{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,14]],"date-time":"2026-10-14T08:15:44Z","timestamp":1791965744478,"version":"4.3.4"},"reference-count":36,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2022,6,14]],"date-time":"2022-06-14T00:00:00Z","timestamp":1655164800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Immunol."],"abstract":"<jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Differential diagnosis of demyelinating diseases of the central nervous system is a challenging task that is prone to errors and inconsistent reading, requiring expertise and additional examination approaches. Advancements in deep-learning-based image interpretations allow for prompt and automated analyses of conventional magnetic resonance imaging (MRI), which can be utilized in classifying multi-sequence MRI, and thus may help in subsequent treatment referral.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>Imaging and clinical data from 290 patients diagnosed with demyelinating diseases from August 2013 to October 2021 were included for analysis, including 67 patients with multiple sclerosis (MS), 162 patients with aquaporin 4 antibody-positive (AQP4+) neuromyelitis optica spectrum disorder (NMOSD), and 61 patients with myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD). Considering the heterogeneous nature of lesion size and distribution in demyelinating diseases, multi-modal MRI of brain and\/or spinal cord were utilized to build the deep-learning model. This novel transformer-based deep-learning model architecture was designed to be versatile in handling with multiple image sequences (coronal T2-weighted and sagittal T2-fluid attenuation inversion recovery) and scanning locations (brain and spinal cord) for differentiating among MS, NMOSD, and MOGAD. Model performances were evaluated using the area under the receiver operating curve (AUC) and the confusion matrices measurements. The classification accuracy between the fusion model and the neuroradiological raters was also compared.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>The fusion model that was trained with combined brain and spinal cord MRI achieved an overall improved performance, with the AUC of 0.933 (95%CI: 0.848, 0.991), 0.942 (95%CI: 0.879, 0.987) and 0.803 (95%CI: 0.629, 0.949) for MS, AQP4+ NMOSD, and MOGAD, respectively. This exceeded the performance using the brain or spinal cord MRI alone for the identification of the AQP4+ NMOSD (AUC of 0.940, brain only and 0.689, spinal cord only) and MOGAD (0.782, brain only and 0.714, spinal cord only). In the multi-category classification, the fusion model had an accuracy of 81.4%, which was significantly higher compared to rater 1 (64.4%, p=0.04&amp;lt;0.05) and comparable to rater 2 (74.6%, p=0.388).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>The proposed novel transformer-based model showed desirable performance in the differentiation of MS, AQP4+ NMOSD, and MOGAD on brain and spinal cord MRI, which is comparable to that of neuroradiologists. Our model is thus applicable for interpretating conventional MRI in the differential diagnosis of demyelinating diseases with overlapping lesions.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fimmu.2022.897959","type":"journal-article","created":{"date-parts":[[2022,6,14]],"date-time":"2022-06-14T11:26:12Z","timestamp":1655205972000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":24,"title":["Transformer-Based Deep-Learning Algorithm for Discriminating Demyelinating Diseases of the Central Nervous System With Neuroimaging"],"prefix":"10.3389","volume":"13","author":[{"given":"Chuxin","family":"Huang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Weidao","family":"Chen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Baiyun","family":"Liu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ruize","family":"Yu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiqian","family":"Chen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fei","family":"Tang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jun","family":"Liu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Wei","family":"Lu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2022,6,14]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"59","DOI":"10.1212\/WNL.0000000000011153","article-title":"Worldwide Incidence and Prevalence of Neuromyelitis Optica: A Systematic Review","volume":"96","author":"Papp","year":"2021","journal-title":"Neurology"},{"key":"B2","doi-asserted-by":"publisher","DOI":"10.1177\/1352458520970841","article-title":"Rising Prevalence of Multiple Sclerosis Worldwide: Insights From the Atlas of MS, Third Edition","volume":"26","author":"Walton","year":"2020","journal-title":"Mult Scler"},{"key":"B3","doi-asserted-by":"publisher","DOI":"10.1093\/brain\/awf151","article-title":"A Role for Humoral Mechanisms