{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,24]],"date-time":"2026-04-24T18:37:55Z","timestamp":1777055875728,"version":"3.51.4"},"reference-count":33,"publisher":"Public Library of Science (PLoS)","issue":"4","license":[{"start":{"date-parts":[[2026,4,24]],"date-time":"2026-04-24T00:00:00Z","timestamp":1776988800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["www.plosdigitalhealth.org"],"crossmark-restriction":false},"short-container-title":["PLOS Digit Health"],"abstract":"<jats:p>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.<\/jats:p>","DOI":"10.1371\/journal.pdig.0001374","type":"journal-article","created":{"date-parts":[[2026,4,24]],"date-time":"2026-04-24T17:45:18Z","timestamp":1777052718000},"page":"e0001374","update-policy":"https:\/\/doi.org\/10.1371\/journal.pdig.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["Predicting multiple sclerosis from radiologically isolated syndrome using generative artificial intelligence"],"prefix":"10.1371","volume":"5","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-3713-2416","authenticated-orcid":true,"given":"Christine","family":"Lebrun-Frenay","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0045-3611","authenticated-orcid":true,"given":"Felix","family":"Renard","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Lydiane","family":"Mondot","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Cassandre","family":"Landes-Chateau","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Adeline","family":"Stewart","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3985-1297","authenticated-orcid":true,"given":"Mikael","family":"Cohen","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6499-1523","authenticated-orcid":true,"given":"Darin T.","family":"Okuda","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Arnaud","family":"Atty\u00e9","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"340","published-online":{"date-parts":[[2026,4,24]]},"reference":[{"issue":"11","key":"pdig.0001374.ref001","doi-asserted-by":"crossref","first-page":"1075","DOI":"10.1016\/S1474-4422(23)00281-8","article-title":"Radiologically isolated syndrome","volume":"22","author":"C Lebrun-Frenay","year":"2023","journal-title":"Lancet Neurol"},{"issue":"9","key":"pdig.0001374.ref002","doi-asserted-by":"crossref","first-page":"800805","DOI":"10.1212\/01.wnl.0000335764.14513.1a","article-title":"Incidental MRI anomalies suggestive of multiple sclerosis: the radiologically isolated syndrome","volume":"72","author":"DT Okuda","year":"2009","journal-title":"Neurology"},{"issue":"8","key":"pdig.0001374.ref003","doi-asserted-by":"crossref","first-page":"3431","DOI":"10.1093\/brain\/awad073","article-title":"The radiologically isolated syndrome: revised diagnostic criteria","volume":"146","author":"C Lebrun-Frenay","year":"2023","journal-title":"Brain"},{"issue":"10","key":"pdig.0001374.ref004","doi-asserted-by":"crossref","DOI":"10.1001\/jamanetworkopen.2021.28271","article-title":"Risk factors and time to clinical symptoms of multiple sclerosis among patients with radiologically isolated syndrome","volume":"4","author":"C Lebrun-Fr\u00e9nay","year":"2021","journal-title":"JAMA Netw Open"},{"key":"pdig.0001374.ref005","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pone.0090509","article-title":"Radiologically isolated syndrome: 5-year risk for an initial clinical event","volume":"9","author":"DT Okuda","year":"2014","journal-title":"PLoS ONE"},{"key":"pdig.0001374.ref006","doi-asserted-by":"crossref","first-page":"407","DOI":"10.1002\/ana.25799","article-title":"Radiologically isolated syndrome: 10\u2010year risk estimate of a clinical event","volume":"88","author":"C Lebrun\u2010Frenay","year":"2020","journal-title":"Ann Neurol"},{"issue":"11","key":"pdig.0001374.ref007","doi-asserted-by":"crossref","first-page":"591","DOI":"10.1038\/nrneurol.2009.157","article-title":"Unanticipated demyelinating pathology of the CNS","volume":"5","author":"DT Okuda","year":"2009","journal-title":"Nat Rev Neurol"},{"issue":"1","key":"pdig.0001374.ref008","doi-asserted-by":"crossref","first-page":"4273","DOI":"10.1038\/s41467-018-05892-0","article-title":"Uncovering the heterogeneity and temporal complexity of