This study uses unsupervised machine learning and brain MRI scans from 6322 MS patients (training dataset) and 3068 (validation dataset) to identify MS subtypes based on pathological features. Three subtypes were defined: cortex-led, normal-appearing white matter (NAWM)-led, and lesion-led. The lesion-led subtype showed the highest risk of disability progression and relapse rate, but also a positive treatment response in selected clinical trials. Findings suggest that MRI-based subtypes predict MS disability progression and treatment response, potentially improving patient stratification in interventional trials.
Publisher
Nature Communications
Published On
Apr 06, 2021
Authors
Arman Eshaghi, Alexandra L. Young, Peter A. Wijeratne, Ferran Prados, Douglas L. Arnold, Sridar Narayanan, Charles R. G. Guttmann, Frederik Barkhof, Daniel C. Alexander, Alan J. Thompson, Declan Chard, Olga Ciccarelli
Tags
multiple sclerosis
machine learning
MRI scans
subtypes
disability progression
treatment response
patient stratification
Related Publications
Explore these studies to deepen your understanding of the subject.