multiple metasurfaces. The types of mathematical operators will be adapted, optimized, and approximated such that they can be implemented with practical cascaded metasurface systems.

The novelty of this research is that the proposed metasurface system implicitly measures and transforms the incident information-carrying EM waves thereby directly solving the inverse problem of interest. The result of this research can significantly reduce the cost and increase the speed of imaging and remote sensing systems, which are essential in various sectors such as security, medical, industrial, and environmental applications.

Zoonotic viruses circulate among different animal species and risk spilling over to novel hosts causing severe disease and mortality, and/or potentially the establishment of new host reservoirs. Our understanding of zoonotic virus ecology and ecology has expanded significantly with molecular detection and pathogenomics. Although the largely descriptive body of work generated by these approaches provides valuable insights into the diversity and complexity of viral populations, viral ecology and niche connectivity remain relatively unexplored. Coronaviruses (CoVs) are now among the most studied mammalian viruses, however, CoV-wildlife hosts-habitats and niche dynamics remain relatively uncharacterized.

Coronaviruses are a highly diverse group of viruses infecting numerous mammalian species in broad geographic areas. We have detected CoVs in Eptesicus fuscus and Myotis lucifugus, two species of bats, and in Peromyscus species, all indigenous to Canada. We have established a collaborative surveillance program as well as diagnostic and genomics platforms to enable the detection and preliminary characterization of CoVs in Canadian wildlife, including these small mammals and white-tailed deer. We thus identified the first cases of severe acute respiratory syndrome CoV-2 (SARS-CoV-2) in Canadian deer, including a highly divergent virus that spilled back into at least one human.

Our overarching objectives focus on the following fundamental questions on virus dynamics and ecology among animal hosts: 1. What are the ecological features of viral activity in specific wildlife hosts? 2. What are the principle molecular drivers of viral evolution and their covariates? 3. What is the breadth and distribution of genetic diversity for endemic CoVs?

To address these objectives, we will establish and diversify novel experimental systems with primary cells and organoids from wildlife species to study CoV ecology, evolution, and potential for inter-species transmission among mammals, including humans. This requires a multidisciplinary, mixed methods approach that incorporates applied/diagnostic and molecular virology, computational biology and host biology, ranging from population-level data to finer, molecular spatial scales. This work will address fundamental knowledge gaps, informal viral risk assessments to the health of a range of mammalian species, including humans, and guide future surveillance programs for wildlife and public health.

Chronic overcrowding in Canadian EDs has reached crisis levels, with approximately 1,000 patients occupying Ontario hospital hallways daily. Termed "hallway medicine," this problem results from a disconnect in ED admission and discharge, leading to "boarded" patients treated in hallways. Prolonged ED stays worsen overall hospital experiences and health outcomes.

When "boarded" in EDs, patients face two issues: sitting in structurally inadequate chairs, and receiving care in ill-suited spaces such as hallways. This project proposes reimagining hospital space, treating hallways as clinical areas. Inspired by modular designs like Lego, Paris Métro seats, and ergonomic features of gaming chairs, we introduce the Modular Emergency Department Seating (M-EDS) System. This innovation transforms ED hallways into patient-centric spaces, customizable with backrests, footrests, side headrests, lumbar support, and diagnostic "plug-ins" like IV poles, fall risk monitoring, vitals monitoring, and trays, creating a tailored ED environment.

Achieving these goals involves design/engineering and humanities-based research and the combination of both scientific and humanities-based methodologies. Using ergonomics, occupational health and structural engineering methods, our hardware design proposal will marry the “iterative design process” method which incorporates qualitative feedback from users (patients, emergency personnel, porters, and cleaners) through surveys, contextual inquiry, interviews and usability tests. Uniting these two methodologies is what makes the project unique, and ensures that patients and staff have agency in shaping the M-EDS to meet ED needs. Access to a Design and Technology Lab enables rapid prototyping. Partnering with Brampton Civic Emergency Department, our M-EDS pilot project aims to reinvent hospital hallways into safe, efficient spaces for high-quality care, with three objectives:

1. Improved Patient Outcomes: Enhancing comfort and reducing stress during waiting times. 2. Enhanced Efficiency: Optimizing seating to improve ED flow and care delivery. 3. Higher Patient Satisfaction: Creating a more compassionate care environment.

This interdisciplinary project prioritizes patient comfort, feedback, and innovative seating solutions, potentially transforming healthcare delivery. The M-EDS project addresses "hallway medicine," prompts spatial rethinking in hospitals, and leaves a lasting impact on emergency medicine practices.

Packaging that is recyclable or compostable (e.g., bioplastics) will displace single-use packaging in the coming years, but it doesn’t address the pressing global issue of microplastics. Most recyclable and compostable packaging still ends up in landfills and oceans, and if bioplastics aren’t composted in an industrial facility, they can take as long to degrade as traditional plastics, producing persistent microbioplastics. Packaging that rapidly degrades in the environment is urgently needed. To address this, researchers are adding plastic-eating enzymes into plastics. However, this requires new processes for every type of plastic and impacts the functional properties of the plastics and enzymes.

We propose a transformative paradigm shift: incorporating enzymes into vapor-barrier coatings used in flexible plastic packaging. Novel enzyme-laden nanocoatings will be developed that can be applied to virtually any flexible plastic packaging without compromising the plastic’s properties.

This will be done by integrating computational biophysics and enzymes (biochemistry) with novel industrial-coating expertise (mechanical engineering). This project will (1) develop novel manufacturing methods to integrate plastic-eating enzymes into novel barrier coatings, (2) build atomistic models to study enzymatic hydrolysis in these coatings, and (3) optimize the coatings for plastic biodegradation. Enzymes that break down polyethylene terephthalate (PET) and compostable polylactic acid (PLA) in environmental conditions will be used. Novel spatial-atomic-layer-deposition coaters will be developed to encase the enzymes in a vapor-barrier coating. Additionally, new coatings composed of nanoscale layers of metal oxides and metalcones will be developed to provide the required vapor barrier when dry but become porous when wet to activate the enzymes. Our atomistic models will deepen our understanding of new and complex plastic-enzyme-coating interactions, helping to prevent enzyme damage during the coating process, minimize enzyme quantities, ensure enzyme immobilization on the plastics, and maximize the biodegradation of PET and PLA.

This project will push the boundaries of industrial coating technology by integrating plastic-eating enzymes into a novel packaging-coating process. This ground-breaking technology has the potential to disrupt the $250B flexible packaging market and eliminate microplastics that are damaging ecosystems and human health globally.

We propose designing novel interactions and sensor technology (NSE) using textile weaving (SSHRC) for Health Professions Education (HPE) (CIHR, SSHRC), to improve the health of Canadians (CIHR). In the post-pandemic health crisis, there is an urgent need to train more healthcare providers in the face of limited human resources including expert educators. Existing simulation technologies are costly, lack hands-on tactility and universal accessibility, and have thus had limited success. Our approach focuses on creating affordable, tangible technology to improve clinical skills, such as performing physical examinations, by embedding sensing textiles into standard medical apparel. Smart fabrics, crafted with e-textile metallic threads, will detect hand gestures (like touch, pressure, pinch, squeeze, and stretch) and transmit data to both physical devices and virtual systems, enabling real-time evaluation, measurement, and feedback. We aim to pioneer a new generation of medical apparel featuring these sensors through digital weaving technology, using our state-of-the-art Norwegian computational loom.

To succeed, our project requires a multidisciplinary team, engaging a range of stakeholders, including medical students, engineers, surgeons, and designers. We will employ a Human-Computer Interaction (HCI) co-design methodology, allowing users to participate in the interaction design and digital fabrication processes. Our plan involves prototyping 3 medical garments embedded with wire-free woven sensors: 1) a patient gown, 2) pants, and 3) a mannikin tank top. These prototypes will be evaluated in real-world educational and clinical settings, focusing on usability and learners’ experience. Knowledge and technology developed in this project will be transferable to different disciplines within HPE such as training of nurses and personal support workers in Long Term Care (LTC) homes.

