Generative AI for Faster and better MRI

Mathews Jacob
UVA ECE

Time: 2024-10-16, 12:00 - 13:00 ET
Location: Rice 540 and Zoom

Abstract Magnetic Resonance Imaging (MRI) is an indispensable tool in modern medicine, renowned for its exceptional soft-tissue contrast, non-invasive nature, and absence of known long-term side effects. Its ability to provide both functional and molecular imaging makes it crucial for diagnosing a wide range of conditions. However, a significant challenge with MRI is the inherently slow acquisition process, which contributes to patient discomfort and motion artifacts, but also increases medical costs and limits access.

This talk will explore how recent advances in generative AI can be capitalized to revolutionize MRI by dramatically accelerating scan times, enhancing image quality, and unlocking new diagnostic capabilities. I will discuss the unique challenges in medical imaging, such as limited training data, the need for high accuracy and reliability, and fast processing times, along with opportunities like the well-established imaging physics and models that can be leveraged. Recent work from our group that combine imaging physics with AI models with additional constraints to guarantee robustness and converge will be presented. Finally, I will discuss remaining challenges and opportunities in integrating the AI-driven innovations to improve patient experiences, optimize healthcare efficiency, and broaden the potential for personalized medicine.

Bio Mathews Jacob is a professor in the Department of Electrical and Computer Engineering and is heading the Computational Biomedical Imaging Group (CBIG). His research interests include image reconstruction, image analysis, and quantification in the context of magnetic resonance imaging. Dr. Jacob is the recipient of the NSF CAREER award, American Cancer Society Research Scholar Award, and the University of Iowa Faculty Excellence. He is a distinguished lecturer of the IEEE Signal Processing Society, 2025 and was elected as a Fellow of the IEEE (2022) for contributions to computational biomedical imaging.