7/15/2020: Machine Learning (ML) Generated Indices of ‘Brain Health’ from Medical Images

Presented on July 15, 2020 by trainees in the MIT Health Sciences Technology (HST) division to their peers and faculty of the Neuroimaging Training Program.

Advances in technologies that enable human brain imaging to inform clinical care are proliferating at an astounding pace. The half-day virtual symposium, “Machine Learning (ML) Generated Indices of ‘Brain Health’ from Medical Images”, opened with an expert overview of the field followed by succinct reviews of a range of topics relevant to the use of neuroimaging to improve and/or maintain brain health. These include applications of machine learning to high-fidelity brain image reconstruction, statistical techniques for robust analysis of brain data, machine learning applications to treatment design, and ethical and regulatory challenges to deploying any of these solutions. Each of the graduate student presenters selected their specific topic to harmonize across the goal of the symposium and their own thesis project domain.

The program was sponsored by the McCance Center for Brain Health at MGH, as part of our mission to harness these advances to identify and study neuroimaging indicators of brain health. As these presentations make clear, while there is still significant work to be done before neuroimaging becomes a cornerstone of primary care, the value of neuroimaging to yield meaningful information about brain health is unquestionable.

Recorded Sessions

Simon Eickhoff, PhD: “Quantitative Markers of Brain Health”
Mitchell Robinson: “Using Machine Learning to Work with Physiological Signals in fMRI Data”
John Gustaf Wilhelm Samuelsson: “Machine Learning in Health Care: Regulatory Framework”
Jordan Harrod: “Applications of Machine Learning for Transcutaneous Treatment to Heal the Depressed Brain”
Katharina Hoebel: “Role of Bias in Automatic Brain Health Assessment”
Jay Patel: “Assessment of Single (Subject) fMRI Task Based Variance”
Ellen Degennaro: “Current Status of Brain Pathology Informatics”