Title: Leveraging Large-Scale Datasets for Dynamic Brain Network Modelling

Speaker: Chetan Gohil (Research Fellow, Research Department of Epilepsy, Institute of Neurology, UCL)

Location: In Person & Online via Zoom

Abstract: Many neuroimaging studies are designed to address specific cognitive or clinical questions and consequently rely on relatively small datasets. While such 'boutique' studies are well targeted, small sample sizes can limit statistical power and robustness. At present, large-scale neuroimaging datasets are becoming increasingly available. In this talk I will discuss opportunities for exploiting these large-scale datasets in smaller boutique studies, focusing on dynamic network modelling with MEG and EEG. I will review generative modelling approaches such as the Hidden Markov Model and emerging approaches from deep learning known as 'foundation models'. I will highlight both the potential benefits and limitations of these approaches.