{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,18]],"date-time":"2026-08-18T05:08:17Z","timestamp":1787029697063,"version":"build-2736575974"},"reference-count":22,"publisher":"MDPI AG","issue":"4","license":[{"start":{"date-parts":[[2026,2,20]],"date-time":"2026-02-20T00:00:00Z","timestamp":1771545600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>Freezing of gait (FoG) affects 38\u201365% of advanced Parkinson\u2019s disease patients, yet automated detection algorithms are often validated solely on laboratory datasets. This study quantifies the critical performance gap between laboratory and real-world performance\u2014a prerequisite for clinical deployment. Using temporal convolutional networks (TCNs), we trained models on two public datasets representing ecological extremes: a daily living dataset (Figshare; n = 35, single-sensor) and a laboratory dataset (DAPHNET; n = 10, multi-sensor). We compared five training configurations to address class imbalance. Results showed that F1-based early stopping outperformed Area Under the Curve (AUC)-based stopping by 47% (F1: 0.55 vs. 0.37, p = 0.0008). Combining multiple imbalance corrections (focal loss, weighting, sampling) paradoxically degraded precision to 33% due to a ~60-fold over-weighting of the minority class. Most importantly, cross-dataset validation revealed an 83% performance gap: laboratory F1 reached 0.9999 \u00b1 0.0002, whereas daily living F1 dropped to 0.55 \u00b1 0.26 (p &lt; 0.0001), with a 1299-fold increase in variance. These findings demonstrate that laboratory success does not guarantee real-world utility. We propose that the observed gap represents a \u201cdeployment gap\u201d reflecting the combined influence of environmental complexity, sensor constraints, and physiological variability. These results provide an empirical framework for evaluating deployment readiness of wearable FoG detection systems and offer concrete training strategy recommendations for clinical translation.<\/jats:p>","DOI":"10.3390\/s26041352","type":"journal-article","created":{"date-parts":[[2026,2,20]],"date-time":"2026-02-20T10:32:37Z","timestamp":1771583557000},"page":"1352","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":1,"title":["Deep Learning for Freezing of Gait Detection: Cross-Dataset Validation Reveals Critical Deployment Gaps Between Laboratory and Daily Living Wearable Monitoring"],"prefix":"10.3390","volume":"26","author":[{"given":"Wei","family":"Lin","sequence":"first","affiliation":[{"name":"Department of Neurosurgery, The 904th Hospital of the Joint Logistics Support Force of People\u2019s Liberation Army, Wuxi 214000, China"},{"name":"Department of Neurologic Surgery, Mayo Clinic Florida, Jacksonville, FL 32224, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sanjeet S.","family":"Grewal","sequence":"additional","affiliation":[{"name":"Department of Neurologic Surgery, Mayo Clinic Florida, Jacksonville, FL 32224, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2026,2,20]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"4853","DOI":"10.1038\/s41467-024-49027-0","article-title":"A machine learning contest enhances automated freezing of gait detection and reveals time-of-day effects","volume":"15","author":"Salomon","year":"2024","journal-title":"Nat. Commun."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"734","DOI":"10.1016\/S1474-4422(11)70143-0","article-title":"Freezing of gait: Moving forward on a mysterious clinical phenomenon","volume":"10","author":"Nutt","year":"2011","journal-title":"Lancet Neurol."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"431","DOI":"10.1007\/s10072-023-07017-y","article-title":"Detection and prediction of freezing of gait with wearable sensors in Parkinson disease","volume":"45","author":"Zhang","year":"2024","journal-title":"Neurol. Sci."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"102459","DOI":"10.1016\/j.artmed.2022.102459","article-title":"Real-time detection of freezing of gait in Parkinson disease using multi-head convolutional neural networks and a single inertial sensor","volume":"135","author":"Mazzetta","year":"2023","journal-title":"Artif. Intell. Med."