{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,30]],"date-time":"2026-09-30T00:54:45Z","timestamp":1790729685558,"version":"4.1.0"},"reference-count":60,"publisher":"Public Library of Science (PLoS)","issue":"9","license":[{"start":{"date-parts":[[2025,9,16]],"date-time":"2025-09-16T00:00:00Z","timestamp":1757980800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/100009633","name":"Eunice Kennedy Shriver National Institute of Child Health and Human Development","doi-asserted-by":"publisher","award":["R21084231"],"award-info":[{"award-number":["R21084231"]}],"id":[{"id":"10.13039\/100009633","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100000055","name":"National Institute on Deafness and Other Communication Disorders","doi-asserted-by":"publisher","award":["R21DC021029-01A1"],"award-info":[{"award-number":["R21DC021029-01A1"]}],"id":[{"id":"10.13039\/100000055","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["www.plosdigitalhealth.org"],"crossmark-restriction":false},"short-container-title":["PLOS Digit Health"],"abstract":"<jats:p>Accurate diagnosis of gait impairments is often hindered by subjective or costly assessment methods, with current solutions relying on either expensive multi-camera equipment or subjective clinical observation. There is a critical need for accessible, objective tools that can aid in gait assessment while preserving patient privacy. In this work, we present a mobile phone-based, privacy-preserving artificial intelligence (AI) system for classifying gait impairments that leverages a novel dataset of 743 videos capturing seven distinct gait types. The dataset consists of frontal and sagittal views of clinicians simulating normal gait and six types of pathological gait (circumduction, Trendelenburg, antalgic, crouch, Parkinsonian, and vaulting), recorded using standard mobile phone cameras. Our system achieved 86.5% accuracy using combined frontal and sagittal views, with sagittal views generally outperforming frontal views except for specific gait types like circumduction. Model feature importance analysis revealed that frequency-domain features and entropy measures were critical for classification performance. Specifically, lower limb keypoints proved most important for classification, aligning with clinical understanding of gait assessment. These findings demonstrate that mobile phone-based systems can effectively classify diverse gait types while preserving privacy through on-device processing. The high accuracy achieved using simulated gait data suggests their potential for rapid prototyping of gait analysis systems, though clinical validation with patient data remains necessary. This work represents a significant step toward accessible, objective gait assessment tools for clinical, community, and tele-rehabilitation settings.<\/jats:p>","DOI":"10.1371\/journal.pdig.0001004","type":"journal-article","created":{"date-parts":[[2025,9,16]],"date-time":"2025-09-16T17:36:54Z","timestamp":1758044214000},"page":"e0001004","update-policy":"https:\/\/doi.org\/10.1371\/journal.pdig.corrections_policy","source":"Crossref","is-referenced-by-count":7,"title":["Classifying simulated gait impairments using privacy-preserving explainable artificial intelligence and mobile phone videos"],"prefix":"10.1371","volume":"4","author":[{"ORCID":"https:\/\/orcid.org\/0009-0007-0621-8130","authenticated-orcid":true,"given":"Lauhitya","family":"Reddy","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0000-3308-0949","authenticated-orcid":true,"given":"Ketan","family":"Anand","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0885-3836","authenticated-orcid":true,"given":"Shoibolina","family":"Kaushik","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Corey","family":"Rodrigo","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"J. Lucas","family":"McKay","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Trisha M.","family":"Kesar","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5693-3278","authenticated-orcid":true,"given":"Hyeokhyen","family":"Kwon","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2025,9,16]]},"reference":[{"key":"pdig.0001004.ref001","doi-asserted-by":"crossref","first-page":"128","DOI":"10.1159\/000456541","article-title":"Association between various brain pathologies and