Advancing microstructure imaging via diffusion MRI: harnessing modern machine learning and advanced computational models

Abstract :

Diffusion MRI has emerged as one of the most powerful non-invasive tools for probing tissue microstructure, offering unprecedented opportunities to characterize the cellular architecture of the brain and other organs in health and disease. In this talk, I will present recent advances in microstructure imaging that combine cutting-edge diffusion MRI, computational biophysical modelling, and modern machine learning to push beyond the limits of conventional imaging. Drawing on pioneering work on generative models of brain tissue microstructure, the presentation will illustrate how advanced computational frameworks can reveal cellular features such as soma size, neurite density, membrane permeability, and water exchange processes in vivo. The talk will further highlight how Monte Carlo simulations, deep learning, and uncertainty-aware inference methods are enabling more robust and biologically meaningful biomarkers from diffusion MRI data.

Bio:

Marco Palombo is a UKRI Future Leaders Fellow and Associate Professor (Senior Lecturer) in Microstructure Imaging at Cardiff University, based at the Cardiff University Brain Research Imaging Centre (CUBRIC). He leads the Microstructure Imaging research programme at CUBRIC, where his work combines physics, computer science, and neuroscience to develop non-invasive imaging technologies for the early diagnosis and prognosis of neurological and psychiatric disorders.

Website: https://profiles.cardiff.ac.uk/staff/palombom