Our project is inherently high-risk due to the novelty of digital weaving e-textiles and the experimental processes involved. However, we are confident that collaboration leverages our expertise enabling us to develop and test the educational impact of this technology with HPE learners using novel interactive garments. Our innovative method promises high impact, offering higher fidelity and lower costs compared to existing HPE technologies. This impact can be scaled for use in lower-resourced settings, ultimately improving patient outcomes and quality of healthcare education in Canada.

Background: Localized scleroderma (LS), is a rare autoimmune fibrotic skin disease. It is presumed to be externally triggered in genetically predisposed individuals. While its pathogenesis is poorly understood, it's known to follow 3 clinical stages: inflammation, fibrosis and atrophy. Despite eventual resolution, permanent sequelae often lead to dysmorphism and functional impairments, especially in children.

Preliminary data: Using RNAseq data from 2 independent patients cohorts, viral and interferon pathways were most upregulated. This led us to hypothesize that idiopathic LS may be triggered by a virus. Using bioinformatic techniques (BLAST), we identified a CRESS-DNA virus in 56/60 adult LS samples with mean read per million (rPM) of 4.063 (range, 0-69.88). This virus was seen in 6/43 healthy controls (HC) with mean rPM of 0.108, (range, 0-3.221) and 23/28 chronic spontaneous urticaria (CSU) patients with mean rPM of 1.93 (range, 0-12.52). However, the complete viral genome was seen in LS samples only.

Hypothesis: We hypothesize that idiopathic LS is triggered by CRESS-DNA infection. The proposed research aims to:

Aim 1: Confirm presence of CRESS-DNA virus in LS (60) vs. HC (43) and CSU (28).

Methods: 1.1 A custom library of CRESS-DNA viruses will be constructed from the literature to determine the exact viral sequences. 1.2 Electron microscopy will be performed on the existing biosamples to identify intracellular viral inclusion bodies in the tissue. 1.3 DNA will be isolated from biosamples and amplified using rolling circle amplification (RCA) to confirm viral presence and species. 1.4 Design a probe/perform in situ hybridization (ISH) to confirm/localize virus in affected tissue. 1.5 Develop/run custom RT-PCR. 1.6 Confirm findings in an independent cohort of LS (fresh cryopreserved).

Aim 2: Elucidate host-pathogen interaction.

Methods: STRING database will be used to explore virus-host protein-protein interactions and build a network of differentially expressed genes which will be analyzed with Cytoscape and MCODE plugins to identify potential hub genes. Prognostic and clinical correlation of Hub genes will be analyzed via cBioPortal. The potential of these hub genes and pathways as druggable targets for therapeutic development will be studied.

Novelty: Aim 1 will confirm viral association with idiopathic LS, informing pathogenesis. Aim 2 will instruct host-viral interaction, identify prognostic biomarkers and druggable targets.

Micro/nanoplastics (MNPs) are pervasive and can have significant health and ecological impacts. MNP detection is crucial for accurate risk assessment, but current methods are inadequate. Techniques such as electron microscopy, confocal laser scanning microscopy, particle tracking analysis, inductively coupled plasma-mass spectrometry, surface-enhanced Raman scattering and fluorescence labelling–each effective in its own way–are either expensive, labour-intensive, require highly specialized personnel, or lack the portability needed for real-world on-site applications.

To address these challenges, we propose to create a low-cost miniaturized scanning electron microscope (SEM) offering user-friendly analysis. This field-deployable SEM would fit in a backpack allowing on-site imaging of MNPs in complex environments.

Our approach is to build a novel device, based on a uniquely simple and flexible SEM platform developed by the NPI, which will feature an electron beam within a permanently sealed vacuum tube, that will pass through an electron-transparent membrane to strike the specimen outside the tube. This eliminates the need for a vacuum pump, making the device compact, portable, and capable of imaging specimens in their natural environment without restrictions on cleanliness, dryness, or size. We will apply digital deconvolution to remove blurs in SEM images. For added detection accuracy and speed, we will use machine learning techniques and algorithms commonly used in computer vision, such as YOLO (You Only Look Once), which has been shown to achieve over 100 frames per second for real-time object detection.

If successful, this project will transform the SEM from a lab-bound instrument into a portable tool that integrates seamlessly into the user's workflow, broadening its applicability across diverse fields such as healthcare, mining, manufacturing, and forensics.

The proposed SEM will not match commercial instruments in imaging performance, but aims to define what a portable low-cost electron microscope can do. The goal is to make electron-beam imaging widely accessible, potentially even in everyday settings like schools (science education) or remote communities (monitoring drinking water quality). Attaining the desired resolution and achieving operation in complex environments are formidable challenges, but the expertise of the interdisciplinary team and the unique SEM platform on which the project is based help mitigate these risks.

Chimeric Antigen Receptor T-cell (CAR-T) therapy represents a groundbreaking approach in cancer treatment. CAR-T cells are produced using a patient’s own T cells that have been isolated and engineered to express a cancer antigen-specific CAR. After infusing back to the patient, CAR redirects the T cells to bind and destroy the patient’s cancer cells with more specificity. Over the past few years, the pipeline for CAR-T cell therapies has been increasing at a rapid pace. To address the increasing demands, biomanufacturing techniques to produce the CAR-T cells have taken the center stage. Some of the current limitations of CAR-T production includes their complex manufacturing process, high production cost, and long preparation time. Furthermore, CAR-T cells are generated using viral vectors which causes a potential safety concern. Although some studies have investigated strategies to generate CAR-T cells without using viruses, sub-optimal CAR expression and low homogeneity of final products still remain a concern.

This project aims to introduce a potentially transformative method for rapid and efficient CAR-T cell therapy. Specifically, we propose to take an interdisciplinary approach and develop a microfluidic high-throughput platform that can rapidly transport plasmid DNA molecules into T cells without using any supplementary viral or non-viral vectors. Under clinical settings, this device will effectively deliver recombinant CAR plasmids into T cells (isolated from patients own blood) and transform them into therapeutic CAR-T cells, such as CD19-targeted CAR-T cells. These cells can then be infused back to the patients for selective and effective killing of CD19-positive tumor cells. Towards this clinical vision, specific objectives of current project are: 1) to develop a high-throughput microfluidic device that can efficiently deliver naked CAR plasmids into T cells, 2) to test the formation and efficient expression of antigen receptors on CAR-T cell surface using the microfluidic device, 3) to validate that the CAR-T cells exhibit potency against different types of tumor cells.

At the completion of this project, we expect to have established a new paradigm for rapid and cost-effective personalized CAR-T cell therapy using non-toxic virus-free conditions. This potentially transformative and sustainable technology can have significant positive impacts in biomanufacturing of other clinical-grade therapeutic cells that need genetic modifications.

Glioma and brain metastases (GBMet) are among the deadliest cancers, urgently needing better treatments. GBMet disrupts the brain’s glymphatic system – a crucial network of perivascular spaces that clears toxic metabolites, imparts immunity, and regulates fluid. Despite these critical functions, only three studies have explored the link between glymphatics and human brain tumours. Our objective is to explore this vital and overlooked system in GBMet, promising a HIGH REWARD return.

We HYPOTHESIZE glymphatic disruption promotes GBMet invasion, especially in eloquent brain regions crucial to movement and language which are more metabolically active and therefore vulnerable to glymphatic disruption.

Our HIGH RISK approach leverages our patented Intelligent Surgical System (ISS), an AI-driven platform that securely analyzes real-time surgical and imaging data to provide immediate feedback to the surgeon. The ISS integrates pre-surgery MRI data highlighting glymphatic disruption and eloquent areas near the GBMet with data from direct brain stimulation during surgery—the gold standard for identifying eloquence. Linked to an advanced language model, similar to ChatGPT, the ISS answers critical questions like, “Which GBMet region has the worst glymphatic disruption and is closest to eloquent areas?” This guides us during surgery towards maximal safe resection while preserving eloquence. Additionally, tissue from these regions can be sampled for future molecular analysis, reinforcing our leadership in treating brain tumour glymphatic disruption and advancing its bench-to-bedside study.