gait disturbance","volume":"43","author":"AMV Wennberg","year":"2017","journal-title":"Dement Geriatr Cogn Disord"},{"issue":"2","key":"pdig.0001004.ref002","doi-asserted-by":"crossref","first-page":"216","DOI":"10.1016\/0003-9993(94)90399-9","article-title":"Procedures for gait analysis","volume":"75","author":"GF Harris","year":"1994","journal-title":"Archives of Physical Medicine and Rehabilitation"},{"issue":"1","key":"pdig.0001004.ref003","doi-asserted-by":"crossref","first-page":"2","DOI":"10.1016\/S1356-689X(99)80003-4","article-title":"Gait analysis in the therapeutic environment","volume":"4","author":"F Coutts","year":"1999","journal-title":"Man Ther"},{"issue":"2","key":"pdig.0001004.ref004","first-page":"353","article-title":"Gait analysis: normal and pathological function","volume":"9","author":"J Perry","year":"2010","journal-title":"Journal of Sports Science & Medicine"},{"issue":"7","key":"pdig.0001004.ref005","doi-asserted-by":"crossref","first-page":"1027","DOI":"10.1093\/ptj\/65.7.1027","article-title":"Reliability of observational kinematic gait analysis","volume":"65","author":"DE Krebs","year":"1985","journal-title":"Phys Ther"},{"key":"pdig.0001004.ref006","doi-asserted-by":"crossref","first-page":"17","DOI":"10.1186\/1471-2474-6-17","article-title":"Reliability of videotaped observational gait analysis in patients with orthopedic impairments","volume":"6","author":"JJ Brunnekreef","year":"2005","journal-title":"BMC Musculoskelet Disord"},{"issue":"372","key":"pdig.0001004.ref007","first-page":"217","article-title":"The effect of preoperative gait analysis on orthopaedic decision making","author":"RM Kay","year":"2000","journal-title":"Clin Orthop Relat Res"},{"key":"pdig.0001004.ref008","doi-asserted-by":"crossref","first-page":"75","DOI":"10.1016\/j.gaitpost.2023.06.014","article-title":"The impact of three-dimensional gait analysis in adults with pathological gait on management recommendations","volume":"105","author":"M McGrath","year":"2023","journal-title":"Gait Posture"},{"issue":"133","key":"pdig.0001004.ref009","first-page":"57063","article-title":"Clinical-oriented three-dimensional gait analysis method for evaluating gait disorder","author":"M Mukaino","year":"2018","journal-title":"J Vis Exp"},{"key":"pdig.0001004.ref010","doi-asserted-by":"crossref","first-page":"901331","DOI":"10.3389\/fmedt.2022.901331","article-title":"Present and future of gait assessment in clinical practice: towards the application of novel trends and technologies","volume":"4","author":"AA Hulleck","year":"2022","journal-title":"Front Med Technol"},{"issue":"11","key":"pdig.0001004.ref011","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pone.0223549","article-title":"Quantifying normal and parkinsonian gait features from home movies: Practical application of a deep learning-based 2D pose estimator","volume":"14","author":"K Sato","year":"2019","journal-title":"PLoS One"},{"issue":"4","key":"pdig.0001004.ref012","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pcbi.1008935","article-title":"Two-dimensional video-based analysis of human gait using pose estimation","volume":"17","author":"J Stenum","year":"2021","journal-title":"PLoS Comput Biol"},{"key":"pdig.0001004.ref013","article-title":"BlazePose: on-device real-time body pose tracking","author":"V Bazarevsky","year":"2020","journal-title":"arXiv preprint"},{"key":"pdig.0001004.ref014","article-title":"OpenPose: realtime multi-person 2D pose estimation using part affinity fields","author":"Z Cao","year":"2019","journal-title":"arXiv preprint"},{"issue":"3","key":"pdig.0001004.ref015","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pdig.0000467","article-title":"Clinical gait analysis using video-based pose estimation: multiple perspectives, clinical populations, and measuring change","volume":"3","author":"J Stenum","year":"2024","journal-title":"PLOS Digit Health"},{"issue":"1","key":"pdig.0001004.ref016","doi-asserted-by":"crossref","first-page":"3840","DOI":"10.1038\/s41598-024-53217-7","article-title":"Validation of portable in-clinic video-based gait analysis for prosthesis users","volume":"14","author":"A Cimorelli","year":"2024","journal-title":"Sci Rep"},{"issue":"1","key":"pdig.0001004.ref017","doi-asserted-by":"crossref","first-page":"4054","DOI":"10.1038\/s41467-020-17807-z","article-title":"Deep neural networks enable quantitative movement analysis using single-camera videos","volume":"11","author":"\u0141 Kidzi\u0144ski","year":"2020","journal-title":"Nat Commun"},{"key":"pdig.0001004.ref018","doi-asserted-by":"crossref","first-page":"102871","DOI":"10.1016\/j.media.2023.102871","article-title":"Automatic labeling of Parkinson\u2019s disease gait videos with weak supervision","volume":"89","author":"M Gholami","year":"2023","journal-title":"Med Image Anal"},{"key":"pdig.0001004.ref019","unstructured":"Cost of a Data Breach Report 2024 . 