Related Experiment Video
Updated: Apr 26, 2026
Author Spotlight: Creating a Versatile Experimental Autoimmune Encephalomyelitis Model Relevant for Both Male and Female Mice
Published on: October 13, 2023
Predicting multiple sclerosis from radiologically isolated syndrome using generative artificial intelligence
Christine Lebrun-Frenay1,2, Felix Renard3, Lydiane Mondot1,2
1UR2CA-URRIS, Université Nice Côte d'Azur, Nice, France.
Abstract:
Radiologically Isolated Syndrome (RIS) is characterized by incidental MRI findings indicative of multiple sclerosis (MS) in asymptomatic individuals. Factors such as younger age, positive cerebrospinal fluid biomarkers, and specific lesion locations have been previously linked to a higher risk of conversion from RIS to clinical MS. Predicting which individuals will develop clinical MS remains challenging. Based on widely available cross-sectional patient studies, unsupervised machine learning has been proposed to uncover MRI-driven MS phenotypes with distinct temporal progression patterns. We evaluated whether an unsupervised artificial intelligence framework based on generative manifold learning could stratify RIS patients by conversion risk. BrainGML-MS analyzed imaging biomarkers and generated individualized digital twins from MRI data. We studied 152 RIS individuals (32 converters, RIS-C), 152 MS patients, and 152 healthy controls. The model identified four RIS clusters with distinct five-year conversion risks ranging from 10% to 39%. The brain age gap increased progressively from healthy controls to RIS non-converters, RIS-C, and MS. RIS converters showed greater structural atrophy and greater similarity to MS profiles. These findings indicate that MRI-derived brain aging biomarkers and structural deviations measured at the first RIS scan may improve early risk stratification and support clinical decision-making in preclinical MS.
Insights
Radiologically Isolated Syndrome (RIS) patients can be stratified by conversion risk using AI analysis of MRI scans. This approach identifies distinct patient clusters, aiding in early prediction of conversion to multiple sclerosis (MS).
Area of Science:
- Neurology
- Radiology
- Artificial Intelligence
Background:
- Radiologically Isolated Syndrome (RIS) presents incidental MRI findings suggesting multiple sclerosis (MS) in asymptomatic individuals.
- Predicting conversion from RIS to clinical MS is challenging, despite known risk factors like age and biomarkers.
- Unsupervised machine learning offers potential for uncovering MRI-driven MS phenotypes and progression patterns.
Purpose of the Study:
- To evaluate an unsupervised artificial intelligence (AI) framework using generative manifold learning for stratifying RIS patients by conversion risk.
- To assess the utility of AI-generated digital twins from MRI data for risk stratification.
- To identify distinct RIS patient clusters with varying five-year conversion risks.
Main Methods:
- Studied 152 RIS individuals (32 converters), 152 MS patients, and 152 healthy controls.
- Utilized an AI framework (BrainGML-MS) for analyzing imaging biomarkers and generating individualized digital twins from MRI data.
- Applied unsupervised generative manifold learning to stratify RIS patients.
Main Results:
- The AI model identified four RIS clusters with five-year conversion risks from 10% to 39%.
- A progressive increase in brain age gap was observed from healthy controls to RIS non-converters, RIS converters, and MS patients.
- RIS converters exhibited greater structural atrophy and resembled MS patient profiles more closely.
Conclusions:
- MRI-derived brain aging biomarkers and structural deviations at the initial RIS scan can improve early risk stratification.
- AI-driven analysis of MRI data shows promise for supporting clinical decision-making in preclinical MS.
- This approach may enhance the identification of individuals at higher risk of developing clinical MS from RIS.
