開催日時：2023年02月21日 11:00-12:00

開催場所：545演習室・オンライン（※学外の方でも参加できます。）

講演者： Thomas A. W. Bolton (Connectomics Laboratory, Department of Radiology, Centre Hospitalier Universitaire Vaudois, Switzerland)

題目：
Novel analytical approaches for functional magnetic resonance imaging: graph signal processing and the arrow-of-time

概要：
Functional magnetic resonance imaging (fMRI) is a non-invasive way to indirectly quantify brain activity over the course of time. Over the last two decades, many have demonstrated how statistical interdependence between regional brain activity time-courses, termed functional connectivity (FC), could advance our understanding of the healthy and diseased brain.
Recently, the focus has shifted towards exploiting more sophisticated features of fMRI data. Here, I will outline two emerging directions for this purpose. First, I will explain how diffusion-weighted MRI data enables the quantification of physical wiring between brain regions, or structural connectivity (SC). SC can then be conceived as an irregular lattice over which regional functional signals are analyzed, a burgeoning field known as graph signal processing (GSP).
Second, I will introduce a novel approach that quantifies the extent of irreversibility in fMRI data, often referred to as the arrow-of-time (AoT). A first-order autoregressive model is applied to forward and backward time-courses, and the non-Gaussianity of the model’s residuals is contrasted as a measure of AoT strength, yielding structured and temporally evolving patterns distinct from those obtained with more classical FC-based assessments.

お問い合わせは以下のアドレスにメールでご連絡ください
dser-center@biwako.shiga-u.ac.jp