Prof. Pascal Tyrell of Department of Medical Imaging, Temerty Faculty of Medicine, University of Toronto

The development and deployment of machine learning algorithms for medical imaging has become increasingly popular in recent years. These algorithms rely on large amounts of training data to learn patterns and features in the images that can be used for tasks such as segmentation and classification. However, machine learning algorithms can suffer from various forms of bias, such as sampling bias, measurement bias, and algorithmic bias. Sampling bias can occur when the training data do not accurately represent the population that the algorithm will be used on. Measurement bias can occur when the measurements used in the training data are not consistent or standardized. Algorithmic bias can occur when the algorithm is designed or trained in a way that reflects societal biases. Addressing bias in machine learning algorithms for medical imaging is critical for ensuring accurate and equitable healthcare.

Biography:

Dr. Pascal Tyrrell is a data scientist—a combination of research methodologist, computer/database solutions architect, and innovator. He received his PhD in medical sciences from the University of Toronto working in pediatric rheumatology at SickKids (Toronto, ON). Currently, he is the Director of Data Science and Associate Professor with the Department of Medical Imaging, Temerty Faculty of Medicine, University of Toronto where he is also the founding director of the MiDATA Data Science program. Dr. Tyrrell is cross appointed to the Institute of Medical Science and the Department of Statistical Sciences where his research aims to introduce statistically sound and innovative Artificial Intelligence and Machine Learning approaches to the study of medical images in health-related outcomes research. He has previous work experience in the computer, financial, and medical device industries, and is the CEO and co-founder of the software startup company SofTx Innovations Inc.”