{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,7,30]],"date-time":"2026-07-30T13:59:00Z","timestamp":1785419940062,"version":"3.56.0"},"posted":{"date-parts":[[2026,7,17]]},"group-title":"Neurology","reference-count":36,"publisher":"openRxiv","license":[{"start":{"date-parts":[[2026,7,17]],"date-time":"2026-07-17T00:00:00Z","timestamp":1784246400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/publicdomain\/zero\/1.0\/"}],"funder":[{"name":"Intramural Research Program of the National Institute of Neurological Disorders and Stroke","award":["Z01 NS003119"],"award-info":[{"award-number":["Z01 NS003119"]}]},{"name":"NIHR Oxford Health Biomedical Research Centre","award":["NIHR203316"],"award-info":[{"award-number":["NIHR203316"]}]},{"award":["FG-2208-40289"],"award-info":[{"award-number":["FG-2208-40289"]}],"id":[{"id":"https:\/\/ror.org\/03019wv17","id-type":"ROR","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2026,7,17]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:p>White matter lesion (WML) identification, assessment, and characterization using magnetic resonance imaging (MRI) are fundamental for diagnosis and monitoring of multiple sclerosis (MS). Portable ultra-low field (pULF) MRI at 64 millitesla (mT) has been shown to visualize WML with at least one dimension greater than 4 mm. An automated WML segmentation tool catered to pULF-MRI can provide standardized and accurate quantitative measurements of WML volume. In this study, we sought to investigate and compare the accuracy of machine-learning (ML) and deep-learning (DL) pULF MRI segmentation tools. Same-day paired pULF (64mT) and high-field (HF, 3T) MRI scans from 84 adults with MS or suspected-MS (mean age\u00b1SD: 48\u00b113, 62 females) included T2-FLAIR and T1w images. Reference WML segmentations were manually annotated on pULF T2-FLAIR for all scans, with WML confirmed with registered HF T2-FLAIR. HF reference WML segmentations were created. Four automated segmentation methods were applied to pULF scans: Method for Inter-Modal Segmentation Analysis (MIMoSA), an ML algorithm trained on HF WML masks; WMH-SynthSeg, a convolutional neural network model with flexible segmentation capabilities across field strengths and resolution; nnU-Net, a DL algorithm trained on pULF reference WML masks; and Pseudo-Label Assisted nnU-Net (PLAn), a DL algorithm pre-trained on HF reference WML masks and refined with 64mT reference WML masks. Two models were trained with nnU-Net, one using T2-FLAIR images only (nnU-Net-FL) and one using T1w and T2-FLAIR images (nnU-Net-FL\/T1). The same was done with PLAn, creating PLAn-FL and PLAn-FL\/T1. The six automated WML segmentation outputs were compared to the manual segmentations to determine Dice Similarity Coefficient (DSC) scores.<\/jats:p>\n                <jats:p>Associations of WML volume estimates with clinical measures were investigated.<\/jats:p>\n                <jats:p>DSC scores with pULF reference WML masks from PLAn-FL (DSC mean\u00b1SD: 0.50\u00b10.24) outperformed MIMoSA (0.24\u00b10.20, p&lt;0.0001), WMH-SynthSeg (0.30\u00b10.18, p&lt;0.0001), nnU-Net-FL (0.41\u00b10.24, p&lt;0.0001), and nnU-Net-FL\/T1 (0.41 \u00b1 0.26, p = 0.0004). Worse Expanded Disability Status Scale (EDSS) and Scripps Neurologic Rating Scale (SNRS) scores were correlated with higher WML volumes in the pULF and HF reference masks. They were also correlated with WML volumes derived from WHM-SynthSeg, nnU-Net-FL, nnU-Net-FL\/T1, PLAn-FL, and PLAn-FL\/T1, but not MIMoSA. After adjusting for age, WHM-SynthSeg, nnU-Net FL, nnU-Net-FL\/T1, PLAn-FL, and PLAn-FL\/T1 had significant associations with EDSS and SNRS scores.