During Michaelmas Term, OxTalks will be moving to a new platform (full details are available on the Staff Gateway).
For now, continue using the current page and event submission process (freeze period dates to be advised).
If you have any questions, please contact halo@digital.ox.ac.uk
Teams link: www.win.ox.ac.uk/events/win_seminar_mar_2025
In this talk, I will summarize work done at my group over the last decade, on brain image analysis of brain MRI. I will start with our initial efforts on high-resolution atlasing using ex vivo MRI and histology, with an emphasis on our recently released NextBrain, which is the first probabilistic atlas of the whole human brain built with high-resolution 3D histology, and which can be used to automatically segment MRI scans (in vivo or ex vivo) into 300+ regions of interest. Then I will move on to our family of deep convolutional neural networks trained on synthetic, domain-randomized data. These networks can, for the first time, segment MRI scans of any orientation, resolution, and contrast “out of the box”, without retraining. I will show applications to large-scale analysis of clinical scans, models of pathology, analysis of low-field scans acquired with a portable device, and foundation models.