Also, we will leverage routine serial imaging in GBMet patients to construct a novel dataset in the ISS, capturing disease progression, glymphatic disruption, and eloquence. This dataset will train a deep learning model to predict brain regions at increased risk of invasion. For the first time, glymphatic disruption and eloquence will be used to predict brain tumour invasion, redefining research paradigms. By predicting invasion patterns, we shift from the usual REACTIVE treatment of GBMet to PROACTIVE treatment by targeting regions at risk of invasion with more aggressive interventions such as surgery or radiation, transforming treatment paradigms.

Our multidisciplinary team, including experts in neurosurgery, surgical technology, imaging, and AI/deep learning, aims to redefine the role of glymphatics in GBMet invasion and provide new tools to enhance cancer treatment.

The Xingu Indigenous Reserve, a vital part of the Brazilian Amazon, has suffered significant deforestation, reaching 30% of its original area due to land encroachment, wildfires, and agricultural expansion. This has adversely impacted Indigenous communities, jeopardizing their livelihoods and cultural heritage. To address these challenges, an innovative technological platform is proposed to support Indigenous communities to sustainably harvest and trade native seeds. A cornerstone of this platform is a novel seed quality assessment tool that leverages image recognition technology. By capturing a seed image with a smartphone, users can rapidly and accurately determine its viability, quality, and species. Focusing on species with both economic potential and ecological value, such as Brachiaria

humidicola, Andropogon gayanus, Euterpe oleracea (açaí palm), and Bertholletia excelsa (Brazil nut), this project aims to incentivize sustainable harvesting and cultivation practices among Indigenous communities. These species not only provide valuable resources like forage, fruits, and nuts, but also contribute to soil improvement, biodiversity, and forest regeneration. Accurate seed identification, combined with information on species, harvest time, and cultivation, is crucial for rapid regeneration of degraded areas. By promoting the cultivation of these species, the project can create a sustainable economic base for Indigenous communities, reducing their dependence on unsustainable activities and fostering a sense of stewardship for the forest. Image analysis technology offers a rapid and non-destructive evaluation of seed characteristics. Seed images contain information about morphology, color, texture, and other features that can reveal factors such as viability, germination potential, and disease susceptibility. By optimizing image processing algorithms and incorporating machine learning techniques, we can effectively adapt existing image analysis methods to work with smartphone camera data, benefiting researchers, seed banks, and conservation organizations globally. This project represents a significant advancement in seed science, offering a solution to the challenges faced by Indigenous communities. By ensuring traceability and authenticity of seeds, the platform contributes to the creation of a more transparent and reliable seed and seedling market, incentivizing the production of high-quality seedlings and the restoration of degraded areas.

TEAM: This application brings together a bioengineer, vascular and exosome biologist and brain cancer geneticist representing a wide range of research interests and stages in career. We team up to develop a nanobiology-based novel blood test to diagnose and monitor brain cancer progression by remote sensing of the tumour vasculature.

IDEA/RATIONALE: Blood vessels are sensors of the state of tissue they supply. Tumour blood vessels are composed of endothelial cells that lie at the boundary of between tumour mass and circulating blood, to which they have a unique and direct access. All cancers are vascularized, and their associated endothelial cells undergo a profound biological change stimulated by nearby cancer cells and reflective of their properties, such as genetic profiles and factors they release. These changes are central to tumour growth, metastasis, interactions with immune system, drug delivery and other key features.

WE HYPOTHESIZE that endothelial changes associated with specific types of brain cancer are accompanied by a release by these cells of small bubble-like cellular fragments, known as extracellular vesicles/exosomes (EVs), which contain unique physical and molecular fingerprints of endothelia and mirror their cancer-specific responses. Endothelial EVs are likely more abundant in the circulation than material released from any other cellular population trapped within the tumour mass. This exceptional property would render analysis of endothelial EVs (endobiopsy) more informative than any other form of traditional ‘liquid biopsy’, which focus on cancer cell secretions and breakdown products and suffer from major limitations.

APPROACH: To detect cancer-specific changes in endothelial EVs we will deploy a novel nanotechnology-based platform capable of detecting single EVs and assembling them into high resolution disease-specific ‘landscapes’ using Raman spectroscopy and machine learning (MoSERS). We will further develop the molecular profiles of diagnostic endothelial EVs and link them with specific disease states, therapeutic responses, and progression using experimental models and patient samples in adult and pediatric glioblastoma and other brain cancers.

NOVELTY AND SIGNIFICANCE: Monitoring cancer by remote sensing of the tumour vasculature, through EV based ‘endobiopsy’, has never been explored and may fail. However, this approach may also add a new dimension to brain cancer diagnostics and have application beyond cancer.

Every day, 76 Canadian men are diagnosed with prostate cancer (PCa) and 12 succumb to the disease. A biopsy is the only way to definitively diagnose PCa. Men with risk factors will undergo an MRI of the prostate to detect potential areas of clinically significant PCa. When suspicious lesions are observed on MRI, a targeted biopsy is performed, typically under ultrasound (US) guidance. In US biopsy the MRI and US images are fused together to place a sampling biopsy needle into the MRI-observed lesion(s) using real-time US guidance. While a positive biopsy result informs treatment decisions, a negative result has uncertain meaning: Is it due to inadequate sampling, missing the lesion, or benign diseases mimicking cancer on MRI? These uncertainties create doubt in the clinical care pathway and confusion about what to do next. Biopsy under MRI may be more accurate than US, but this has not been confirmed and MRI biopsies are lengthy, costly and require sedation. Contrast-enhanced and micro-US have shown favourable preliminary results but also lack validation, are not available in practice, and have not yet been incorporated into the MRI-biopsy pathway.

To assess whether the biopsy needle is positioned in the MRI target or whether repositioning or additional biopsies are required, our multidisciplinary team will develop methods to detect cancerous tissue at the needle tip and create a tomographic image of its margins. We will use specialized needles that measure the effect of US pressure on the dielectric properties of the tissue to reveal inhomogeneities in its acoustic and electric properties, both of which are strong biomarkers of PCa. Real-time information about the location of the target will augment a biopsy operator’s ability to find suspicious MRI lesions using US and improve biopsy yield, potentially replacing MRI-fusion systems which are prone to image misregistration and poor lesion localization.

Rather than an incremental contribution to PCa biopsy, we propose a fundamentally new imaging method specifically devised for PCa. This paradigm shift in PCa diagnosis will improve confidence in biopsy results, reduce procedure time and the number of required biopsies, while providing new data to guide diagnosis and treatment decisions and significantly improving risk stratification. This technology also can enable tumour bracketing and thus has applications in radiotherapy and focal therapy of PCa and could also be applied breast and thyroid cancers.

This research project aims to evaluate and enhance the resiliency of autonomous systems, including unmanned vessels, drones, and vehicles, against AI-related faults and security attacks these platforms may encounter. By employing chaos engineering, we will simulate and analyze AI-driven disruptions across these platforms to understand their impact on operations. The project’s high-risk nature lies in applying chaos engineering—traditionally used in distributed software systems—to the AI components of physical autonomous systems. This approach not only identifies vulnerabilities but also enables the design of new fault tolerance and security control strategies, significantly enhancing the resilience of these systems.

The novelty of this project lies in its interdisciplinary approach, which combines expertise in AI, cybersecurity, and engineering across multiple autonomous platforms. This comprehensive approach will push the boundaries of current knowledge by introducing a method to systematically test and improve the robustness of AI in real-world, high-stakes environments. The expected outcome is the development of strategies and tools that ensure the reliable operation of autonomous systems, even in the face of unexpected AI failures. This work will contribute to the reliable deployment of autonomous systems across maritime, aviation, and transportation industries, offering a critical advancement in the field.

