Join the Machine Learning in Medical Imaging Consortium (MaLMIC) for a networking opportunity
Monthly Virtual Forum Series on Zoom
Friday, July 16, 2021
3:30 to 5:30 p.m. EDT
Sandra Black, a neurologist, talked about the clinical perspective followed by two researchers, Maged Goubran and Lyndon Boone, describing their machine learning research in the area. This was followed by discussion focused on lessons learned, opportunities to collaborate and sharing of resources.
Monthly Virtual Forum Series on Zoom
Friday, July 16, 2021
3:30 to 5:30 p.m. EDT
Sandra Black, a neurologist, talked about the clinical perspective followed by two researchers, Maged Goubran and Lyndon Boone, describing their machine learning research in the area. This was followed by discussion focused on lessons learned, opportunities to collaborate and sharing of resources.
Reconciling complexity and precision medicine: Can Artificial Intelligence really capture real-world ground truth in the common dementias?
Sandra E Black, MD
Professor, University of Toronto, and Senior Scientist, Sunnybrook Research Institute
Talk summary: The Sunnybrook Dementia study exemplifies research embedded in care, a 25 year venture combining standardized, comprehensive neurocognitive, mood, behaviour and functional measures, with real-world pipelines for MRI volumetrics, developed collaboratively with radiology, medical biophysics, and computer programmers, to map topographies of the common dementias and elucidate cerebral small vessel disease. This provided essential ground truth for developing Computational Neural Networks, turning hours of labour into minutes of computation that generate quick, reliable structural diagnostics to support personalized treatments and outcome evaluation.
Building artificial neural networks for neuroimage analysis: applications in dementia and stroke
Maged Goubran, PhD
Assistant Professor, University of Toronto, and Scientist, Sunnybrook Research Institute
Talk summary: Maged will describe some of the novel ML and computational tools that they are building for neuroscience applications and precision medicine, and their impact in the context of Alzheimer’s disease (AD) and traumatic brain injury (TBI). Maged will specifically present their networks for analyzing structural MRI biomarkers, and highlight how they are improving on existing segmentation methods. He then presented their novel ML pipelines for high-throughput analysis and characterization of neuronal and vascular network alterations in animal models of AD and TBI. Finally, Maged described their recent AI model for predicting subject-specific cognitive decline in preclinical AD.
Sensitivity of Convolutional Neural Networks to Common Imaging parameters, pertubations and artifacts in MRI
Lyndon Boone, MSc student
University of Toronto and Sunnybrook Research Institute
Talk summary: A number of studies have shown that deep learning methods are capable of achieving near-human-level performance on neuroimaging segmentation tasks. With that said, most of the results quoted in the literature supporting this statement are in the context of test sets drawn from the same overarching dataset as the training data. Clinically-deployed models faced with out-of-distribution data (i.e. data that does not resemble the training set) may severely underperform relative to the standards set in the literature if they aren’t designed specifically with robustness to out-of-distribution data in mind. In this talk, Lyndon highlighted the sensitivity of modern CNN-based architectures to image corruptions, artifacts, and post-processing transforms commonly found in MRI. He then presented a methodology for benchmarking different architectures and models on the basis of robustness to out-of-distribution data, inspired by similar work in the computer vision literature.