Convergent Research Themes
Academic Year - 2022/2023
Ìý
Predicting the local impact of regional extreme weather events in smart cities
Overview:ÌýThis theme explores the feasibility of coupling Numerical Weather Prediction models with Computational Fluid Dynamic models in order to quantify local influences of severe weather on smart cities. It will also explore the best strategies to communicate the results to decision-makers.
It represents areas in atmospheric sciences, structural and wind engineering, geographic information systems, and urban sustainability and resilience.
Core Team:
Djordje Romanic
Atmospheric and Oceanic Sciences, Faculty of Science
Laxmi Sushama
Civil Engineering, Faculty of Engineering
Raja Sengupta
Geography, Faculty of Science
Ìý
Applications of natural language processing in clinical care at ÎÛÎÛ²ÝÝ®ÊÓƵ
Overview: This theme aims at identifying clinical needs that can be best addressed with NLP-based tools, in order to improve patient outcomes. Research questions include structuring clinical text (from typed medical reports or interviews), the use of health chatbots, and mining medical literature to discover latent associations.
It represents areas of NLP, clinical outcomes, evaluative research and health services delivery.
Core Team:
Dan Poenaru
Pediatric Surgery, Faculty of Medicine
Jackie Cheung
Computer Science, Faculty of Science
Esli Osmanlliu
Pediatrics, Faculty of Medicine
Samira Rahimi
Family Medicine, Faculty of Medicine
Ìý
Using machine learning and natural language processing to predict real-world consumer decision-making and evaluation
Overview: This theme will explore applying machine learning and NLP tools to a very large data set of consumer choices and reviews, in order to predict decision-making and textual content of reviews. On a broader scale, this theme will develop computational methods for generating psychological insights from text.
It represents areas of cognitive neuroscience, decision-making, big data methodologies, machine learning and natural language processing.
Core Team:
Ìý
Ross Otto
Psychology, Faculty of Science
Bruce Doré
Marketing, Desautels Faculty of Management
Brendan Johns
Psychology, Faculty of Science
Ìý
Challenges and rewards of developing an intelligent technology for high-risk biomedical environments
Overview: This theme will explore the development of an Intelligent Personal Assistant that will aid in planning, safety and day-to-day operations in high risk environments such as Containment Level 3 (CL3) laboratories. The first stage of the project will involve identifying needs and limitations of CL3 environments and creating software testing protocols to be evaluated first in lower risk laboratories.
It represents areas of software engineering, computer-human interactions, machine learning and natural language processing, and biomedical methods and protocols.
Core Team:
Jérôme Waldispühl
Computer Science, Faculty of Science
Silvia Vidal
Human Genetics, Faculty of Medicine
Elena Nazarova
Computer Science, Faculty of Science
Ìý
Developing a deep learning algorithm to improve cancer treatments
Overview: This theme aims at developing a deep learning algorithm for auto-segmentation of extremity soft tissue sarcomas (STS), and evaluating radiation doses to the areas that will be irradiated. Emphasis will be placed on the evaluation of the different volumes to be irradiated, which will give insights into the clinical significance of auto-segmentation.
It represents areas of STS imaging, DL auto-segmentation, and radiation therapy planning.
Core Team:
James Tsui
Radiation Oncology, ÎÛÎÛ²ÝÝ®ÊÓƵ University Health Centre
Carolyn Freeman
Radiation Oncology, ÎÛÎÛ²ÝÝ®ÊÓƵ University Health Centre
Shirin Enger
Medical Physics Unit, Gerald Bronfman Department of Oncology, ÎÛÎÛ²ÝÝ®ÊÓƵ University.
Ahmed Aoude
Orthopaedic Surgery, Faculty of Medicine
Anthony Bozzo
Orthopedic Oncology, Memorial Sloan Kettering Cancer Center
Orthopaedic Surgery, Faculty of Medicine
Sungmi Jung
Pathology, Faculty of Medicine