The project will directly address concerns over the widespread adoption of autonomous technologies by ensuring these systems can operate safely despite AI failures. The outcomes of this research could not only redefine but also significantly elevate industry standards for the safety and resilience of unmanned systems, positioning Canada as a leader in developing secure, autonomous technologies. The knowledge generated through this work will have far-reaching implications, improving current systems and guiding the design of future autonomous technologies. Ultimately, society will benefit by promoting safer and more reliable AI-driven solutions.

Forest restoration is a popular “natural climate solution” that would mitigate climate change by increasing carbon stored in woody biomass. Identifying priority areas for restoration that maximize benefits and minimize costs is essential, yet attempts to do so are limited by major blind spots in remote sensing and other “big” data sources. Data are widely available for socio-economic and ecological features that are relatively easy to map, while other features of landscapes that are central to just and effective outcomes, but hard to map, remain “hidden” in global analyses. We propose a high-risk, high-reward project to address the two principle gaps in existing assessments of restoration priority: 1) belowground carbon (BGC), which is measured infrequently and with great difficulty and 2) cryptic land uses that are important to the livelihoods of resource-poor and marginalized people yet left out of evaluations of opportunity costs of restoration.

Our specific objectives are: 1) Develop methods for rapid field assessment and prediction of BGC from aboveground features to create maps of BGC at high spatial resolution. 2) Assess and generate landscape-scale maps of cryptic land uses by combining approaches from development and rural livelihoods and remote sensing. 3) Update restoration priority maps to assess the impacts of incorporating hidden carbon and land uses on restoration planning.

This research incorporates theory and methods from geomatics and remote sensing, social-ecological systems, ecosystem ecology, development and rural livelihoods. We will work across gradients of tree cover, restoration and management intensity in a global priority landscape for restoration and climate mitigation. Our interdisciplinary team includes expertise on geomatics and remote sensing, ecology of grassy biomes, natural climate solutions, restoration social science, and participatory research, and includes non-academic collaborators from the study area. Our new methods will facilitate mapping and monitoring BGC at multiple scales (plot to landscape) and will demonstrate how remote sensing can facilitate understanding cryptic land uses. These cutting-edge tools can be extended to other regions, enhancing our ability to measure and monitor important but elusive components of multifunctional landscapes. Our restoration priority maps will demonstrate how incorporating hidden features can improve both environmental and social outcomes of natural climate solutions.

Infant formulas (IF) are designed to provide as close-to-optimal nutrition as possible for infants. They typically contain whey and casein proteins from cow’s milk, vegetable oils, lactose, and other micronutrients. Despite substantial improvement in the nutritional quality of IF, there remains a measurable gap in performance compared to human breast milk when assessing early childhood development. Correspondingly, human milk is still universally the preferred feeding modality for developing infants.

There are several factors that contribute to the gap between IF and human milk including a fixed whey-to-casein ratio, allergenicity associated with bovine or plant-based proteins and deficiencies in bioactive milk components such as lactoferrin and milk fat globule membrane. Recent severe supply chain shortages, melamine contamination, and issues of antimicrobial resistance and greenhouse gas emissions associated with large scale dairy farming motivates the need and opportunity to adopt alternative approaches to producing IF and key ingredients therein.

The objective of this project is high-risk and aims to fundamentally alter the composition and approach to producing IF by humanizing the protein content using precision fermentation. To accomplish this, we will engineer fungal cell factories to convert low-cost raw materials into human milk proteins by combining fungal genetics and strain engineering, and protein biochemistry. This approach will produce proteins that mimic the structure and function of proteins found in human milk. Outcomes from this project are high reward, with the potential to reduce allergenicity, enhance bioactivity and functionality, decentralize production and improve the sustainability of IF production.

The project will rely on an interdisciplinary approach to evaluate the quality, safety, and consumer acceptance of human milk proteins produced through precision fermentation. We will employ nutritional immunology and mycotoxicology to validate that the milk proteins display reduced allergenicity, while confirming they are free from compounds that could raise food safety concerns. Consumer acceptance of IF produced using fungal-derived ingredients will help guide bioprocess development to proactively address consumer concerns. Focus will be placed on acceptance in Canadian Indigenous communities, which have the lowest rates of breast feeding in Canada and consequently are most susceptible to gaps between IF and human milk.

Alzheimer’s disease (AD) is the most common type of dementia characterized by the deposit of misfolded proteins (neurotoxic Aβ peptides forming dense fibrillary plaques) causing neuronal cell death leading to dementia and eventually death. More than 700,000 Canadians are currently living with AD and worldwide more than 40 million people are afflicted by this neurodegenerative disease. Finding a cure for AD constitutes one of the greatest challenges to modern medicine and billions of research money is currently invested in finding a cure. A major limiting factor in developing effective treatment strategy for AD is the blood-brain barrier (BBB) preventing entry of drugs/agents into the brain. Over the last decade, nanoparticles, which are engineered materials less than 100 nm in diameter with unique physicochemical properties, have been explored extensively as an area of novel therapeutic modalities to overcome the BBB. Acidic poly (D, L-lactide-coglycolide) (PLGA) nanoparticles which constitute a family of FDA-approved biodegradable polymers have long been studied as delivery vehicles for drugs, proteins, and other macromolecules. Interestingly, our recent data indicate that native PLGA nanoparticles can also ameliorate not only Aβ aggregation/toxicity but also ADrelated pathology in cellular and 5xFAD mouse models of AD. Considering the evidence that native PLGA nanoparticles can attenuate Aβ aggregation under in vitro conditions, we will develop a theranostic platform for the diagnosis and treatment of AD. The diagnostic part is based on labeling the PLGA nanoparticles with eligible positron emitting radionuclides (e.g. fluorine-18 or longer lived nuclides such as 64Cu) for PET imaging, whereas the therapeutic efficacy is rooted in the PLGA nanoparticle's ability to successfully remove the plaque deposits from the living brain. Fundamental questions about PLGA nanoparticle pharmacokinetic profile and its curative efficiency can only be answered if one can follow the distribution of radiolabeled PGLA nanoparticles in real-time. PET imaging will provide the necessary tool to answer these questions alongside the verification that PGLA nanoparticles can remove plaques deposits.

Background: Neurological disorders resulting from the deterioration of healthy brain functions affect nearly one billion people globally leading to a poor quality of life and high medical burden to society. Neuroscientists use stand-alone tools such as optogenetics or electrophysiology to study neurons, the basic building blocks of the brain. However, existing neural implants fail to provide long-term, high-resolution data due to rigidity, and issues related to biocompatibility and biostability.

Current limitations include:

- Planar bioelecctrodes based on rigid materials have mismatch with the soft neuronal tissue leading to inflammation, delamination and corrosion.

- Multifunctional electrode designs are tiled preventing precise localizing of neurons.

- Stand-alone electrophysiology electrodes make it impossible to identify the location of study post implantation.

- The high cost (~$2000 CAD/probe), complex usability, and low reusability limit pre-clinical applications.

Goal: To develop a coaxially engineered soft micro-fiber electrodes that can optically stimulate, record electronic action potentials, neurotransmitters, and deliver drugs to the same neuronal target coupled with facile tailorability.

Proposed Methodology: Our interdisciplinary project will actively involve experts from various disciplines (flexible electronics, textile engineering, materials chemistry, biosensing) along with end-users ‘neuroscientists’ in the device development. We will engineer a core-shell fiber structure with a custom roll-to-roll manufacturing setup. A doped elastomer will be spun as the fiber core for optical stimulation. Then it will be coated with a porous and fibrous composite of conductive polymers and ionic liquid for electrophysiological recording, stimulation and drug delivery.

Schedule:

1. Neuroscience team consultation and fabrication methodology setup (months 1- 4)

2. Refinement, Optimization and Validation with Ex-Vivo studies (months 4-14)

3. In-Vivo animal study on Huntington’s mice (months 15-24)

Expected results and significance:

1. A soft electro-photonic fibre that is mechanically and chemically biocompatible and provide biostability.

2. Reduction in the number of animals used for neurological disease studies due to longitudinal reliability and specifity.