<\/jats:p>\n                <jats:p>nnU-Net and PLAn performed best in segmenting WML on pULF-MRI at 64 mT, providing accurate quantitative estimates of WML burden. Moreover, WML volumes estimated by these algorithms were associated with clinical measures of disability, underscoring their utility for reflecting clinical and radiological disease severity. Given pULF-MRI\u2019s mobility and lower cost, these findings highlight its relevance in clinical trials, particularly in involving more participants who face logistical constraints and barriers.<\/jats:p>\n                <jats:sec>\n                  <jats:title>Highlights:<\/jats:title>\n                  <jats:list list-type=\"bullet\">\n                    <jats:list-item>\n                      <jats:p>Deep learning algorithms were trained on 20 paired, same-day pULF and HF MRI<\/jats:p>\n                    <\/jats:list-item>\n                    <jats:list-item>\n                      <jats:p>Automated pULF segmentation tools were evaluated using 84 paired, same day MRI<\/jats:p>\n                    <\/jats:list-item>\n                    <jats:list-item>\n                      <jats:p>Automated pULF WML segmentations align with reference pULF and HF segmentations<\/jats:p>\n                    <\/jats:list-item>\n                    <jats:list-item>\n                      <jats:p>Trained deep learning algorithms demonstrate strong pULF MRI lesion segmentation<\/jats:p>\n                    <\/jats:list-item>\n                    <jats:list-item>\n                      <jats:p>Automated pULF WML estimates are associated with clinical disability scores<\/jats:p>\n                    <\/jats:list-item>\n                  <\/jats:list>\n                <\/jats:sec>","DOI":"10.64898\/2026.07.15.26357954","type":"posted-content","created":{"date-parts":[[2026,7,17]],"date-time":"2026-07-17T17:35:12Z","timestamp":1784309712000},"source":"Crossref","is-referenced-by-count":0,"title":["Portable Ultra-Low Field MRI Deep-Learning Algorithms for White Matter Lesion Segmentation Improve Accuracy and Reflect Clinical Disability in Multiple Sclerosis"],"prefix":"10.64898","author":[{"ORCID":"https:\/\/orcid.org\/0009-0004-1117-8987","authenticated-orcid":false,"given":"Ashley A.","family":"Thommana","sequence":"first","affiliation":[{"name":"Neuroimmunology Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD, USA"},{"name":"Sidney Kimmel Medical College, Thomas Jefferson University, Philadelphia, PA, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-4856-8422","authenticated-orcid":false,"given":"Corinne A.","family":"Donnay","sequence":"additional","affiliation":[{"name":"Neuroimmunology Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD, USA"},{"name":"Oxford Centre for Integrative Neuroimaging - University of Oxford, Nuffield Department of Clinical Neurosciences, University of Oxford, Oxford, United Kingdom"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1324-0217","authenticated-orcid":false,"given":"Gina","family":"Norato","sequence":"additional","affiliation":[{"name":"Office of Biostatistics, Division of Intramural Research, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4455-5641","authenticated-orcid":false,"given":"Mar\u00eda I.","family":"Gait\u00e1n","sequence":"additional","affiliation":[{"name":"Neuroimmunology Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0540-9353","authenticated-orcid":false,"given":"Ludovica","family":"Griffanti","sequence":"additional","affiliation":[{"name":"Oxford Centre for Integrative Neuroimaging, Oxford Centre for Human Brain Activity - University of Oxford, Department of Psychiatry, University of Oxford, Oxford, United Kingdom"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3725-615X","authenticated-orcid":false,"given":"Govind","family":"Nair","sequence":"additional","affiliation":[{"name":"qMRI Core Facility, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2628-4334","authenticated-orcid":false,"given":"Daniel