{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,16]],"date-time":"2026-09-16T02:29:54Z","timestamp":1789525794325,"version":"build-2803163510"},"reference-count":24,"publisher":"MDPI AG","issue":"3","license":[{"start":{"date-parts":[[2026,3,19]],"date-time":"2026-03-19T00:00:00Z","timestamp":1773878400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"Ministry of Education","award":["2025-RISE-10-001"],"award-info":[{"award-number":["2025-RISE-10-001"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Bioengineering"],"abstract":"<jats:p>Parkinson\u2019s disease (PD) is associated with progressive gait deterioration; however, widely used clinical scales such as the Hoehn &amp; Yahr (H&amp;Y) stage are limited in capturing continuous severity changes due to subjectivity and discrete grading. This study proposes a two-stage explainable AI framework using vertical ground reaction force (VGRF) signals to achieve reproducible PD detection and continuous severity estimation. In the first stage, three deep learning models, temporal convolutional network (TCN), BiGRU with attention, and FCNN-Transformer, were trained using windowed VGRF signals under repeated subject-wise data segmentation. All models achieved high discrimination performance (AUC \u2265 0.93), with FCNN-Transformer showing the highest mean AUC (0.940) and statistically superior performance (paired Wilcoxon test, p &lt; 0.05). Stability-based explainable AI using Integrated Gradients consistently identified variability-related VGRF features as the most informative, which were also significantly different between groups at the data level (p &lt; 0.001, FDR-corrected). In the second stage, XGBoost regression was applied to PD subjects to predict continuous H&amp;Y severity, achieving strong correlation with clinical grades (Spearman \u03c1 = 0.921, p &lt; 0.001), low error (MAE = 0.158, RMSE = 0.241), and high determination (R2 = 0.953). This shows that gait-based features are a sensitive enough signal to continuously quantify disease progression. In addition, in the TREND prospective longitudinal cohort (n = 696), wearable walking indicators differed significantly from those of non-patients prior to diagnosis, and a decline in walking pace was observed approximately four years before Parkinson\u2019s disease diagnosis, providing the basis for early screening and monitoring using gait-based digital biomarkers. These results demonstrate that gait-based digital biomarkers can objectively quantify both PD presence and disease progression. The proposed framework provides a reproducible, explainable, and clinically interpretable AI-based decision support approach for PD assessment.<\/jats:p>","DOI":"10.3390\/bioengineering13030360","type":"journal-article","created":{"date-parts":[[2026,3,19]],"date-time":"2026-03-19T11:50:36Z","timestamp":1773921036000},"page":"360","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":5,"title":["Digital Gait Biomarkers for Parkinson\u2019s Disease: Subject-Wise Validated Explainable AI Framework Using Vertical Ground Reaction Force Signals"],"prefix":"10.3390","volume":"13","author":[{"given":"Moonhyeok","family":"Choi","sequence":"first","affiliation":[{"name":"Department of Electronic and Communication Engineering, Catholic Kwandong University, 24 Beomil-ro 579 Beongil, Gangneung-si 25601, Republic of Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jaehyun","family":"Jo","sequence":"additional","affiliation":[{"name":"Department of Digital Healthcare, Catholic Kwandong University, 24 Beomil-ro 579 Beongil, Gangneung-si 25601, Republic of Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-7273-7279","authenticated-orcid":false,"given":"Jinhyoung","family":"Jeong","sequence":"additional","affiliation":[{"name":"Department of Healthcare Management, Catholic Kwandong University, 24 Beomil-ro 579 Beongil, Gangneung-si 25601, Republic of Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2026,3,19]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"187","DOI":"10.1038\/s41531-025-01025-9","article-title":"Machine learning for Parkinson\u2019s disease: A comprehensive review of datasets, algorithms, and challenges","volume":"11","author":"Shokrpour","year":"2025","journal-title":"npj Park. Dis."