Related Experiment Video
Updated: Nov 10, 2025
Magnetic Resonance Imaging of Multiple Sclerosis at 7.0 Tesla
Published on: February 19, 2021
Identifying multiple sclerosis subtypes using unsupervised machine learning and MRI data
Arman Eshaghi1,2, Alexandra L Young3,4, Peter A Wijeratne3
1Queen Square Multiple Sclerosis Centre, Department of Neuroinflammation, UCL Queen Square Institute of Neurology, Faculty of Brain Sciences, University College London, London, UK. a.eshaghi@ucl.ac.uk.
Abstract:
Multiple sclerosis (MS) can be divided into four phenotypes based on clinical evolution. The pathophysiological boundaries of these phenotypes are unclear, limiting treatment stratification. Machine learning can identify groups with similar features using multidimensional data. Here, to classify MS subtypes based on pathological features, we apply unsupervised machine learning to brain MRI scans acquired in previously published studies. We use a training dataset from 6322 MS patients to define MRI-based subtypes and an independent cohort of 3068 patients for validation. Based on the earliest abnormalities, we define MS subtypes as cortex-led, normal-appearing white matter-led, and lesion-led. People with the lesion-led subtype have the highest risk of confirmed disability progression (CDP) and the highest relapse rate. People with the lesion-led MS subtype show positive treatment response in selected clinical trials. Our findings suggest that MRI-based subtypes predict MS disability progression and response to treatment and may be used to define groups of patients in interventional trials.
Insights
Machine learning identified three multiple sclerosis (MS) subtypes from MRI scans: cortex-led, normal-appearing white matter-led, and lesion-led. The lesion-led subtype shows the highest disability progression and relapse rates.
Area of Science:
- Neurology
- Radiology
- Artificial Intelligence
Background:
- Multiple sclerosis (MS) classification into four phenotypes lacks clear pathophysiological boundaries, hindering treatment stratification.
- Machine learning offers a method to identify patient subgroups with similar features from complex datasets.
Purpose of the Study:
- To classify multiple sclerosis (MS) subtypes using pathological features identified via unsupervised machine learning on brain MRI scans.
- To validate MRI-based MS subtypes in an independent patient cohort.
Main Methods:
- Unsupervised machine learning applied to brain MRI scans from a training dataset of 6322 MS patients.
- Independent validation performed on a cohort of 3068 MS patients.
- Subtypes defined based on the location of earliest pathological abnormalities: cortex-led, normal-appearing white matter-led, and lesion-led.
Main Results:
- Three distinct MRI-based MS subtypes were identified: cortex-led, normal-appearing white matter-led, and lesion-led.
- The lesion-led MS subtype demonstrated the highest risk of confirmed disability progression (CDP) and the highest relapse rate.
- Patients with the lesion-led MS subtype showed a positive treatment response in specific clinical trials.
Conclusions:
- MRI-based MS subtypes can predict disease disability progression and treatment response.
- These subtypes may serve as valuable criteria for patient stratification in future interventional trials.
- This approach enhances understanding of MS heterogeneity and personalized medicine strategies.
