Gabriel Devenyi


Gabriel comes to Neuroscience after a PhD in Engineering Physics at McMaster University where he studied the crystal structure of materials for energy and electronics applications. In 2014 he joined the Computational Brain Anatomy (CoBrA) Lab at the Douglas McGill Mental Health Institute and Department of Psychiatry. At the Douglas, Gabriel spends his day as an “expert generalist”, where he develops neuroimaging software, manages a cluster, provides experimental and statistical consulting, and investigates methodological neuroimaging questions.

One of the most important inputs to modern machine learning applications is large amounts of data in which to find or learn patterns to make predictions. Neuroimaging has a huge amount of publicly available data in the form of raw MRI scans of subjects with neuropsychiatric disorders, but these raw images are ill suited by themselves to be used in machine learning. Machine learning experts rarely have the background in neuroimaging necessary to convert these raw images into meaningful data for use. The PONDR AI project aims to fill this gap, by collecting public images, processing them into meaningful measurements of human brain anatomy, and making the resulting data publicly available for use by machine learning researchers.