Meeting Banner
Abstract #3955

Evaluation of automatic segmentation of perivascular spaces in multiple sclerosis patients and healthy controls on 3T MRI

Joan Brewer1, Ilena George2, Gale Edison1, Joy Zhang2, Sarah King2, James Sumowski2, Priti Balchandani2, Gaurav Verma2, Rebecca Feldman1, and Sam Horng2
1University of British Columbia Okanagan, Kelowna, BC, Canada, 2Icahn School of Medicine at Mount Sinai, New York, NY, United States

Synopsis

Multiple Sclerosis (MS) is an autoimmune disease; disease can be measured on MRI by the appearance or enlargement of lesions. PVS measurements on MRI may represent a more predictive biomarker of disease. Because manual measurement of PVSs is time intensive, our collaborators developed a tool to automatically segment PVSs. Here, we investigated the utility of the automatic segmentations. We correlated the automated PVS counts with manual segmentations (0.66), as well as the PVS areas (0.55). Differences between healthy control and MS groups in this preliminary analysis were not statistically significant. Future work will expand from 45 to 300 patients.

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

Join Here