Virtual Open Forum

Machine Learning in Breast Cancer Imaging

Machine Learning for Medical Imaging Consortium
Friday, November 19, 2021
3:30-5:30PM Eastern Time

Dr. Emily Conant, a radiologist from the University of Pennsylvania, will give the lead-off talk presenting the clinical problems and opportunities where machine learning could make significant impact. She will be followed by Professor Nico Karssemeijer from Radboud University in the Netherlands. He has developed important algorithms for breast density and is implementing AI approaches to detection of breast cancer. Grey Kuling, PhD candidate from Sunnybrook Research Institute will give the third talk covering the more technical approach of their research work. The talks will be followed by discussion focused on lessons learned, opportunities to collaborate and sharing of resources.

Director

Parvin Mousavi

Prof. Comp/ECE/Pathology, member of the Royal Soc of Canada, College of New Scholars

As the world’s population grows to 10B by 2050, technology-enabled innovation is the critical to ensuring well-being for all. While AI and informatics show great promise in rapid analysis of next-generation large scale data, future workers will require highly specialized training fit to the unique nature of health data. We aim to solidify Canada’s competitive advantage in the global space through concerted efforts to train computer scientists with specialized multi-disciplinary experience in medical informatics and digital health, and engage diverse groups and lived experiences in our training.

Projects

Integration, federation, and retrieval of large-scale data repositories in Canada

We collaborate with the Canadian Institute of Health Information (CIHI) and the Ontario Health Data Platform (OHDP) to tackle novel challenges in data management and devise evaluation frameworks that center on security, performance and fairness.

Next generation of actionable prescriptive analysis using multi-resolution, multi-modality data

We build computation models of disease from multi-omics data using machine learning, deep learning and evolutionary algorithms, focusing on innovations that address the unique nature of health data and the unavailability and vagueness of gold-standard labels (e.g., pathology labels).

Democratization of AI, software and data for healthcare

We're international leaders in developing and disseminating open source software using low-cost point-of-care imaging. Our positive impact on the global computer–assisted medical interventions community has is constantly expanding. We'll meet new challenges for discovery, refinement of methods and innovative AI-powered solutions using large open source data, hence democratizing access to care and positive outcomes of AI.