{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,4]],"date-time":"2026-09-04T01:39:22Z","timestamp":1788485962976,"version":"build-2803163510"},"reference-count":41,"publisher":"Wiley","issue":"8","license":[{"start":{"date-parts":[[2023,6,9]],"date-time":"2023-06-09T00:00:00Z","timestamp":1686268800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/"}],"content-domain":{"domain":["onlinelibrary.wiley.com"],"crossmark-restriction":true},"short-container-title":["Ann Clin Transl Neurol"],"published-print":{"date-parts":[[2023,8]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:sec><jats:title>Objective<\/jats:title><jats:p>Myasthenia gravis (MG) is an autoimmune disease leading to fatigable muscle weakness. Extra\u2010ocular and bulbar muscles are most commonly affected. We aimed to investigate whether facial weakness can be quantified automatically and used for diagnosis and disease monitoring.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>In this cross\u2010sectional study, we analyzed video recordings of 70 MG patients and 69 healthy controls (HC) with two different methods. Facial weakness was first quantified with facial expression recognition software. Subsequently, a deep learning (DL) computer model was trained for the classification of diagnosis and disease severity using multiple cross\u2010validations on videos of 50 patients and 50 controls. Results were validated using unseen videos of 20 MG patients and 19 HC.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>Expression of anger (<jats:italic>p<\/jats:italic>\u2009=\u20090.026), fear (<jats:italic>p<\/jats:italic>\u2009=\u20090.003), and happiness (<jats:italic>p<\/jats:italic>\u2009&lt;\u20090.001) was significantly decreased in MG compared to HC. Specific patterns of decreased facial movement were detectable in each emotion. Results of the DL model for diagnosis were as follows: area under the curve (AUC) of the receiver operator curve 0.75 (95% CI 0.65\u20130.85), sensitivity 0.76, specificity 0.76, and accuracy 76%. For disease severity: AUC 0.75 (95% CI 0.60\u20130.90), sensitivity 0.93, specificity 0.63, and accuracy 80%. Results of validation, diagnosis: AUC 0.82 (95% CI: 0.67\u20130.97), sensitivity 1.0, specificity 0.74, and accuracy 87%. For disease severity: AUC 0.88 (95% CI: 0.67\u20131.0), sensitivity 1.0, specificity 0.86, and accuracy 94%.<\/jats:p><\/jats:sec><jats:sec><jats:title>Interpretation<\/jats:title><jats:p>Patterns of facial weakness can be detected with facial recognition software. Second, this study delivers a \u2018proof of concept\u2019 for a DL model that can distinguish MG from HC and classifies disease severity.<\/jats:p><\/jats:sec>","DOI":"10.1002\/acn3.51823","type":"journal-article","created":{"date-parts":[[2023,6,9]],"date-time":"2023-06-09T07:19:49Z","timestamp":1686295189000},"page":"1314-1325","update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":22,"title":["Assessing facial weakness in myasthenia gravis with facial recognition software and deep learning"],"prefix":"10.1002","volume":"10","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-8608-5871","authenticated-orcid":false,"given":"Annabel M.","family":"Ruiter","sequence":"first","affiliation":[{"name":"Department of Neurology Leiden University Medical Center  Leiden the Netherlands"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ziqi","family":"Wang","sequence":"additional","affiliation":[{"name":"Vision Lab Delft University of Technology  Delft the Netherlands"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zhao","family":"Yin","sequence":"additional","affiliation":[{"name":"Vision Lab Delft University of Technology  Delft the Netherlands"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Willemijn C.","family":"Naber","sequence":"additional","affiliation":[{"name":"Department of Neurology Leiden University Medical Center  Leiden the Netherlands"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jerrel","family":"Simons","sequence":"additional","affiliation":[{"name":"Department of Neurology Leiden University Medical Center  Leiden the Netherlands"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jurre T.","family":"Blom","sequence":"additional","affiliation":[{"name":"Medical Illustrator at www.jurreblom.nl  Apeldoorn the Netherlands"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jan C.","family":"van Gemert","sequence":"additional","affiliation":[{"name":"Vision Lab Delft University of Technology  Delft the Netherlands"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jan J. G. M.","family":"Verschuuren","sequence":"additional","affiliation":[{"name":"Department of Neurology Leiden University Medical Center  Leiden the Netherlands"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Martijn R.","family":"Tannemaat","sequence":"additional","affiliation":[{"name":"Department of Neurology Leiden University Medical Center  Leiden the Netherlands"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2023,6,9]]},"reference":[{"key":"e_1_2_10_2_1","doi-asserted-by":"publisher","DOI":"10.3109\/08916931003602130"},{"key":"e_1_2_10_3_1","doi-asserted-by":"publisher","DOI":"10.1002\/mus.20030"},{"key":"e_1_2_10_4_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.nmd.2019.07.006"},{"key":"e_1_2_10_5_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41572-019-0079-y"},{"key":"e_1_2_10_6_1","doi-asserted-by":"publisher","DOI":"10.1212\/WNL.0000000000002383"},{"key":"e_1_2_10_7_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.jns.2007.04.045"},{"key":"e_1_2_10_8_1","volume-title":"Real\u2010World