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"6452","DOI":"10.1007\/s00415-024-12611-x","article-title":"Machine learning and wearable sensors for automated Parkinson\u2019s disease diagnosis aid: A systematic review","volume":"271","author":"Pecoraro","year":"2024","journal-title":"J. Neurol."},{"key":"ref_3","doi-asserted-by":"crossref","unstructured":"Franco, A., Russo, M., Amboni, M., Ponsiglione, A.M., Di Filippo, F., Romano, M., Amato, F., and Ricciardi, C. (2024). The Role of Deep Learning and Gait Analysis in Parkinson\u2019s Disease: A Systematic Review. Sensors, 24.","DOI":"10.3390\/s24185957"},{"key":"ref_4","doi-asserted-by":"crossref","unstructured":"Emamzadeh, F.N., and Surguchov, A. (2018). Parkinson\u2019s disease: Biomarkers, Treatment, Risk Factors. Front. Neurosci., 12.","DOI":"10.3389\/fnins.2018.00612"},{"key":"ref_5","doi-asserted-by":"crossref","unstructured":"Meral, M., and Ozbilgin, F. (2025). Parkinson\u2019s Disease Diagnosis and Severity Assessment from Gait Signals via Bayesian-Optimized Deep Learning. Diagnostics, 15.","DOI":"10.3390\/diagnostics15162046"},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"24","DOI":"10.1007\/s44163-025-00241-9","article-title":"A review of machine learning and deep learning for Parkinson\u2019s disease detection","volume":"5","author":"Rabie","year":"2025","journal-title":"Discov. Artif. Intell."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"100","DOI":"10.1109\/JBHI.2025.3589772","article-title":"VGRF Signal-Based Gait Analysis for Parkinson\u2019s Disease Detection: A Multi-Scale Directed Graph Neural Network Approach","volume":"30","author":"Wang","year":"2026","journal-title":"IEEE J. Biomed. Health Inform."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"16633","DOI":"10.1038\/s41598-023-43542-8","article-title":"Risk of data leakage in estimating the diagnostic performance of a deep-learning-based computer-aided system for psychiatric disorders","volume":"13","author":"Lee","year":"2023","journal-title":"Sci. Rep."},{"key":"ref_9","unstructured":"Wesarg, S., Ant\u00f3n, E.P., Baxter, J.S.H., Erdt, M., Drechsler, K., Laura, C.O., Freiman, M., Chen, Y., Rekik, I., and Eagleson, R. (2023). How You Split Matters: Data Leakage and Subject Characteristics Studies in Longitudinal Brain MRI Analysis. Clinical Image-Based Procedures, Fairness of AI in Medical Imaging, and Ethical and Philosophical Issues in Medical Imaging. CLIP EPIMI FAIMI 2023, Springer. Lecture Notes in Computer Science."},{"key":"ref_10","doi-asserted-by":"crossref","unstructured":"Balletti, N., Guglielmi, E., Laudato, G., Oliveto, R., Simeone, J., and Zinni, R. (2025, January 20\u201322). Integrating Gait and Clinical Data with Explainable Artificial Intelligence for Parkinson\u2019s Prediction: The EDAM System. Proceedings of the 18th International Conference on Health Informatics, Porto, Portugal.","DOI":"10.5220\/0013179400003911"},{"key":"ref_11","doi-asserted-by":"crossref","unstructured":"Ndlovu, B., Maguraushe, K., and Mabikwa, O. (2025). Machine Learning and Explainable AI for Parkinson\u2019s Disease Prediction: A Systematic Review. Indones. J. Comput. Sci., 14.","DOI":"10.33022\/ijcs.v14i2.4837"},{"key":"ref_12","first-page":"103541","article-title":"Explainable AI for Parkinson\u2019s disease prediction: A machine learning approach with interpretable models","volume":"73","author":"Esan","year":"2025","journal-title":"Curr. Res. Transl. Med."