in the Pathogenesis of Devic's Neuromyelitis Optica","volume":"125","author":"Lucchinetti","year":"2002","journal-title":"Brain"},{"key":"B4","doi-asserted-by":"publisher","DOI":"10.1177\/1352458519868796","article-title":"MOG-Antibody-Associated Disease is Different From MS and NMOSD and Should be Considered as a Distinct Disease Entity - Yes","volume":"26","author":"Leite","year":"2020","journal-title":"Mult Scler"},{"key":"B5","doi-asserted-by":"publisher","DOI":"10.1186\/1742-2094-10-8","article-title":"The History of Neuromyelitis Optica","volume":"10","author":"Jarius","year":"2013","journal-title":"J Neuroinflammat"},{"key":"B6","doi-asserted-by":"publisher","DOI":"10.1016\/S1474-4422(07)70216-8","article-title":"The Spectrum of Neuromyelitis Optica","volume":"6","author":"Wingerchuk","year":"2007","journal-title":"Lancet Neurol"},{"key":"B7","doi-asserted-by":"publisher","DOI":"10.1097\/WCO.0000000000000828","article-title":"Neuromyelitis Optica Spectrum Disorders and Myelin Oligodendrocyte Glycoprotein Antibody-Associated Disease: Current Topics","volume":"33","author":"Fujihara","year":"2020","journal-title":"Curr Opin Neurol"},{"key":"B8","doi-asserted-by":"publisher","DOI":"10.1097\/WCO.0000000000000694","article-title":"Neuromyelitis Optica Spectrum Disorders: Still Evolving and Broadening","volume":"32","author":"Fujihara","year":"2019","journal-title":"Curr Opin Neurol"},{"key":"B9","doi-asserted-by":"publisher","DOI":"10.1016\/S1474-4422(17)30470-2","article-title":"Diagnosis of Multiple Sclerosis: 2017 Revisions of the McDonald Criteria","volume":"17","author":"Thompson","year":"2018","journal-title":"Lancet Neurol"},{"key":"B10","doi-asserted-by":"publisher","DOI":"10.1212\/WNL.0000000000001729","article-title":"International Consensus Diagnostic Criteria for Neuromyelitis Optica Spectrum Disorders","volume":"85","author":"Wingerchuk","year":"2015","journal-title":"Neurology"},{"key":"B11","doi-asserted-by":"publisher","DOI":"10.1007\/s00415-021-10727-y","article-title":"Application of Deep-Learning to the Seronegative Side of the NMO Spectrum","volume":"269","author":"Cacciaguerra","year":"2021","journal-title":"J Neurol"},{"key":"B12","doi-asserted-by":"publisher","DOI":"10.3389\/fneur.2018.00888","article-title":"Diagnosis and Treatment of NMO Spectrum Disorder and MOG-Encephalomyelitis","volume":"9","author":"Borisow","year":"2018","journal-title":"Front Neurol"},{"key":"B13","doi-asserted-by":"publisher","DOI":"10.1001\/archneurol.2010.188","article-title":"Interferon Beta Treatment in Neuromyelitis Optica: Increase in Relapses and Aquaporin 4 Antibody Titers","volume":"67","author":"Palace","year":"2010","journal-title":"Arch Neurol"},{"key":"B14","doi-asserted-by":"publisher","DOI":"10.1016\/j.msard.2018.09.011","article-title":"Severe Disease Exacerbation in a Patient With Neuromyelitis Optica Spectrum Disorder During Treatment With Dimethyl Fumarate","volume":"26","author":"Popiel","year":"2018","journal-title":"Mult Scler Relat Disord"},{"key":"B15","doi-asserted-by":"publisher","DOI":"10.1002\/ana.25648","article-title":"Comparison of the Response to Rituximab Between Myelin Oligodendrocyte Glycoprotein and Aquaporin-4 Antibody Diseases","volume":"87","author":"Durozard","year":"2020","journal-title":"Ann Neurol"},{"key":"B16","doi-asserted-by":"publisher","DOI":"10.1136\/jnnp-2020-324826","article-title":"Brain Structural Alterations in MOG Antibody Diseases: A Comparative Study With AQP4 Seropositive NMOSD and MS","volume":"92","author":"Duan","year":"2021","journal-title":"J Neurol Neurosurg Psychiatry"},{"key":"B17","doi-asserted-by":"publisher","DOI":"10.1136\/jnnp-2020-325121","article-title":"Brainstem and Cerebellar Involvement in MOG-IgG-Associated Disorder Versus Aquaporin-4-IgG and MS","author":"Banks","year":"2020","journal-title":"J Neurol Neurosurg Psychiatry"},{"key":"B18","doi-asserted-by":"publisher","DOI":"10.3389\/fneur.2020.599042","article-title":"Deep Learning-Based Method to Differentiate Neuromyelitis Optica Spectrum Disorder From Multiple Sclerosis","volume":"11","author":"Kim","year":"2020","journal-title":"Front Neurol"},{"key":"B19","doi-asserted-by":"publisher","DOI":"10.3389\/fphys.2020.612928","article-title":"3d Compressed Convolutional Neural Network Differentiates Neuromyelitis Optical Spectrum Disorders From Multiple Sclerosis Using Automated White Matter Hyperintensities Segmentations","volume":"11","author":"Wang","year":"2020","journal-title":"Front Physiol"},{"key":"B20","doi-asserted-by":"publisher","DOI":"10.1097\/RLI.0000000000000735","article-title":"Deep Learning on Conventional Magnetic Resonance Imaging Improves the Diagnosis of Multiple Sclerosis Mimics","volume":"56","author":"Rocca","year":"2021","journal-title":"Invest Radiol"},{"key":"B21","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2104.06399","article-title":"Co-Scale