Stroke: Vascular and Interventional Neurology logo
URL with Digital Object Identifier
The cause of ischemic stroke often remains elusive even after full stroke workup is completed. Cardioembolic mechanisms in particular are frequently presumed but challenging to definitively diagnose. Quantitative thrombus texture analysis is emerging as a powerful tool for stroke characterization, having shown the ability to predict response to stroke treatment,1 but its ability to predict stroke cause and complement machine learning models built from standard clinical features has not been studied.2, 3 The purpose of this study is to evaluate the ability of radiomics features extracted from quantitative magnetic resonance images of retrieved ischemic stroke thrombi (R2*(=1/T2*), quantitative susceptibility mapping, and fat fraction) to improve the accuracy of machine learning models built from clinical data for the prediction of cardioembolic stroke.
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.