3. Adaptable soft electronic fibers enable customizable electrode length and positioning for different disease studies.

4. Enabling technology for nerve regeneration

Numerous Canadians annually seek assisted reproductive technologies (ART). The most common ARTs, in vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI), involve essential yet invasive egg retrieval and embryo transfer procedures, which are conducted while patients are under sedation. The invasive IVF/ICSI procedures pose risks such as damage to nearby organs, reproductive tract infections, and anesthesia-related complications. The use of rigid surgical tools (needles, catheters, and ultrasound probes) also impacts the quality of eggs and embryos, and subsequently, fertilization success rate.

We aim to revolutionize Assisted Reproductive Technologies by Soft Microrobots (SMART) and enable non-invasive retrieval, manipulation, and transport of eggs and embryos. SMART is high-risk research seeking solutions via an interdisciplinary approach from biomedical robotics, reproductive medicine, macromolecular and materials chemistry, microtechnology, and biomaterials. Microrobots have made remarkable progress toward non-invasive medical procedures such as drug delivery, microsurgery, and biopsy. Nonetheless, they have never been used in the real world for ARTs due to their limited biocompatibility, programmability, scalability, and challenges in vivo powering, navigating, controlling, and monitoring.

SMART will rely on novel biocompatible soft materials responsive to changes in their environment, including exposure to chemicals, biomolecules, and magnetic fields, in a programmable fashion. SMART will be equipped with active capture-release, remote guidance, and transport mechanisms while being monitored by medical imaging modalities during navigation and function. SMART will be tested in vitro, ex vivo, and in vivo in animal models. We believe SMART will provide much less invasiveness and complications, greater precision, control, and success rate, and reduced time and cost of treatment than existing procedures.

Accessing IVF/ICSI entails lengthy waiting periods, with each cycle costing around $15,000, typically not covered by insurance. This places a considerable burden on many Canadians. SMART bears significant benefits for Canadians as it will offer accessible, affordable, non-invasive, more successful ARTs, and better mental health to couples struggling with infertility. SMART will put Canada at the forefront of ART research and offer potential translational technologies for other fields such as endocrinology, urology, and surgery.

Pancreatic cancer is one of the deadliest cancers, with high mortality and limited treatment options. Its aggressive nature, late diagnosis, and resistance to therapies contribute to a poor prognosis. The genetic complexity and heterogeneity of pancreatic tumors complicate the identification of therapeutic targets, requiring innovative strategies. Gender-specific differences in incidence, symptoms, and treatment response further highlight the need for tailored approaches. Generative Artificial Intelligence (AI) is increasingly recognized for its potential to synthesize new data and insights. Our project aims to leverage this technology to address the genomic complexity of pancreatic cancer by:

1. Generating Novel Mutational Signatures: Identifying and synthesizing new mutational signatures to uncover previously unrecognized therapeutic targets. 2. Predicting Optimal Treatment Strategies: Using these novel signatures to identify key mutations driving cancer progression and design personalized treatment strategies tailored to each patient’s genetic profile.

Research Approach: We will train our Generative AI framework on genomic data from The Cancer Genome Atlas, validated with external datasets. This includes gene expression profiles, somatic mutations, copy number variations, DNA methylation, and miRNA expression levels. The model will analyze existing mutational patterns, simulate interactions among mutations, and generate new signatures. We will also employ Conditional Generative Adversarial Networks to generate gender-specific mutations, ensuring equitable representation in mutational signature identification. Furthermore, deep learning models, such as recurrent neural networks, will be used to predict optimal treatment strategies by integrating these novel signatures.

Novelty and Significance: This research goes beyond traditional methods by generating new hypotheses about mutational signatures. Creating these profiles is challenging due to complex biological mechanisms and the significant validation required for clinical relevance, but once successful, the potential rewards are equally high. This project could redefine pancreatic cancer treatment by shifting from descriptive to generative models that predict and create new therapeutic avenues. Ultimately, we will develop a digital platform with our trained AI algorithm as the backend, ensuring broad research use, scalable across various cancers, making it a versatile tool in oncology.

In Canada, over half of adults aged 40 and older experience hearing loss (HL), a major public health concern due to its negative impacts on communication, social participation, quality of life and collateral effects on health (depression, risk of falls, cognitive decline). Despite the many benefits provided by hearing aids, commonly used to compensate HL, users still face performance issues with speech perception in noise (SPIN) and music appreciation. Addressing these limits requires innovative approaches. Even if the sense of hearing shares many similarities with touch, few technologies exist to transform sound into vibrotactile (VT) input despite the fact that some studies revealed that tactile cues can improve SPIN and melody recognition. To further explore such under-used potential offered by the tactile system, we developed a first VT device to feel sounds with hands. Experimental results from our studies confirmed that deaf people can perceive frequencies and emotions in music through touch. More surprisingly, our recent investigations suggest that the ventral stream associated with auditory perception in the brain can also be activated by touch. Nevertheless, the current prototype of VT glove is not suitable for daily usage and end-users now need to be involved to design a functional system that goes beyond a delicate prototype conceived for scientific research. Our project aims to involve end-users -starting with hard-of-hearing individuals and their relatives- in the design process while advancing the prototype’s technological development. A first step will be to assess the needs, social acceptability and satisfaction related to the current device combining questionnaires and Delphi method strategy. Subsequently, based on these findings, improvements will be made to the technology and re-evaluated by users to ensure the identified limitations are being addressed. Technological development is inherently challenging, and this project is considered high-risk due to issues like portability, energy consumption, and connectivity. The VT glove has the potential to promote social interactions by facilitating communication in noisy environments, improve music enjoyment and provide a new tool to clinicians to face the numerous collateral effects of HL. This project, with its high impact potential, combines the health and technology sectors and could also provide access to an important cultural activity: music.

While breast cancer (BC) remains the most common cancer globally and is among the leading cancer death causes worldwide, the specificity and effectiveness of the common diagnostic procedure (X-ray mammography sometimes coupled with breast MRI) is severely limited and results in three-quarters falsely prescribed biopsies. The situation is the most urgent for women with a known high and intermediate lifetime risk for breast cancer who are also at high risk of developing two of the most aggressive types of breast cancer: HER2+ and triple-negative breast cancer (TNBC). Despite immunohistochemistry (IH) being known for decades and the concept of molecular imaging MRI biosensors being proposed around twenty years ago, there has been no success in performing non-invasive IH in a living organism.

Our interdisciplinary team is challenging this healthcare burden by implementing hyperpolarized (HP) xenon-129 (129Xe) MRI - an imaging modality that had never been used prior in oncology. We aim to develop first-in-kind molecular MRI imaging biosensors capable not only of breast cancer detection with high accuracy but also of performing IH in living bodies and distinguishing TNBC from HER2+ BC. This high-risk high-reward project will be conducted through joint efforts of chemists, molecular biologists, electrical engineers, radiophysicists, and computer scientists and will be designed based on the advice of clinical oncologists and patient advocates. Briefly, the resorcinarene trimer (R3) methanesulfonate (R3-Noria-MeSO3H) macrocycle will be linked to nanobodies developed to have extreme affinity for either HER2+ or TNBC. A dedicated imaging system composed of an MRI loop coil coupled with novel Lenz Resonators will be built for superior signal sensitivity and implemented for molecular biosensor detection via HP 129Xe MRI. The achieved images will be reconstructed using a dedicated AI approach and converted into an IH map which will identify the presence/absence of the BC and its genetic subtype. The sensitivity and specificity of the developed biosensor as well as image resolution will be evaluated and compared with the identified requirements by clinical oncologists and patient advocates.

This innovative approach will not just be the pioneering take on 129Xe molecular imaging biosensors, but also a fist-in-kind approach to combine molecular imaging and IH to detect and identify BC subtypes non-invasively at an early stage during a single session of MRI screening.