Conv-Attentional Image Transformers","author":"Xu","year":"2021","journal-title":"ArXiv E-print"},{"key":"B22","doi-asserted-by":"publisher","first-page":"134","DOI":"10.1186\/s12974-018-1144-2","article-title":"MOG Encephalomyelitis: International Recommendations on Diagnosis and Antibody Testing","volume":"15","author":"Jarius","year":"2018","journal-title":"J Neuroinflammat"},{"key":"B23","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1802.04712","article-title":"Attention-Based Deep Multiple Instance Learning","volume":"80","author":"Ilse","year":"2018","journal-title":"Proc 35 Th Int Conf Mach Learn"},{"key":"B24","doi-asserted-by":"publisher","DOI":"10.1016\/j.compbiomed.2017.08.001","article-title":"Pre-Trained Convolutional Neural Networks as Feature Extractors for Tuberculosis Detection","volume":"89","author":"Lopes","year":"2017","journal-title":"Comput Biol Med"},{"key":"B25","doi-asserted-by":"publisher","first-page":"145011","DOI":"10.1088\/1361-6560\/ab857d","article-title":"DCT-MIL: Deep CNN Transferred Multiple Instance Learning for COPD Identification Using CT Images","volume":"65","author":"Xu","year":"2020","journal-title":"Phys Med Biol"},{"key":"B26","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1710.09412","article-title":"Mixup: Beyond Empirical Risk Minimization","author":"Zhang","year":"2018","journal-title":"ArXiv E-print"},{"key":"B27","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1905.04899","article-title":"CutMix: Regularization Strategy to Train Strong Classifiers With Localizable Features","author":"Yun","year":"2019","journal-title":"Proc IEEE\/CVF Int Conf Comput Vision (ICCV), ArXiv E-print"},{"key":"B28","doi-asserted-by":"publisher","DOI":"10.1007\/s11548-019-02068-z","article-title":"Memory-Efficient 2.5D Convolutional Transformer Networks for Multi-Modal Deformable Registration With Weak Label Supervision Applied to Whole-Heart CT and MRI Scans","volume":"14","author":"Hering","year":"2019","journal-title":"Int J Comput Assist Radiol Surg"},{"key":"B29","doi-asserted-by":"crossref","first-page":"13050","DOI":"10.1007\/978-3-030-89847-2_2","article-title":"Multi-Scale Hybrid Transformer Networks: Application to Prostate Disease Classification","volume-title":"International Workshop on Multimodal Learning for Clinical Decision Support Lecture Notes in Computer Science","author":"Santhirasekaram","year":"2021"},{"key":"B30","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-015-0816-y","article-title":"ImageNet Large Scale Visual Recognition Challenge","volume":"115","author":"Russakovsky","year":"2015","journal-title":"Int J Comput Vision"},{"key":"B31","doi-asserted-by":"publisher","DOI":"10.3389\/fonc.2020.601869","article-title":"Deep Convolutional Neural Network Based on Computed Tomography Images for the Preoperative Diagnosis of Occult Peritoneal Metastasis in Advanced Gastric Cancer","volume":"10","author":"Huang","year":"2020","journal-title":"Front Oncol"},{"key":"B32","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-019-01228-7","article-title":"Grad-CAM: Visual Explanations From Deep Networks via Gradient-Based Localization","volume":"128","author":"Selvaraju","year":"2020","journal-title":"Int J Comput Vision"},{"key":"B33","doi-asserted-by":"publisher","DOI":"10.1007\/s00330-019-06026-w","article-title":"Radiomics in Multiple Sclerosis and Neuromyelitis Optica Spectrum Disorder","volume":"29","author":"Liu","year":"2019","journal-title":"Eur Radiol"},{"key":"B34","doi-asserted-by":"publisher","DOI":"10.3174\/ajnr.A5543","article-title":"Deep Learning in Neuroradiology","volume":"39","author":"Zaharchuk","year":"2018","journal-title":"AJNR Am J Neuroradiol"},{"key":"B35","doi-asserted-by":"publisher","DOI":"10.1001\/jamanetworkopen.2020.5842","article-title":"Development and Validation of a Deep Learning Model for Non-Small Cell Lung Cancer Survival","volume":"3","author":"She","year":"2020","journal-title":"JAMA Netw Open"},{"key":"B36","doi-asserted-by":"publisher","DOI":"10.1093\/bib\/bbaa310","article-title":"Deep Learning for Brain Disorders: From Data Processing to Disease Treatment","volume":"22","author":"Burgos","year":"2021","journal-title":"Brief Bioinform"}],"container-title":["Frontiers in Immunology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fimmu.2022.897959\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,6,14]],"date-time":"2022-06-14T11:26:19Z","timestamp":1655205979000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fimmu.2022.897959\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,6,14]]},"references-count":36,"alternative-id":["10.3389\/fimmu.2022.897959"],"URL":"https:\/\/doi.org\/10.3389\/fimmu.2022.897959","relation":{},"ISSN":["1664-3224"],"issn-type":[{"value":"1664-3224","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,6,14]]},"article-number":"897959"}}