2024. https:\/\/www.ibm.com\/reports\/data-breach"},{"key":"pdig.0001004.ref020","article-title":"MediaPipe: a framework for building perception pipelines","author":"C Lugaresi","year":"2019","journal-title":"arXiv preprint"},{"key":"pdig.0001004.ref021","unstructured":"Gandbhir VN, Lam JC, Lui F, Rayi A. Trendelenburg Gait. StatPearls. Treasure Island (FL): StatPearls Publishing; 2025."},{"key":"pdig.0001004.ref022","article-title":"Reducing circumduction and hip hiking during hemiparetic walking through targeted assistance of the paretic limb using a soft robotic exosuit","volume":"96","author":"LN Awad","year":"2017","journal-title":"Am J Phys Med Rehabil"},{"issue":"2","key":"pdig.0001004.ref023","doi-asserted-by":"crossref","first-page":"026113","DOI":"10.1063\/1.3147408","article-title":"Gait dynamics in Parkinson\u2019s disease: common and distinct behavior among stride length, gait variability, and fractal-like scaling","volume":"19","author":"JM Hausdorff","year":"2009","journal-title":"Chaos"},{"issue":"6","key":"pdig.0001004.ref024","doi-asserted-by":"crossref","first-page":"623","DOI":"10.1016\/j.clinbiomech.2015.03.022","article-title":"Evolution of vaulting strategy during locomotion of individuals with transfemoral amputation on slopes and cross-slopes compared to level walking","volume":"30","author":"C Villa","year":"2015","journal-title":"Clin Biomech (Bristol)"},{"issue":"12","key":"pdig.0001004.ref025","doi-asserted-by":"crossref","first-page":"1913","DOI":"10.1007\/s43465-023-01002-5","article-title":"Crouch gait in cerebral palsy: current concepts review","volume":"57","author":"RA Pandey","year":"2023","journal-title":"Indian J Orthop"},{"issue":"3","key":"pdig.0001004.ref026","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/2499621","article-title":"A tutorial on human activity recognition using body-worn inertial sensors","volume":"46","author":"A Bulling","year":"2014","journal-title":"ACM Comput Surv"},{"key":"pdig.0001004.ref027","first-page":"265","article-title":"On the algorithmic implementation of multiclass kernel-based vector machines","volume":"2","author":"K Crammer","year":"2001","journal-title":"Journal of Machine Learning Research"},{"issue":"1","key":"pdig.0001004.ref028","doi-asserted-by":"crossref","first-page":"5","DOI":"10.1023\/A:1010933404324","article-title":"Random forests","volume":"45","author":"L Breiman","year":"2001","journal-title":"Machine Learning"},{"key":"pdig.0001004.ref029","doi-asserted-by":"crossref","unstructured":"Chen T, Guestrin C. XGBoost. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 2016. p. 785\u201394. https:\/\/doi.org\/10.1145\/2939672.2939785","DOI":"10.1145\/2939672.2939785"},{"key":"pdig.0001004.ref030","doi-asserted-by":"crossref","unstructured":"Hammerla NY, Kirkham R, Andras P, Ploetz T. On preserving statistical characteristics of accelerometry data using their empirical cumulative distribution. In: Proceedings of the 2013 International Symposium on Wearable Computers. 2013. p. 65\u20138. https:\/\/doi.org\/10.1145\/2493988.2494353","DOI":"10.1145\/2493988.2494353"},{"key":"pdig.0001004.ref031","doi-asserted-by":"crossref","first-page":"72","DOI":"10.1016\/j.neucom.2018.03.067","article-title":"Time series FeatuRe extraction on basis of scalable hypothesis tests (tsfresh \u2013 a python package)","volume":"307","author":"M Christ","year":"2018","journal-title":"Neurocomputing"},{"key":"pdig.0001004.ref032","article-title":"Distributed and parallel time series feature extraction for industrial big data applications","author":"M Christ","year":"2017","journal-title":"arXiv preprint"},{"issue":"4","key":"pdig.0001004.ref033","doi-asserted-by":"crossref","DOI":"10.1214\/aos\/1013699998","article-title":"The