.png&w=3840&q=75)
Location
Date
Title
Pushing the resolution limits of image analysis in brain MRI (in both directions)
Abstract
Most tools for analyzing brain MRI are designed for T1-weighted scans with a resolution of approximately one millimeter in all three dimensions. Yet modern MRI scans often reveal brain anatomy at finer scales, while hospital archives contain enormous collections of scans acquired at lower resolution. In this talk, I will describe how we are pushing the resolution limits of image analysis in both directions: exploiting the anatomical detail in modern high-resolution images and unlocking the wealth of information in existing clinical scans. I will begin with ex vivo atlases, which are detailed anatomical maps built from high-resolution images of donated human brains outside the body, and which can be used for fine-grained analysis of high-resolution MRI. I will then turn to the vast collections of clinical brain scans stored in hospital imaging archives around the world. Many of these scans have thick slices (i.e., low resolution in one of the three spatial dimensions) and widely varying imaging characteristics, making them difficult to analyze with conventional tools. I will describe generalizable neural networks that enable analysis across these differences, opening up existing data for large-scale studies, with applications extending to portable MRI. Finally, I will close with a look at ongoing work and future directions, particularly in neuropathology and cell atlasing.
Speaker
Juan Eugenio Iglesias, Associate Professor of Radiology, MGH & Harvard Medical School, Research affiliate at Massachusetts Institute of Technology, and Honorary Research Fellow at University College London.
Bio
Juan Eugenio Iglesias holds M.Sc. degrees in Telecom and Electrical Engineering from the University of Seville (Spain) and the Royal Institute of Technology (KTH, Stockholm, Sweden), respectively. He received his Ph.D. in Biomedical Engineering from UCLA sponsored by a Fulbright grant. After two postdoctoral positions at the Martinos Center for Biomedical Imaging (MGH / Harvard Medical School) and the BCBL (sponsored by a Marie Curie fellowship), he moved to University College London as junior faculty with a Starting Grant of the ERC. In 2019, he moved back to the Martinos Center in Boston, where he is now Associate Professor. He also holds an affiliate appointment at MIT.