Risk Estimation of Deterioration for patients in the cardiac ICU

The Artificial Intelligence (AI) initiative at SickKids, supporting the future of individualized paediatric care at SickKids and beyond. We are developing a unique paediatric methodology for the integration of AI and data science into clinical care. Critically ill children in pediatric and cardiac intensive care units are at a significant risk of clinical deterioration, which […]

Read More
L2M – Chase Biotherapeutics: Breaking barriers towards clinical translation of ChASE37-AR for stroke recovery

Stroke is a leading cause of disability, affecting 880,000 Canadians, with 109,000 new cases annually and no approved therapies for neuroregeneration. Current standard of care for ischemic stroke includes acute thrombolysis and/or thrombectomy to restore blood flow to the injured tissue, yet these treatments are time sensitive and only apply to specific stroke etiology, leaving […]

Read More
Development of Climate and Infrastructure Forensic Analysis Systems: A BayesianPerspective

This project will upgrade the Climate and Infrastructure Forensic Analysis System (CIFAS), originally developed to characterize snow- and permafrost-related impacts to ground transportation and mine access roads in the Canadian North. My first objective is to enhance the analytical capacity of CIFAS by improving its ability to quantify uncertainties associated with empirical knowledge used to […]

Read More
Leveraging Large Language Models for Sales & Customer Success Automation

Mash helps revenue teams at technical B2B companies streamline pre- and post-sales activities by automating knowledge-intensive tasks such as answering product questions, managing bug reports and feature requests, and prepping for meetings. Its AI platform leverages data buried in messaging platforms, wikis, CRMs, and other internal tools to deliver timely, context-aware assistance to individuals in […]

Read More
Optimizing Lightstage Capture for High Fidelity 3D Facial Reconstruction

Ubisoft is one of the world’s largest video game studios, specializing in 3D open-world games that require precise 3D character representations. In particular, achieving high-quality facial features is crucial, as humans are highly sensitive to small details in facial expressions. Currently, creating 3D facial representations first requires a Lightstage capture pipeline. This process begins with […]

Read More
A Risk-Based Continuous Authentication Engine Using a Probabilistic Model around Behavioral Biometrics

Traditional static authentication systems have a fundamental deficiency; it assumes the presence of the validated user through the length of the session. Continuous authentication algorithms periodically validate the identity of a user during the entire session. It relies on information that can be automatically extracted from the user such as biometrics and behavior patterns. A […]

Read More
Increasing dairy intake for the reduction of obesity and diabetes in adolescents

Dairy consumption has decreased in the last 30 years. Canada’s Food Guide no longer recommends three servings of dairy and emphasizes plant-based diets and dairy alternatives. Associated with this decrease is an increase in overweight and the risk of obesity and Type 2 Diabetes (T2D) in adolescents. Consistent with this advice, many adolescents have increased […]

Read More
Development of Natural Language Queries Embedded Within a School Data Hub to Support Data-Informed Decision-Making by School-Based Practitioners

The University of Toronto Schools (UTS), through the Eureka Research Institute, is committed to advancing data-driven decision-making in K-12 education. This aligns with the growing recognition of the importance of data literacy in education, where students and teachers need to “read the world with data” and “write the world with data” (Louie, 2022). Recent research […]

Read More
ESROP – KMUTT – Hybrid LSTM-GRU Architecture with Adaptive Attention for Financial Data

This research project focuses on using advanced machine learning techniques to better predict stock prices, specifically targeting stocks from the S&P 500. By combining powerful deep learning methods—such as LSTM and GRU networks—with adaptive attention mechanisms inspired by Transformer models, the project aims to create forecasting systems that can dynamically adapt to changing market conditions, […]

Read More
Expanding and Enhancing Awesense’s Digital Twin Sandbox to Support the Clean Energy Transition

Awesense is a clean tech company on a mission to accelerate the transition to clean energy by simplifying the creation of data-driven applications for a decarbonized, decentralized grid. To enable the complex planning and operational decisions required by distributed energy resources (e.g., solar, wind, batteries, EVs), Awesense developed its Digital Energy Platform. This platform allows […]

Read More