The objective of this project is to develop an artificial intelligence-driven solution to improve the outcomes of infertility treatments. Infertility affects approximately 1 in 6 couples in Canada [1], presenting significant emotional and financial challenges. Infertility treatments can cost over $20,000 per cycle, imposing financial burdens on both patients and the public healthcare system. Despite advancements in assisted reproductive medicine, such as improved embryo culture techniques, live birth success rates have stagnated at around 29.2% per oocyte retrieval [2].

Preimplantation genetic testing for aneuploidy (PGT-A) is increasingly utilized worldwide to select euploid embryos for transfer. However, PGT is an invasive diagnostic test that requires the removal of trophectoderm cells from embryos, and the interpretation and reporting of sequencing results currently rely mainly on subjective human analysis. We propose developing machine learning algorithms to utilize embryo morphology and other characteristics to predict embryos’ competence and euploidy status. Additionally, training AI on large datasets from PGT-A, correlated with clinical outcomes, could help resolve ambiguous results and may outperform human capabilities in selecting embryos with optimal chromosomal integrity.

The goal of this interdisciplinary research is to shift from a "one-size-fits-all" approach to infertility treatment to personalized protocols based on individual patient data. By leveraging machine learning advances, the project aims to improve fertility treatment outcomes, increase live birth success rates, and reduce the financial and emotional burdens on couples facing infertility. This personalized approach seeks to enhance the precision and effectiveness of infertility treatments, ultimately contributing to better healthcare experiences and results for patients.

[1] https://cfas.ca/Canadian_Fertility_Awareness [2] Cumulative Clinical Pregnancy and Live Birth Rates from Autologous In Vitro Fertilization Treatment Cycles in Canada (bornontario.ca)

We are experiencing the 6th mass extinction, driven by humans. Activities such as deforestation, direct exploitation, pollution, and invasive species introduction all contribute to declines in biodiversity. These actions mean humans and wildlife habitats are progressively overlapping leading to increased disease transmission. We will apply an interdisciplinary, One Health approach that incorporates microbiology, spatial ecology, participatory development, public health, biological and social/cultural anthropology, and international development to understand the potential reservoirs and transmission routes of zoonotic diseases in NW Madagascar.

This ambitious study will take place in Ankarafantsika National Park, NW Madagascar, where people, and their livestock live in complex landscapes shared with eight species of lemurs. Lemurs are the most endangered group of mammals in the world and people in Madagascar are facing a humanitarian crisis. People, domestic animals, and lemurs are all potential carriers and susceptible to infection by zoonotic pathogens, which can result in diarrheal diseases causing illness and death.

We will create a socio-cultural-ecological map using participant mapping, interviews, focus groups, and Geographic Information Systems (GIS). This map will show how people, lemurs, and domestic animals use and are distributed within the landscape. To determine the occurrence of zoonotic pathogens in people, lemurs and domestic animals we will obtain stool/fecal samples from children under the age of 5, collect samples from latrines and open defecation sites, and via fecal samples respectively. Water samples will be collected and concentrated using dead-end ultrafiltration or membrane filtration using field pumps. Water and fecal samples will be stored on ice and then pre-processed and preserved before being transported to the Water, Health and Applied Microbiology (WHAM) lab for molecular analysis.

Our novel methodology will combine qualitative data regarding human and animal behaviours and will be added to our socio-cultural-ecological map to identify the most likely transmission routes and to propose interventions for reducing transmission.

This study is the first to leverage this type of unique interdisciplinary approach assessing a One Health issue and will provide much-needed data to help conserve species at risk of extinction and improve people’s health and livelihoods.

Generative Artificial Intelligence (GAI) systems play vital roles in automated decision-making and the generation of content in the form of source code or contents such as images. Ensuring that automatically generated content is ethical has become a global concern with authorities like UNESCO and Global Privacy Assembly developing guidelines to mitigate potential risks of misuse of GAI systems. Despite its importance, there is a lack of understanding due to its interdisciplinary nature, and the focus of prior techniques on fairness testing that identifies software discriminations, neglecting other vital ethical aspects (e.g., software harms).

We have assembled a team of researchers with interdisciplinary expertise in the areas of software testing, applied ethics, and quality measurement. We take inspiration from the software testing literature where we design mutation operators to transform inputs to identify unethical content. Our framework assumes that an unethical input can be obtained by starting with an ethical input and applying certain transformations. Our assumption is inspired by Mencius’s and Aristotle’s theory of innate goodness: human behaviors are benign initially and then unethical behaviors are derived from them. Our interdisciplinary approach aims to improve the quality of generated content via four objectives: (1) To qualitatively analyze unethical behavior in generated content and construct a dataset of unethical keywords with behavior and underlying principles; (2) To design and implement an automated testing framework that identifies unethical behavior in code generation systems; and (3) To design and implement an automated testing framework that identifies unethical behavior in GAI systems for non-code content; and (4) To evaluate results of prior objectives in Canadian contexts (e.g., education, healthcare) with real applications.

This project is high risk because it is among the first to combine software testing, applied ethics, and quality measurement to detect unethical content by GAI systems. Despite the teams’ expertise, the task is complex as GAI systems can be challenging to test due to the instability of generated outputs and rapid evolution, thus there is uncertainty in our potential to reliably detect unethical content. Despite the risk, this project has high reward potential, from accelerating responsible AI education, to identifying societal risks of unethical content, and improving quality of generated content.

The research proposal aims to address the pressing issue of global warming by developing innovative microbe-assisted technologies that mitigate nitrous oxide (N2O) emissions, a potent greenhouse gas. N2O is a significant contributor to global warming due to its high heat-trapping potential and long atmospheric lifespan. The project proposes a novel approach to genetically rewire soil microbes to limit N2O production utilizing engineered bacteriophages, viruses that selectively infect bacteria, as vectors for precise gene editing.

The central hypothesis of the research is that by genetically modifying soil microbes, particularly those involved in the nitrogen cycle, the production of N2O can be significantly reduced, thereby mitigating its contribution to climate change. An innovative and high-risk aspect of this research is the use of bacteriophages to deliver genetic cues/instructions to soil microbes in situ. Bacteriophages are highly specific to their bacterial hosts, making them ideal vectors for targeted gene delivery. The proposal outlines the production of bacteriophages designed to genetically alter the soil microbes in its intact soil environment. In addition to the novel microbial genetic engineering approach, the proposal will integrate machine learning (ML) techniques to identify target gene modifications that will curb N2O production. ML algorithms will be used to analyze vast datasets related to microbial genomes, enzyme activities, and environmental conditions. This would facilitate the identification of optimal genetic targets and the prediction of the most effective bacteriophage delivery modules. Moreover, ML could be used to model and simulate the outcomes of microbial modifications in various soil environments, helping to refine and optimize the engineering strategies before they are implemented in the field.

The project is highly interdisciplinary, bringing together expertise in microbiology, bioengineering, environmental science, and computational biology. In summary, this proposal presents a groundbreaking approach to climate change mitigation through the genetic rewiring of soil microbes with the assistance of functionally repurposed viruses. By incorporating ML strategies, the project aims to enhance the precision and effectiveness of these interventions, ultimately contributing to a significant reduction in greenhouse gas emissions and offering a novel solution to one of the most critical environmental challenges of our time.

Advanced microchips are in growing demand in many industries, such as artificial intelligence, medical devices, electronic weaponry, consumer electronics, and automotive. However, the manufacturing process of these chips involves complex steps that produce toxic wastes harmful to human health and the environment, including heavy metals and contaminated metal oxide-based nanoparticles (NPs). This proposal is a proactive response to the urgent need for an efficient and sustainable solution in chip manufacturing by ecodesigning new chip processing methods to minimize their harmful effects while promoting a circular economy. While we acknowledge the high risk involved, we are confident that our approach can overcome the technological and economic challenges that have led to previous failures in the chip industry.

