Meeting Banner
Abstract #1680

WMH and PVS mapping from clinical data using semi-supervised multi-modal convolutional neural network

Farshid Sepehrband1, Giuseppe Barisano2, Hyun-Joon Yang2, Jeiran Choupan2, and Arthur W Toga2
1Stevens Neuroimaging and Informatics Institute, Keck School of Medicine, USC, Los Angeles, CA, United States, 2USC, Los Angeles, CA, United States

Perivascular space (PVS), also known as Virchow-Robin space, is a pial-lined, fluid-filled structure that surrounds vessels in the cerebral cortex [1], [2], and occupies a large portion of the cerebral tissue. While PVS mapping becomes more clinically relevant, separation of PVS from white matter hyperintensities (WMH) make it challenging to map PVS on clinical data. Here we present a semi-supervised multi-modal approach to extract both PVS and WMH using T1w and FLAIR image modalities automatically. We also evaluated multi-site, multi-vendor reliability of the technique.

This abstract and the presentation materials are available to members only; a login is required.

Join Here