neurodegenerative diseases with subtype and stage inference","volume":"9","author":"AL Young","year":"2018","journal-title":"Nature Communications"},{"key":"pdig.0001374.ref009","doi-asserted-by":"crossref","first-page":"613261","DOI":"10.3389\/frai.2021.613261","article-title":"Ordinal SuStaIn: subtype and stage inference for clinical scores, visual ratings, and other ordinal data","volume":"4","author":"AL Young","year":"2021","journal-title":"Front Artif Intel"},{"key":"pdig.0001374.ref010","doi-asserted-by":"crossref","first-page":"104869","DOI":"10.1016\/j.msard.2023.104869","article-title":"Assessing the utility of magnetic resonance imaging-based \u201cSuStaIn\u201d disease subtyping for precision medicine in relapsing-remitting and secondary progressive multiple sclerosis","volume":"77","author":"X Jiang","year":"2023","journal-title":"Mult Scler Relat Disord"},{"key":"pdig.0001374.ref011","unstructured":"McInnes L, Healy J, Melville J. UMAP: uniform manifold approximation and projection for dimension reduction. 2020."},{"issue":"1","key":"pdig.0001374.ref012","doi-asserted-by":"crossref","first-page":"7563","DOI":"10.1038\/s41598-024-58141-4","article-title":"Data-driven normative values based on generative manifold learning for quantitative MRI","volume":"14","author":"A Atty\u00e9","year":"2024","journal-title":"Sci Rep"},{"key":"pdig.0001374.ref013","first-page":"466","article-title":"AssemblyNet: a novel deep decision-making process for whole brain MRI Segmentation.","volume-title":"Lecture notes in computer science","author":"P Coup\u00e9","year":"2019"},{"issue":"6","key":"pdig.0001374.ref014","doi-asserted-by":"crossref","first-page":"990","DOI":"10.1002\/acn3.51782","article-title":"\u201cBrain age\u201d predicts disability accumulation in multiple sclerosis","volume":"10","author":"MR Brier","year":"2023","journal-title":"Ann Clin Transl Neurol"},{"issue":"2","key":"pdig.0001374.ref015","doi-asserted-by":"crossref","first-page":"162","DOI":"10.1016\/S1474-4422(17)30470-2","article-title":"Diagnosis of multiple sclerosis: 2017 revisions of the McDonald criteria","volume":"17","author":"AJ Thompson","year":"2018","journal-title":"Lancet Neurol"},{"issue":"2","key":"pdig.0001374.ref016","doi-asserted-by":"crossref","first-page":"288","DOI":"10.1002\/ana.24564","article-title":"Primary progressive multiple sclerosis evolving from radiologically isolated syndrome","volume":"79","author":"OH Kantarci","year":"2016","journal-title":"Ann Neurol"},{"issue":"3","key":"pdig.0001374.ref017","doi-asserted-by":"crossref","first-page":"604","DOI":"10.1002\/ana.26555","article-title":"Dimethyl fumarate delays multiple sclerosis in radiologically isolated syndrome","volume":"93","author":"DT Okuda","year":"2023","journal-title":"Ann Neurol"},{"issue":"10","key":"pdig.0001374.ref018","doi-asserted-by":"crossref","first-page":"1080","DOI":"10.1001\/jamaneurol.2023.2815","article-title":"Teriflunomide and time to clinical multiple sclerosis in patients with radiologically isolated syndrome: the TERIS randomized clinical trial","volume":"80","author":"C Lebrun-Fr\u00e9nay","year":"2023","journal-title":"JAMA Neurol"},{"issue":"8","key":"pdig.0001374.ref019","doi-asserted-by":"crossref","first-page":"686","DOI":"10.1212\/WNL.0b013e31820d8b1d","article-title":"Asymptomatic spinal cord lesions predict disease progression in radiologically isolated syndrome","volume":"76","author":"DT Okuda","year":"2011","journal-title":"Neurology"},{"issue":"3","key":"pdig.0001374.ref020","doi-asserted-by":"crossref","DOI":"10.1212\/NXI.0000000000000102","article-title":"Early CNS neurodegeneration in radiologically isolated syndrome","volume":"2","author":"CJ Azevedo","year":"2015","journal-title":"Neurol Neuroimmunol Neuroinflamm"},{"issue":"1","key":"pdig.0001374.ref021","doi-asserted-by":"crossref","first-page":"130","DOI":"10.1177\/1352458519887346","article-title":"Cerebellar volume loss in radiologically isolated syndrome","volume":"27","author":"IC George","year":"2021","journal-title":"Mult Scler"},{"issue":"10","key":"pdig.0001374.ref022","doi-asserted-by":"crossref","first-page":"850","DOI":"10.1016\/S1474-4422(25)00270-4","article-title":"Diagnosis of multiple sclerosis: 2024 revisions of the