Our interdisciplinary team, comprising materials engineers, chemists, and sustainability scientists, is set to revolutionize chip manufacturing. We plan to replace the commonly used Chemical Mechanical Process metal oxide-based abrasive NPs with our novel organic waste-derived biopolymer NPs (lignin/chitosan/elemental sulfur). Unlike their conventional counterparts, these biopolymer abrasive NPs are recyclable and can recover heavy metal contaminants, making them a potentially more sustainable and environmentally friendly choice. Our chemist will synthesize new biopolymer abrasive NPs and their recycling methods. At the same time, the materials engineer will evaluate their nano-tribological and wear performance on chips using atomic force microscopy and nanoindentation techniques. The sustainability scientist will assess their potential environmental and social impacts by considering various life cycle assessment (LCA) indicators such as climate change, water use, mineral depletion, and toxicity. New impact indicators will be developed to account for the potential impacts of NPs released to the environment. Through iterative learning cycles and collaboration, our team will produce functional biopolymer NPs that do not negatively impact the environment and society. The successful outcome of this project is significant and will yield high rewards. The new NPs will reduce chip production costs and the industry's carbon and toxic chemical footprint and have a positive ripple effect in other industrial sectors that depend on chips. It could also lead to greater electronic device access in education through reduced chip costs.

The study of natural animal behaviour has led to pioneering discoveries in neuroscience, such as the Nobel Prize-winning discovery of how neurons encode an animal's spatial location. However, human neuroscience research is often constrained by methods that measure neural activity while participants are stationary and tethered. Although efforts have been made to study more natural behaviors through movies and video games, studying human brains in real-world "in the wild" settings have remained elusive. Our research project, however, introduces a novel approach that overcomes these limitations.

Our team of neuroscientists, psychologists and engineers, will address this challenge by leveraging recent technological advancements. We have developed a Health Canada-approved mobile system that continuously records neuron activity, sensory inputs, and physiological states. Additionally, we have introduced a mathematical technique that identifies critical transitions in brain dynamics directly from these recordings. These advancements will enable us to closely monitor and analyze human brain function in real-life, pragmatic scenarios.

Over the next two years, our research will focus on integrating these techniques to transform our understanding of how people perceive and remember their experiences. Studies suggest that distinct transitions in brain dynamics segment ongoing experiences into meaningful, memorable events. However, insights are limited since they predominantly come from passive movie watching—the closest approximation to real-life experiences available before our developments.

We will assess our new approach using two aims: (i) Validation: Using recordings performed during movie watching, we will compare the performance of our state-segmentation technique to commonly used algorithms; (ii) Extension: Using recordings of participants' natural experiences while in our epilepsy monitoring unit, we will determine how unexpected transitions in brain dynamics relate to their sensory experiences, actions, and interactions.

This project represents a paradigm shift in neuroscience, providing the tools to study the human brain with the same precision and flexibility used in animal studies. This shift will bridge the gap between animal models and human research, revolutionizing our understanding of brain function. More importantly, it has the potential to transform the treatment of various brain conditions, offering new insights and strategies for patient care.

This interdisciplinary proposal brings together researchers from medicine, computer science, the social sciences, and humanities, to develop and evaluate novel foundation artificial intelligence (FAI) models on electronic health record (EHR) data for the purpose of clinical risk prediction. FAIs are multipurpose AI models that train through a self-supervised approach, learning associations in data and new representations that are suitable starting points for many different tasks. FAIs have gained widespread popularity as Large Language Foundation Models (LLMs such as ChatGPT). While LLMs are trained on large volumes of available text data, such as on the internet, EHR data is highly sensitive and often siloed away in protected repositories. Thus, efforts toward developing FAI models for EHR data are nascent, and many important questions remain about their performance, reliability, generalizability, biases, and what types of clinical tasks they may be useful for.

This proposal will leverage GEMINI, Canada’s largest multi-institutional hospital EHR dataset for research, containing 30 hospitals and 2.5 million patient records from all regions of Ontario. The project will have 3 aims: 1. Evaluate existing EHR FAI models (which have largely been developed in single U.S. centres) and assess performance on patient risk prediction (e.g. predicting hospital-mortality or 30-day readmission) and clinical note generation (e.g., patient discharge summaries). 2. Refine existing FAI models, or, if necessary, train entirely new models, using GEMINI’s large multi-institutional sample. 3. Assess for bias in the existing and refined models across a wide range of patient sociodemographic characteristics, including age, sex, language, homelessness, disability, and neighborhood-level measures of racialization, income, and education.

FAI models could be highly impactful in medicine, using data generated through routine healthcare encounters to perform many tasks, such as predicting patients’ risk of future events, making treatment recommendations, and generating clinical notes to reduce the burden of clinical documentation. FAI models may also be unreliable, produce unsafe recommendations, exacerbate inequities in care, or increase the workload for clinicians. FAI models are the very definition of high-risk, high-reward, and this proposal brings together world-class interdisciplinary teams and unique data assets to explore the frontiers of this emerging field of research.

Breastfeeding is a complex biological process involving mother, infant, and close interactions between the two in the context of their psychological, social, biological, and cultural environment(s). Each one influences the milk given their own health state. For example, an ill infant signals the breast, a biological pump, through its saliva, and the breast then adapts the nutrient composition of the milk to promote baby’s health and wellbeing (Al-Shehri et al. 2015). However, exclusive breastfeeding is not always possible. Of 13.4 million babies born preterm annually, 80-90% are born 32 to 36 weeks’ gestational age (Goldenberg et al. 2008). These babies need their mother’s milk for the first 6 months to ensure survival and reduce illness (Duijts et al. 2009) but may have weak, ineffective, uncoordinated suck/swallow patterns, ineffective latch, and fatigue or sleepiness while breastfeeding (Currie et al. 2019). When a mother’s breastfeeding experience is poor, it impacts maternal mortality and morbidity, specifically mental health (Bartick et al. 2017). To address these challenges, we will design a new wearable performance enhancement technology to train moderate to late preterm infants to breastfeed while the mother simultaneously pumps. Our innovation challenges the conventional principle of breast pump use in which the breast pump replaces the infant and interferes with mother-infant bonding. We will 1) develop a novel, bio-inspired prototype device which personalizes milk production by using real-time infant feeding performance feedback (sensing) for output regulation (baby-in-the-loop) and 2) gain a holistic understanding of different physical effects (pressure induced extraction combined with mechanical expression) and situational effects (comfort, usability, safety), and relationship of these factors to mother-infant wellness. We will draw on our team’s interdisciplinary knowledge and practical experience in health science fields and engineering methods and technologies with focus on rehabilitation and human performance augmentation. We will be mindful of cost, and work in ways that do not harm the environment but rather benefit it. Our close collaboration will ensure that the ecology of breastfeeding is at the forefront of designing the new wearable technology to make feeding contingent on mother and infant engagement, enabling them to access the well-documented physical and mental health benefits of exclusive breastfeeding.

Psychiatric disorders, such as depression and anxiety are becoming increasingly prominent in modern society, especially after the global pandemic, leading to high individual suffering and significant socioeconomic burdens worldwide. Recent studies have demonstrated the therapeutic benefits of hallucinogens (e.g., psilocybin and MDMA) for a wide range of treatment-resistant mental health issues potentially by modulating the default mode brain network. However, the effects of these drugs on individuals are hard to predict and control due to the complex interplay of multiple factors while the adverse responses are difficult to interrupt during therapy. Also, their long-term risks require further investigation. Virtual reality (VR) technology can create an immersive audiovisual experience to represent a wide range of phenomena. In an effort to mitigate the drawbacks associated with hallucinogenic therapies, recent preliminary studies have explored simulating psychedelic experiences through VR, demonstrating promising therapeutic outcomes. Yet, the strength of “digital psychedelic therapy” is still weak due to passive content feeding and may be greatly enhanced by integrating neurofeedback to modulate the brain circuitry more effectively.

