Abstract #3506

Deep learning prediction of retrieved stroke thrombus RBC content using quantitative, multiparametric MRI

Spencer D. Christiansen^1,2, Junmin Liu², Maria Bres Bullrich³, Manas Sharma⁴, Sachin K. Pandey⁴, Mel Boulton³, Luciano A. Sposato³, and Maria Drangova^1,2

¹Medical Biophyics, Western University, London, ON, Canada, ²Robarts Research Institute, London, ON, Canada, ³Clinical Neurological Sciences, London Health Sciences Centre, London, ON, Canada, ⁴Department of Medical Imaging, Western University, London, ON, Canada

Thrombus red blood cell (RBC) content has been associated with ischemic stroke etiology and responsiveness to recanalization therapies, yet currently can only be analyzed through retrospective histological analysis. We evaluated the ability of a convolutional neural network for predicting thrombus RBC content using multiparametric (R₂^*, QSM, late echo GRE) MR image slices of retrieved stroke thrombi ex vivo. The network predicted thrombus RBC content with an accuracy and mean absolute error of up to 71 and 8%, respectively, when data augmentation was applied. This technique holds potential for in vivo RBC content prediction and improving acute stroke care.

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

Join Here