control of the false discovery rate in multiple testing under dependency","volume":"29","author":"Y Benjamini","year":"2001","journal-title":"Ann Statist"},{"issue":"1","key":"pdig.0001004.ref034","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/1007730.1007733","article-title":"Editorial","volume":"6","author":"NV Chawla","year":"2004","journal-title":"SIGKDD Explor Newsl"},{"key":"pdig.0001004.ref035","doi-asserted-by":"crossref","first-page":"321","DOI":"10.1613\/jair.953","article-title":"SMOTE: synthetic minority over-sampling technique","volume":"16","author":"NV Chawla","year":"2002","journal-title":"jair"},{"issue":"1","key":"pdig.0001004.ref036","doi-asserted-by":"crossref","first-page":"12183","DOI":"10.1038\/s41598-021-91797-w","article-title":"XGBoost based machine learning approach to predict the risk of fall in older adults using gait outcomes","volume":"11","author":"B Noh","year":"2021","journal-title":"Sci Rep"},{"issue":"19","key":"pdig.0001004.ref037","doi-asserted-by":"crossref","first-page":"8330","DOI":"10.3390\/s23198330","article-title":"Classifying tremor dominant and postural instability and gait difficulty subtypes of Parkinson\u2019s disease from full-body kinematics","volume":"23","author":"NJ Gong","year":"2023","journal-title":"Sensors (Basel)"},{"issue":"3","key":"pdig.0001004.ref038","doi-asserted-by":"crossref","first-page":"167","DOI":"10.3390\/bios12030167","article-title":"Explainable artificial intelligence and wearable sensor-based gait analysis to identify patients with osteopenia and sarcopenia in daily life","volume":"12","author":"J-K Kim","year":"2022","journal-title":"Biosensors (Basel)"},{"key":"pdig.0001004.ref039","doi-asserted-by":"crossref","first-page":"115222","DOI":"10.1016\/j.eswa.2021.115222","article-title":"Nested cross-validation when selecting classifiers is overzealous for most practical applications","volume":"182","author":"J Wainer","year":"2021","journal-title":"Expert Systems with Applications"},{"issue":"10","key":"pdig.0001004.ref040","doi-asserted-by":"crossref","first-page":"3093","DOI":"10.1093\/bioinformatics\/btaa046","article-title":"Consensus features nested cross-validation","volume":"36","author":"S Parvandeh","year":"2020","journal-title":"Bioinformatics"},{"key":"pdig.0001004.ref041","doi-asserted-by":"crossref","unstructured":"Akiba T, Sano S, Yanase T, Ohta T, Koyama M. Optuna. In: Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. 2019. p. 2623\u201331. https:\/\/doi.org\/10.1145\/3292500.3330701","DOI":"10.1145\/3292500.3330701"},{"key":"pdig.0001004.ref042","unstructured":"Binomial proportion confidence interval. 2024. https:\/\/en.wikipedia.org\/w\/index.php?title=Binomial_proportion_confidence_interval&oldid=1223672356"},{"issue":"10","key":"pdig.0001004.ref043","doi-asserted-by":"crossref","first-page":"1340","DOI":"10.1093\/bioinformatics\/btq134","article-title":"Permutation importance: a corrected feature importance measure","volume":"26","author":"A Altmann","year":"2010","journal-title":"Bioinformatics"},{"issue":"3","key":"pdig.0001004.ref044","doi-asserted-by":"crossref","first-page":"1937","DOI":"10.1007\/s10462-020-09896-5","article-title":"A comparative analysis of gradient boosting algorithms","volume":"54","author":"C Bent\u00e9jac","year":"2020","journal-title":"Artif Intell Rev"},{"issue":"1","key":"pdig.0001004.ref045","doi-asserted-by":"crossref","first-page":"55","DOI":"10.1023\/B:VISI.0000042934.15159.49","article-title":"Pictorial structures for object recognition","volume":"61","author":"PF Felzenszwalb","year":"2005","journal-title":"International Journal of Computer Vision"},{"key":"pdig.0001004.ref046","doi-asserted-by":"crossref","unstructured":"Toshev A, Szegedy C. DeepPose: human pose estimation via deep neural networks. In: 2014 IEEE Conference on Computer Vision and Pattern Recognition. 