S.","family":"Reich","sequence":"additional","affiliation":[{"name":"Neuroimmunology Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3716-2196","authenticated-orcid":false,"given":"Serhat V.","family":"Okar","sequence":"additional","affiliation":[{"name":"Neuroimmunology Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"54368","reference":[{"key":"2026072014201676000_2026.07.15.26357954v1.1","doi-asserted-by":"publisher","DOI":"10.1002\/ana.1032"},{"key":"2026072014201676000_2026.07.15.26357954v1.2","doi-asserted-by":"publisher","DOI":"10.1093\/brain\/awz144"},{"key":"2026072014201676000_2026.07.15.26357954v1.3","doi-asserted-by":"publisher","DOI":"10.1016\/S1474-4422(22)00289-7"},{"key":"2026072014201676000_2026.07.15.26357954v1.4","doi-asserted-by":"publisher","DOI":"10.1016\/S1474-4422(13)70103-0"},{"key":"2026072014201676000_2026.07.15.26357954v1.5","doi-asserted-by":"crossref","first-page":"103101","DOI":"10.1016\/j.nicl.2022.103101","article-title":"Sensitivity of portable low-field magnetic resonance imaging for multiple sclerosis lesions","volume":"35","year":"2022","journal-title":"Neuroimage Clin"},{"key":"2026072014201676000_2026.07.15.26357954v1.6","doi-asserted-by":"publisher","DOI":"10.1097\/RLI.0000000000000810"},{"key":"2026072014201676000_2026.07.15.26357954v1.7","doi-asserted-by":"publisher","DOI":"10.1259\/bjr.20210708"},{"key":"2026072014201676000_2026.07.15.26357954v1.8","doi-asserted-by":"publisher","DOI":"10.1097\/WCO.0000000000001282"},{"key":"2026072014201676000_2026.07.15.26357954v1.9","doi-asserted-by":"crossref","first-page":"1067","DOI":"10.1212\/WNL.0000000000207074","article-title":"Future of Neurology & Technology: Neuroimaging Made Accessible Using Low-Field, Portable MRI","volume":"100","year":"2023","journal-title":"Neurology"},{"key":"2026072014201676000_2026.07.15.26357954v1.10","doi-asserted-by":"crossref","first-page":"116","DOI":"10.1016\/j.ejmp.2021.04.003","article-title":"Characterization of displacement forces and image artifacts in the presence of passive medical implants in low-field (<100 mT) permanent magnet-based MRI systems, and comparisons with clinical MRI systems","volume":"84","year":"2021","journal-title":"Phys Med"},{"key":"2026072014201676000_2026.07.15.26357954v1.11","doi-asserted-by":"crossref","unstructured":"Okar SV , Nair G , Kawatra KD , et al. High-Field-Blinded Assessment of Portable Ultra-Low- Field Brain MRI for Multiple Sclerosis. Journal of Neuroimaging 2025;35.","DOI":"10.1111\/jon.70005"},{"key":"2026072014201676000_2026.07.15.26357954v1.12","doi-asserted-by":"crossref","first-page":"102903","DOI":"10.1016\/j.msard.2021.102903","article-title":"Low-field portable brain MRI in CNS demyelinating disease","volume":"51","year":"2021","journal-title":"Mult Scler Relat Disord"},{"key":"2026072014201676000_2026.07.15.26357954v1.13","doi-asserted-by":"crossref","first-page":"e01226","DOI":"10.1016\/j.heliyon.2019.e01226","article-title":"Robust, atlas-free, automatic segmentation of brain MRI in health and disease","volume":"5","year":"2019","journal-title":"Heliyon"},{"key":"2026072014201676000_2026.07.15.26357954v1.14","doi-asserted-by":"publisher","DOI":"10.1111\/jon.12506"},{"key":"2026072014201676000_2026.07.15.26357954v1.15","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuroimage.2016.07.018"},{"key":"2026072014201676000_2026.07.15.26357954v1.16","doi-asserted-by":"crossref","unstructured":"Iglesias JE , Billot B , Balbastre Y , et al. SynthSR: A public AI tool to turn heterogeneous clinical brain scans into high-resolution T1-weighted images for 3D morphometry. Sci Adv 2023;9:eadd3607.","DOI":"10.1126\/sciadv.add3607"},{"key":"2026072014201676000_2026.07.15.26357954v1.17","doi-asserted-by":"publisher","DOI":"10.1148\/radiol.220522"},{"key":"2026072014201676000_2026.07.15.26357954v1.18","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuroimage.2021.118206"},{"key":"2026072014201676000_2026.07.15.26357954v1.19","doi-asserted-by":"crossref","unstructured":"Laso