Decoding Neurodegeneration via Quantitative MRI and Brain Network Analytics

Abstract:  

Magnetic Resonance Imaging (MRI) stands as one of the cornerstones of neuroimaging, offering a window into the intricate organization of the human brain. In this talk, I will present my research on the integration of classical signal processing theory and machine learning in MRI, aiming to advance our understanding of neurodegenerative diseases. First, by applying a Bayesian framework to model-based quantitative MRI, my work produces high-quality 3D brain imaging from undersampled data while significantly reducing scan times to enhance the patient experience. This method automates hyper-parameter tuning, traditionally a manual process, and operates seamlessly with various MRI protocols and scanners. Second, my work leverages Graph Neural Networks (GNN) to interpret complex brain networks derived from rs-fMRI data. It leads to a more accurate prediction of brain age, a well-regarded biomarker for age-associated neurodegenerative diseases. Finally, I will conclude my talk by discussing promising directions that aim to shape the trajectory of MRI-centric neuroimaging with AI-driven techniques.

Bio:

Shuai Huang is currently an Assistant Professor in the Department of Electrical and Computer Engineering at Auburn University, he received his PhD in Electrical and Computer Engineering from Johns Hopkins University. His research leverages machine learning and signal processing tools to address challenges in MRI and Neuroimaging, with the goal of advancing our understanding and treatment of a vast array of neurodegenerative diseases. His work has contributed to the automatic reconstruction of quantitative MRI and susceptibility maps for measuring the iron deposition in the brains of patients with Alzheimer’s disease.