},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Alharthi, A.S. (2024). Interpretable machine learning comprehensive human gait deterioration analysis. Front. Neuroinform., 18.","DOI":"10.3389\/fninf.2024.1451529"},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Rodriguez-Martin, D., Catal\u00e0, A., and Cabestany, J. (2025). Improving Parkinson\u2019s disease management through wearable technology: A cost-benefit perspective. PLoS ONE, 20.","DOI":"10.1371\/journal.pone.0334413"},{"key":"ref_15","doi-asserted-by":"crossref","unstructured":"Rodr\u00edguez-Mart\u00edn, D., Cabestany, J., P\u00e9rez-L\u00f3pez, C., Pie, M., Calvet, J., Sam\u00e0, A., Capra, C., Catal\u00e0, A., and Rodr\u00edguez-Molinero, A. (2022). A New Paradigm in Parkinson\u2019s Disease Evaluation With Wearable Medical Devices: A Review of STAT-ONTM. Front. Neurol., 13.","DOI":"10.3389\/fneur.2022.912343"},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Uchaipichat, N., Tangjirachaipoka, A., and Chamninavakul, C. (2025). Machine Learning-Based Detection of Parkinson\u2019s Disease Using Vertical Ground Reaction Force and Gait Analysis. Proceedings of the 2024 9th International Conference on Biomedical Imaging, Signal Processing (ICBSP \u201824), Association for Computing Machinery.","DOI":"10.1145\/3707172.3707195"},{"key":"ref_17","doi-asserted-by":"crossref","unstructured":"Mittal, R., Agarwal, N., Dubey, M., Pathak, V., Shukla, P., Rani, G., Vocaturo, E., and Zumpano, E. (2025). Machine learning approach to gait analysis for Parkinson\u2019s disease detection and severity classification. Front. Robot. AI, 12.","DOI":"10.3389\/frobt.2025.1623529"},{"key":"ref_18","doi-asserted-by":"crossref","unstructured":"Al-Adhaileh, M.H., Wadood, A., Aldhyani, T.H.H., Khan, S., Uddin, M.I., and Al-Nefaie, A.H. (2025). Deep learning techniques for detecting freezing of gait episodes in Parkinson\u2019s disease using wearable sensors. Front. Physiol., 16.","DOI":"10.3389\/fphys.2025.1581699"},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"328","DOI":"10.1038\/s41598-024-83357-9","article-title":"Gait-based Parkinson\u2019s disease diagnosis and severity classification using force sensors and machine learning","volume":"15","author":"Sharma","year":"2025","journal-title":"Sci. Rep."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"106345","DOI":"10.1016\/j.neunet.2024.106345","article-title":"SEGAL time series classification\u2014Stable explanations using a generative model and an adaptive weighting method for LIME","volume":"176","author":"Meng","year":"2024","journal-title":"Neural Netw."},{"key":"ref_21","first-page":"010101","article-title":"Effect sizes for nonparametric tests","volume":"36","year":"2026","journal-title":"Biochem. Med."},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"51","DOI":"10.1038\/s41531-025-00897-1","article-title":"Digital gait biomarkers in Parkinson\u2019s disease: Susceptibility\/risk, progression, response to exercise, and prognosis","volume":"11","author":"Mancini","year":"2025","journal-title":"npj Park. Dis."},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"189","DOI":"10.4258\/hir.2021.27.3.189","article-title":"Impact of the Choice of Cross-Validation Techniques on the Results of Machine Learning-Based Diagnostic Applications","volume":"27","author":"Tougui","year":"2021","journal-title":"Healthc. Inform. Res."},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"34515","DOI":"10.1038\/s41598-025-17703-w","article-title":"Explainability and importance estimate of time series classifier via embedded neural network","volume":"15","author":"Chan","year":"2025","journal-title":"Sci. Rep."}],"container-title":["Bioengineering"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2306-5354\/13\/3\/360\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,21]],"date-time":"2026-03-21T05:24:12Z","timestamp":1774070652000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2306-5354\/13\/3\/360"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3,19]]},"references-count":24,"journal-issue":{"issue":"3","published-online":{"date-parts":[[2026,3]]}},"alternative-id":["bioengineering13030360"],"URL":"https:\/\/doi.org\/10.3390\/bioengineering13030360","relation":{},"ISSN":["2306-5354"],"issn-type":[{"value":"2306-5354","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,3,19]]}}}