The project will develop a novel immersive VR simulator, termed PsyTrip, to create interactive visual and auditory psychedelic experiences. We will leverage electroencephalogram (EEG) and generative artificial intelligence (AI) to establish a neurofeedback loop that dynamically co-adapts audiovisual VR content and brain activities in real-time. Specifically, we will build reinforcement learning algorithms that drive the VR experience to optimize the modulation of the default mode network and global brain connectivity. In addition, we will explore various artistic paradigms for creating complementary visual and auditory elements using generative AI techniques informed by surveys of psychedelic experience descriptions. Finally, we will assess the efficacy of the proposed system using subjective questionnaires. This highly interdisciplinary project will integrate and innovate knowledge and practice from multiple domains, including computer science, neuroscience, and creative arts. With the user-in-the-loop design, our PsyTrip system is expected to offer safe, personalized, and controllable “digital psychedelic therapy”, potentially enhancing treatment outcomes for psychiatric disorders, particularly depression and anxiety.

Invasive freshwater dreissenid mussels, are prevalent across Ontario, Quebec, and Manitoba, and are continuing to expand their range. Controlling these mussels in their early invasion stage is crucial to preventing their spread and minimizing environmental impact.

Objectives: Within the two-year duration of the project, we aim to i) develop an underwater robot that generates oxygen (O2) nanobubbles (NBs); ii) explore whether O2 NBs, combined with biopesticides, can eradicate invasive mussels; iii) construct artificial outdoor ponds to evaluate the ecological impact of this mussel control technology.

Research approach: To address the ongoing spread of invasive mussels and the limitations of current control methods, we will create an octopus-shaped underwater robot capable of generating O2 NBs through hydrodynamic cavitation. The robot’s tentacle-like design will maximize coverage, enhancing the delivery of O2 NBs. We will test the ability of O2 NBs to eradicate zebra and quagga mussels at different life stages - glochidia, juvenile, and adult, by combining O2 NBs with biopesticides. Additionally, we will assess the technology’s ecological impact using artificial outdoor ponds.

Novelty and expected significance: This project introduces an innovative approach to invasive mussel control by integrating advanced robotics, O2 NBs, and biopesticides. The octopus-like robot’s biomimetic design, combined with the novel use of O2 NBs to enhance biopesticide effectiveness, represents a pioneering method in this field. The project faces three risks: i) Engineering challenges in constructing and operating the robot; ii) Uncertainty regarding the effectiveness of O2 NBs combined with biopesticides in eradicating mussels at all life stages; iii) Potentially unpredictable ecological impacts. If successful, this research could revolutionize invasive species management in freshwater ecosystems, particularly in Canada, halting the spread of invasive mussels, protecting biodiversity, and reducing associated economic burdens. The technology could also be applied to other invasive species and aquatic environments, offering a scalable, sustainable solution to a global environmental challenge with significant, environmental, scientific, technological, and economic benefits.

Alzheimer’s disease (AD) is an age-related progressive neurodegenerative disorder that represents the most common type of dementia. The Alzheimer Society of Canada’s Landmark Study estimates that by 2030, the number of people living with dementia is expected to rise to nearly 1 million people, with women being disproportionately affected. Despite substantial progress in understanding AD, by the time clinical symptoms develop, therapeutic intervention is often untimely. In the absence of diagnostic tools for early detection of AD, treatment options are ineffective as neuronal damage at that stage becomes irreparable.

Retinal imaging has been explored for detecting AD biomarkers, but its effectiveness is hindered by pupil size, cataracts, and the impact of systemic diseases such as hypertension and diabetes. In contrast, the cornea is easily accessible for imaging and not affected by these conditions. It also has a rich neural network that may provide insight into neural damage in AD. Our goal is to develop a novel corneal imaging technology as a non-invasive, cost-effective, and precise screening test to detect those at risk of AD.

Hypothesis: A novel 3D optical coherence microscope will be developed to capture and catalog in vivo corneal images of patients with AD, establishing a “corneal signature”.

Objectives: 1-Develop a fast, non-contact 3D microscopy imaging system that integrates line-field illumination optics, hyperspectral, and temporal dynamic imaging technology to identify “corneal signatures” in mouse models 2-Identify and characterize corneal biomarkers of disease in mouse models of AD at different disease stages 3-Explore and characterize these novel biomarkers in AD patients compared to healthy age-matched controls

Team: The project is led by a Canadian corneal specialist renown for diagnostic corneal imaging techniques. The co-PI is an expert in biomedical engineering and ocular imaging system development. The team also includes world-renown experts in AD, neurosciences, and molecular and cellular basis of age-related neurodegenerative disease.

Impact: This novel imaging system will serve as the “window to the brain”, enabling earlier diagnosis and intervention for patients with dementia before they develop debilitating cognitive decline and memory loss. This non-invasive technology will overcome the limitations of current invasive brain tests, and offer a portable and accessible diagnostic tool for remote and underserved populations.

Participatory decision-making in community planning is vital but often difficult because of the conflicting needs of community members and their busy lives. This two-year project aims to overcome these challenges by creating and testing an artificial intelligence (AI)-enhanced digital platform that empowers communities to engage more effectively in decisions about public infrastructure and services.

In the first year, we will collaborate with nonprofit organizations to develop an AI-enabled chatbot system that elicits opinions, provides feedback, and educates organizational staff about infrastructure projects and public services. The system will be accessible via smartphones, tablets, and computers. Partnering with experienced nonprofits ensures that the system is grounded in real-world issues and goals.

In the second year, we will expand this chatbot system into a group communication platform to facilitate continuous public participation through AI personas representing nonprofit staff and community members. These AI personas, informed by data from the first year and household surveys, will enable community members to engage asynchronously, learn from one another, and provide feedback to policymakers at their convenience. The platform will also support the exchange of images and videos among real human members and AI personas, significantly enhancing community communications beyond conventional in-person public sessions and virtual meetings.

Integrating AI into participatory planning presents risks, particularly in building trust. Marginalized individuals are reluctant to engage with other community members, and introducing AI into this sensitive process could increase distrust. However, if the concern is addressed properly, the project has the potential to yield substantial benefits by connecting individuals currently siloed by factors such as ethnicity, language, gender, and physiological conditions.

This interdisciplinary project involves key collaborators from (1) civil infrastructure and mobility service design, (2) digital platform and information system development, (3) political influence of nonprofits, and (4) human behavior and urban science. Together, the collaboration aims to pioneer innovative approaches to community engagement that are responsive, inclusive, and effective across various contexts and scales.

More metals are needed for net-zero transition, likely from both mining and recycling. The minerals sector faces increasing pressure to mitigate its environmental impact and align with global climate goals. Nature-based solutions (NbS), focusing on ecosystem restoration, afforestation, and sustainable land management, offer socially accepted pathways to accelerate the sector’s climate transition. However, this transition requires considerable natural, social, and financial capital. There is a disconnect among climate scientists, engineers and investor communities regarding the need to work together to mobilize different types of capital for climate action. One major barrier is the provision of timely and accurate environmental information, as it is crucial to drive investors’ preference for or confidence in responsible investment. Thus, how to more efficiently connect evidence-based climate solutions with resources for responsible investment is an urgent need to achieve ambitious global climate goals. This project positions the minerals sector as a proactive participant in global climate mitigation efforts, demonstrating the strategic value of nature-based solutions in enhancing climate resilience and responsible investment. Our interdisciplinary team will begin by conducting a comprehensive review of NbS practices in the extractive sector and their impact on carbon sequestration and environmental restoration. 1) We will then apply advanced machine learning techniques to enhance the quality and representativeness of current practices on carbon sink quantification. 2) This refined quantification of changes in carbon sinks and potential co-benefits for the minerals sector will be used to assess how NbS can be further integrated into environmental, social and governance (ESG) investment criteria created by mainstream ESG rating providers. 3) With an updated ESG assessment proposal, we will survey fund managers to seek feedback on evaluating the effectiveness of NbS-related information provision in the ESG framework, facilitating more informed and sustainable investment decisions. We expect this project to produce one of the first tools to support more transparent, reliable, and verifiable carbon measurement for sustainable finance and, through our unique training opportunities, nurture well-qualified talents to promote nature-based solutions for accelerating decarbonization.