2014. p. 1653\u201360. https:\/\/doi.org\/10.1109\/cvpr.2014.214","DOI":"10.1109\/CVPR.2014.214"},{"issue":"10","key":"pdig.0001004.ref047","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pcbi.1011556","article-title":"Discovering individual-specific gait signatures from data-driven models of neuromechanical dynamics","volume":"19","author":"TS Winner","year":"2023","journal-title":"PLoS Comput Biol"},{"key":"pdig.0001004.ref048","doi-asserted-by":"crossref","unstructured":"Winner TS, Rosenberg MC, Berman GJ, Kesar TM, Ting LH. Gait signature changes with walking speed are similar among able-bodied young adults despite persistent individual-specific differences; 2024. https:\/\/www.biorxiv.org\/content\/10.1101\/2024.05.01.591976v1","DOI":"10.1101\/2024.05.01.591976"},{"issue":"8","key":"pdig.0001004.ref049","doi-asserted-by":"crossref","first-page":"1593","DOI":"10.1007\/s10439-013-0852-0","article-title":"New perspectives in human movement variability","volume":"41","author":"T Lockhart","year":"2013","journal-title":"Ann Biomed Eng"},{"issue":"5","key":"pdig.0001004.ref050","doi-asserted-by":"crossref","first-page":"869","DOI":"10.1016\/j.humov.2011.06.002","article-title":"Human movement variability, nonlinear dynamics, and pathology: is there a connection?","volume":"30","author":"N Stergiou","year":"2011","journal-title":"Hum Mov Sci"},{"issue":"5","key":"pdig.0001004.ref051","doi-asserted-by":"crossref","first-page":"518","DOI":"10.1016\/j.clinbiomech.2014.03.013","article-title":"Frontal plane compensatory strategies associated with self-selected walking speed in individuals post-stroke","volume":"29","author":"VA Stanhope","year":"2014","journal-title":"Clin Biomech (Bristol)"},{"issue":"3","key":"pdig.0001004.ref052","doi-asserted-by":"crossref","first-page":"392","DOI":"10.2522\/ptj.20090425","article-title":"Influence of systematic increases in treadmill walking speed on gait kinematics after stroke","volume":"91","author":"CM Tyrell","year":"2011","journal-title":"Phys Ther"},{"issue":"4","key":"pdig.0001004.ref053","doi-asserted-by":"crossref","first-page":"555","DOI":"10.1016\/j.gaitpost.2013.02.006","article-title":"The how and why of arm swing during human walking","volume":"38","author":"P Meyns","year":"2013","journal-title":"Gait Posture"},{"key":"pdig.0001004.ref054","doi-asserted-by":"crossref","first-page":"1021","DOI":"10.3389\/fphys.2018.01021","article-title":"Post-stroke hemiplegic gait: new perspective and insights","volume":"9","author":"S Li","year":"2018","journal-title":"Front Physiol"},{"issue":"2","key":"pdig.0001004.ref055","doi-asserted-by":"crossref","first-page":"97","DOI":"10.3988\/jcn.2013.9.2.97","article-title":"Hemispheric differences in ischemic stroke: is left-hemisphere stroke more common?","volume":"9","author":"VS Hedna","year":"2013","journal-title":"J Clin Neurol"},{"issue":"1","key":"pdig.0001004.ref056","doi-asserted-by":"crossref","first-page":"115","DOI":"10.3390\/s16010115","article-title":"Deep convolutional and LSTM recurrent neural networks for multimodal wearable activity recognition","volume":"16","author":"FJ Ord\u00f3\u00f1ez","year":"2016","journal-title":"Sensors (Basel)"},{"key":"pdig.0001004.ref057","article-title":"Deep ConvLSTM with self-attention for human activity decoding using wearables","author":"SP Singh","year":"2020","journal-title":"arXiv preprint"},{"key":"pdig.0001004.ref058","article-title":"Spatial temporal graph convolutional networks for skeleton-based action recognition","author":"S Yan","year":"2018","journal-title":"arXiv preprint"},{"key":"pdig.0001004.ref059","article-title":"ViTPose: simple vision transformer baselines for human pose estimation","author":"Y Xu","year":"2022","journal-title":"arXiv preprint"},{"key":"pdig.0001004.ref060","unstructured":"Sundararajan M, Najmi A. The many shapley values for model explanation. In: Proceedings of the 37th International Conference on Machine Learning. 2020. p. 9269\u201378. https:\/\/proceedings.mlr.press\/v119\/sundararajan20b.html"}],"container-title":["PLOS Digital Health"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pdig.0001004","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,9,16]],"date-time":"2025-09-16T17:37:05Z","timestamp":1758044225000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pdig.0001004"}},"subtitle":[],"editor":[{"given":"Calvin","family":"Or","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2025,9,16]]},"references-count":60,"journal-issue":{"issue":"9","published-online":{"date-parts":[[2025,9,16]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pdig.0001004","relation":{},"ISSN":["2767-3170"],"issn-type":[{"value":"2767-3170","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,9,16]]}}}