P , Cerri S , Sorby-Adams A , et al. Quantifying white matter hyperintensity and brain volumes in heterogeneous clinical and low-field portable MRI. 2024 IEEE International Symposium on Biomedical Imaging (ISBI); 2024: IEEE: 1\u20135.","DOI":"10.1109\/ISBI56570.2024.10635502"},{"key":"2026072014201676000_2026.07.15.26357954v1.20","doi-asserted-by":"publisher","DOI":"10.1038\/s41592-020-01008-z"},{"key":"2026072014201676000_2026.07.15.26357954v1.21","doi-asserted-by":"publisher","DOI":"10.3389\/fnins.2022.1007453"},{"key":"2026072014201676000_2026.07.15.26357954v1.22","doi-asserted-by":"crossref","unstructured":"Zhou W , Li J , Wang X , Wang Y , Lyu M. Infant Hippocampal Segmentation in Ultra-Low-Field MRI Using External Datasets with Diverse Field Strengths. 2025; Cham: Springer Nature Switzerland: 28\u201337.","DOI":"10.1007\/978-3-031-83008-2_3"},{"key":"2026072014201676000_2026.07.15.26357954v1.23","doi-asserted-by":"crossref","first-page":"1252261","DOI":"10.3389\/fnimg.2023.1252261","article-title":"Pseudo-Label Assisted nnU-Net enables automatic segmentation of 7T MRI from a single acquisition","volume":"2","year":"2023","journal-title":"Front Neuroimaging"},{"key":"2026072014201676000_2026.07.15.26357954v1.24","doi-asserted-by":"publisher","DOI":"10.1016\/S1474-4422(17)30470-2"},{"key":"2026072014201676000_2026.07.15.26357954v1.25","doi-asserted-by":"publisher","DOI":"10.1212\/01.wnl.0000335764.14513.1a"},{"key":"2026072014201676000_2026.07.15.26357954v1.26","doi-asserted-by":"publisher","DOI":"10.1212\/WNL.33.11.1444"},{"key":"2026072014201676000_2026.07.15.26357954v1.27","doi-asserted-by":"publisher","DOI":"10.1212\/WNL.34.10.1368"},{"key":"2026072014201676000_2026.07.15.26357954v1.28","doi-asserted-by":"publisher","DOI":"10.1177\/1352458517690824"},{"key":"2026072014201676000_2026.07.15.26357954v1.29","doi-asserted-by":"publisher","DOI":"10.1177\/1352458506070624"},{"key":"2026072014201676000_2026.07.15.26357954v1.30","doi-asserted-by":"publisher","DOI":"10.1177\/1352458518808204"},{"key":"2026072014201676000_2026.07.15.26357954v1.31","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuroimage.2006.01.015"},{"key":"2026072014201676000_2026.07.15.26357954v1.32","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuroimage.2010.09.025"},{"key":"2026072014201676000_2026.07.15.26357954v1.33","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuroimage.2005.05.046"},{"key":"2026072014201676000_2026.07.15.26357954v1.34","doi-asserted-by":"crossref","first-page":"462","DOI":"10.1016\/j.nicl.2013.10.003","article-title":"Accurate white matter lesion segmentation by k nearest neighbor classification with tissue type priors (kNN-TTPs)","volume":"3","year":"2013","journal-title":"Neuroimage Clin"},{"key":"2026072014201676000_2026.07.15.26357954v1.35","doi-asserted-by":"crossref","unstructured":"Lavery AM , Verhey LH , Waldman AT . Outcome measures in relapsing-remitting multiple sclerosis: capturing disability and disease progression in clinical trials. Mult Scler Int 2014;2014:262350.","DOI":"10.1155\/2014\/262350"},{"key":"2026072014201676000_2026.07.15.26357954v1.36","doi-asserted-by":"publisher","DOI":"10.1212\/01.wnl.0000173061.12776.1f"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.64898\/2026.07.15.26357954","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,20]],"date-time":"2026-07-20T21:20:25Z","timestamp":1784582425000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.64898\/2026.07.15.26357954"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,17]]},"references-count":36,"URL":"https:\/\/doi.org\/10.64898\/2026.07.15.26357954","relation":{},"subject":[],"published":{"date-parts":[[2026,7,17]]},"subtype":"preprint"}}