CHBV: Bruce Fischl, Synthesis-Driven Deep Learning and Multi-Scale Analysis of the Human Brain

Friday, December 6
1:00-2:30 pm
Broad Institute of MIT and Harvard
Our Seminar Series is a quarterly series designed to make cutting-edge neuroscience research accessible to all. Our lineup features speakers from a variety of career stages with diverse backgrounds and expertise. We believe in the power of diversity and inclusion, recognizing that a multitude of voices contributes to a richer scientific dialogue.
Register for our upcoming events, read about the speakers, watch previous talks, and delve into the fascinating world of neuroscience. Join us as we connect minds, share knowledge, and pave the way for a more accessible and collaborative scientific community.
Synthesis-Driven Deep Learning and Multi-Scale Analysis of the Human Brain
Bruce Fischl
Harvard Medical School
MGH AA Martinos Center
Bruce Fischl, PhD (MGH, HMS) discusses technical advancements in deep learning applied to neuroimaging in his talk entitled, Synthesis-driven deep learning and multi-scale analysis of the human brain.
Bruce Fischl is a Professor of Radiology at Harvard Medical School and Director of the Computational Core at the MGH AA Martinos Center. His talk will discuss technical advancements in deep learning applied to neuroimaging, particularly addressing the challenges posed by the domain gap, which affects the performance of algorithms trained on specific magnetic resonance imaging (MRI) image types. To overcome this, his group has developed synthetic image synthesis techniques that enable networks to learn without bias from real MRI data, facilitating supervised learning in cases where manual annotation is challenging. He will review his results from a five-year project focused on creating tools for analyzing data across different scales, from micron-level microscopy to whole-brain MRI, aimed at constructing a detailed cellular and connectional atlas of the human brain.
Copyright Broad Institute, 2024. All rights reserved.