Victor Solo, PhD: Identifying Functional Brain Networks with Conditional Independence Graphs
Victor Solo, PhD
MGH/HST Martinos Center for Biomedical Imaging, Dept. Radiology, MGH
Harvard Medical School, Boston, MA
School of Electrical Engineering, UNSW, Sydney, Australia
Identifying Functional Brain Networks with Conditional Independence Graphs
June 1, 2023
Abstract: We discuss, in a straightforward way, some basic concepts required for the coherent construction of functional brain networks from (neuroimaging) data e.g. signals at each node of an anatomically parcellated network. The dominant approach involves correlation, and we expose its flaws. These are remedied by using partial correlation which leads to conditional independence graphs (developed in the statistics literature beginning in the 1970s) which in turn need sparsity methods to scale up to big networks. We then sketch the extension to autocorrelated signals (aka lagged correlation signals) which leads to frequency domain based functional network construction based on partial coherency and state space models. We illustrate some of the ideas with applications to real data. We have ground down the math to an absolute minimum.
Athinoula A. Martinos Center for Biomedical Imaging
https://www.martinos.org/
Follow @MGHMartinos on Facebook, Instagram + Twitter!
Смотрите видео Victor Solo, PhD: Identifying Functional Brain Networks with Conditional Independence Graphs онлайн без регистрации, длительностью часов минут секунд в хорошем качестве. Это видео добавил пользователь MGH Martinos Center 20 Июнь 2023, не забудьте поделиться им ссылкой с друзьями и знакомыми, на нашем сайте его посмотрели 358 раз и оно понравилось 7 людям.