Assessment and Patient Perceptions of a Prototype Myasthenia Gravis Tracking App Oral Presentation MGFA Scientific Session","author":"Steels J","year":"2022"},{"key":"e_1_2_10_9_1","doi-asserted-by":"publisher","DOI":"10.1089\/tmj.2006.0078"},{"key":"e_1_2_10_10_1","doi-asserted-by":"publisher","DOI":"10.2196\/15429"},{"key":"e_1_2_10_11_1","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0082071"},{"key":"e_1_2_10_12_1","doi-asserted-by":"publisher","DOI":"10.1266\/ggs.15-00017"},{"key":"e_1_2_10_13_1","doi-asserted-by":"publisher","DOI":"10.1212\/WNL.0000000000009053"},{"key":"e_1_2_10_14_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ejrad.2020.109460"},{"key":"e_1_2_10_15_1","doi-asserted-by":"publisher","DOI":"10.3390\/molecules25173886"},{"key":"e_1_2_10_16_1","doi-asserted-by":"publisher","DOI":"10.1097\/PRS.0000000000007251"},{"key":"e_1_2_10_17_1","doi-asserted-by":"publisher","DOI":"10.1007\/s00405-019-05647-7"},{"key":"e_1_2_10_18_1","doi-asserted-by":"publisher","DOI":"10.1109\/TNSRE.2020.3021410"},{"key":"e_1_2_10_19_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijom.2021.01.002"},{"key":"e_1_2_10_20_1","doi-asserted-by":"publisher","DOI":"10.1097\/MAO.0000000000002735"},{"key":"e_1_2_10_21_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-017-16608-7"},{"key":"e_1_2_10_22_1","doi-asserted-by":"publisher","DOI":"10.1212\/WNL.0000000000009782"},{"key":"e_1_2_10_23_1","doi-asserted-by":"publisher","DOI":"10.1002\/mus.23988"},{"key":"e_1_2_10_24_1","unstructured":"https:\/\/www.noldus.com\/facereader\/facial\u2010expression\u2010analysis"},{"key":"e_1_2_10_25_1","volume-title":"Facial Action Coding System: Facial Action Coding System: The Manual: On CD\u2010ROM","author":"Ekman P","year":"2002"},{"key":"e_1_2_10_26_1","unstructured":"SoomroK ZamirAR ShahM.Ucf101: a dataset of 101 human actions classes from videos in the wild. arXiv preprint arXiv:121204022012."},{"key":"e_1_2_10_27_1","unstructured":"KayW CarreiraJ SimonyanK et\u00a0al.The kinetics human action video dataset. arXiv preprint arXiv:1705069502017."},{"key":"e_1_2_10_28_1","unstructured":"https:\/\/github.com\/gsig\/PyVideoResearch"},{"key":"e_1_2_10_29_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.502"},{"key":"e_1_2_10_30_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"e_1_2_10_31_1","unstructured":"AnandK WangZ LoogM vanGemertJ.Black magic in deep learning: how human skill impacts network training. arXiv preprint arXiv:2008059812020."},{"key":"e_1_2_10_32_1","first-page":"754","volume-title":"International Conference on Machine Learning","author":"Hutter F","year":"2014"},{"issue":"1","key":"e_1_2_10_33_1","first-page":"42","article-title":"Deep\u2010learning: investigating deep neural networks hyper\u2010parameters and comparison of performance to shallow methods for modeling bioactivity data","volume":"9","author":"Koutsoukas A","year":"2017","journal-title":"J Chem"},{"key":"e_1_2_10_34_1","doi-asserted-by":"publisher","DOI":"10.1097\/00129492-200411000-00026"},{"key":"e_1_2_10_35_1","doi-asserted-by":"publisher","DOI":"10.1080\/09638288.2022.2092779"},{"key":"e_1_2_10_36_1","doi-asserted-by":"publisher","DOI":"10.3233\/JND-190407"},{"key":"e_1_2_10_37_1","doi-asserted-by":"publisher","DOI":"10.1109\/JBHI.2020.2964520"},{"key":"e_1_2_10_38_1","doi-asserted-by":"publisher","DOI":"10.3390\/ijms21031042"},{"key":"e_1_2_10_39_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-018-0279-0"},{"key":"e_1_2_10_40_1","doi-asserted-by":"publisher","DOI":"10.1002\/ajmg.a.38343"},{"key":"e_1_2_10_41_1","unstructured":"HileyL PreeceA HicksY ChakrabortyS GurramP TomsettR.Explaining motion relevance for activity recognition in video deep learning models. arXiv preprint arXiv:2003142852020."},{"key":"e_1_2_10_42_1","first-page":"1","volume-title":"IEEE 3rd International Conference on Biometrics: Theory, Applications, and Systems","author":"Lui Y","year":"2009"}],"container-title":["Annals of Clinical and Translational Neurology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/acn3.51823","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,8,14]],"date-time":"2023-08-14T13:59:58Z","timestamp":1692021598000},"score":1,"resource":{"primary":{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/10.1002\/acn3.51823"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,6,9]]},"references-count":41,"journal-issue":{"issue":"8","published-print":{"date-parts":[[2023,8]]}},"alternative-id":["10.1002\/acn3.51823"],"URL":"https:\/\/doi.org\/10.1002\/acn3.51823","archive":["Portico"],"relation":{},"ISSN":["2328-9503","2328-9503"],"issn-type":[{"value":"2328-9503","type":"print"},{"value":"2328-9503","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,6,9]]},"assertion":[{"value":"2023-03-16","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2023-05-24","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2023-06-09","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}