McDonald criteria","volume":"24","author":"X Montalban","year":"2025","journal-title":"Lancet Neurol"},{"issue":"10","key":"pdig.0001374.ref023","doi-asserted-by":"crossref","first-page":"1415","DOI":"10.1038\/s41380-019-0441-1","article-title":"Conceptualizing mental disorders as deviations from normative functioning","volume":"24","author":"AF Marquand","year":"2019","journal-title":"Mol Psychiatry"},{"key":"pdig.0001374.ref024","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-031-38747-0","volume-title":"An introduction to statistical learning with applications in python","author":"G James","year":"2023"},{"key":"pdig.0001374.ref025","article-title":"Risk of bias in studies on prediction models developed using supervised machine learning techniques: systematic review","volume":"375","author":"CL Andaur Navarro","year":"2021","journal-title":"BMJ"},{"key":"pdig.0001374.ref026","doi-asserted-by":"crossref","DOI":"10.2196\/53542","article-title":"Digital representation of patients as medical digital twins: data-centric viewpoint","volume":"13","author":"S Demuth","year":"2025","journal-title":"JMIR Med Inform"},{"issue":"1","key":"pdig.0001374.ref027","doi-asserted-by":"crossref","first-page":"16279","DOI":"10.1038\/s41598-023-43618-5","article-title":"Toward precision medicine using a \u201cdigital twin\u201d approach: modeling the onset of disease-specific brain atrophy in individuals with multiple sclerosis","volume":"13","author":"S Cen","year":"2023","journal-title":"Sci Rep"},{"issue":"6","key":"pdig.0001374.ref028","doi-asserted-by":"crossref","DOI":"10.1212\/NXI.0000000000000616","article-title":"Contribution of normal aging to brain atrophy in MS","volume":"6","author":"CJ Azevedo","year":"2019","journal-title":"Neurol Neuroimmunol Neuroinflamm"},{"issue":"1","key":"pdig.0001374.ref029","doi-asserted-by":"crossref","first-page":"2078","DOI":"10.1038\/s41467-021-22265-2","article-title":"Identifying multiple sclerosis subtypes using unsupervised machine learning and MRI data","volume":"12","author":"A Eshaghi","year":"2021","journal-title":"Nat Commun"},{"issue":"10","key":"pdig.0001374.ref030","doi-asserted-by":"crossref","first-page":"1794","DOI":"10.2337\/dc24-0860","article-title":"Diabetes, prediabetes, and brain aging: the role of healthy lifestyle","volume":"47","author":"A Dove","year":"2024","journal-title":"Diabetes Care"},{"issue":"1","key":"pdig.0001374.ref031","doi-asserted-by":"crossref","first-page":"34585","DOI":"10.1038\/s41598-025-18073-z","article-title":"Estimation of reference curves for brain atrophy and analysis of robustness to machine effects","volume":"15","author":"Alzheimer\u2019s Disease Neuroimaging Initiative, Frontotemporal Lobar Degeneration Neuroimaging Initiative","year":"2025","journal-title":"Sci Rep"},{"issue":"1","key":"pdig.0001374.ref032","doi-asserted-by":"crossref","first-page":"29843","DOI":"10.1038\/s41598-025-15283-3","article-title":"Scan-rescan reliability assessment of brain volumetric analysis across scanners and software solutions","volume":"15","author":"E B\u00fcrkle","year":"2025","journal-title":"Sci Rep"},{"issue":"1","key":"pdig.0001374.ref033","doi-asserted-by":"crossref","first-page":"719","DOI":"10.1038\/s42003-022-03628-x","article-title":"Towards a comprehensive evaluation of dimension reduction methods for transcriptomic data visualization","volume":"5","author":"H Huang","year":"2022","journal-title":"Commun Biol"}],"container-title":["PLOS Digital Health"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pdig.0001374","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,24]],"date-time":"2026-04-24T17:45:28Z","timestamp":1777052728000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pdig.0001374"}},"subtitle":[],"editor":[{"given":"Liam G","family":"McCoy","sequence":"first","affiliation":[],"role":[{"role":"editor","vocabulary":"crossref"}]}],"short-title":[],"issued":{"date-parts":[[2026,4,24]]},"references-count":33,"journal-issue":{"issue":"4","published-online":{"date-parts":[[2026,4,24]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pdig.0001374","relation":{},"ISSN":["2767-3170"],"issn-type":[{"value":"2767-3170","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,4,24]]}}}