},{"key":"ref_5","doi-asserted-by":"crossref","unstructured":"May, D.S., Tueth, L.E., Earhart, G.M., and Mazzoni, P. (2023). Using wearable sensors to assess freezing of gait in the real world. Bioengineering, 10.","DOI":"10.3390\/bioengineering10030289"},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"1010","DOI":"10.1109\/TBCAS.2023.3281596","article-title":"Flexible gel-free multi-modal wireless sensors with edge deep learning for detecting and alerting freezing of gait symptom","volume":"17","author":"Hou","year":"2023","journal-title":"IEEE Trans. Biomed. Circuits Syst."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"119","DOI":"10.1016\/j.knosys.2017.10.017","article-title":"Deep learning for freezing of gait detection in Parkinson disease patients in their homes using a waist-worn inertial measurement unit","volume":"139","author":"Camps","year":"2018","journal-title":"Knowl.-Based Syst."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"105613","DOI":"10.1016\/j.bspc.2023.105613","article-title":"Episode-level prediction of freezing of gait based on wearable inertial signals using a deep neural network model","volume":"88","author":"Huang","year":"2024","journal-title":"Biomed. Signal Process. Control"},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1186\/s12984-020-00774-3","article-title":"Measuring freezing of gait during daily-life: An open-source, wearable sensors approach","volume":"18","author":"Mancini","year":"2021","journal-title":"J. Neuroeng. Rehabil."},{"key":"ref_10","doi-asserted-by":"crossref","unstructured":"Pardoel, S., Kofman, J., Nantel, J., and Lemaire, E.D. (2019). Wearable-sensor-based detection and prediction of freezing of gait in Parkinson\u2019s disease: A review. Sensors, 19.","DOI":"10.3390\/s19235141"},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"20","DOI":"10.1186\/s12984-022-00992-x","article-title":"Assessing inertial measurement unit locations for freezing of gait detection and patient preference","volume":"19","author":"Lee","year":"2022","journal-title":"J. Neuroeng. Rehabil."},{"key":"ref_12","unstructured":"Bai, S., Kolter, J.Z., and Koltun, V. (2018). An empirical evaluation of generic convolutional and recurrent networks for sequence modeling. arXiv."},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Ismail, A.R., Jovanovic, S., Ramzan, N., and Rabah, H. (2023). ECG classification using an optimal temporal convolutional network for remote health monitoring. Sensors, 23.","DOI":"10.3390\/s23031697"},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"21247","DOI":"10.1038\/s41598-022-25472-z","article-title":"Temporal convolutional networks and data rebalancing for clinical length of stay and mortality prediction","volume":"12","author":"Bednarski","year":"2022","journal-title":"Sci. Rep."},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"120541","DOI":"10.1016\/j.eswa.2023.120541","article-title":"Deep learning and wearable sensors for the diagnosis and monitoring of Parkinson disease: A systematic review","volume":"229","author":"Sigcha","year":"2023","journal-title":"Expert Syst. Appl."},{"key":"ref_16","unstructured":"(2024, October 15). Parkinson Freezing of Gait Dataset. Available online: https:\/\/figshare.com\/."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"436","DOI":"10.1109\/TITB.2009.2036165","article-title":"Wearable assistant for Parkinson disease patients with the freezing of gait symptom","volume":"14","author":"Bachlin","year":"2010","journal-title":"IEEE Trans. Inf. Technol. Biomed."},{"key":"ref_18","doi-asserted-by":"crossref","unstructured":"Lin, T.Y., Goyal, P., Girshick, R., He, K., and Doll\u00e1r, P. (2017, January 22\u201329). Focal loss for dense object detection. Proceedings of the IEEE International Conference on Computer Vision, Venice, Italy.","DOI":"10.1109\/ICCV.2017.324"},{"key":"ref_19","doi-asserted-by":"crossref","unstructured":"Saito, T., and Rehmsmeier, M. (2015). The precision-recall plot is more informative than the ROC plot when evaluating binary classifiers on imbalanced datasets. PLoS ONE, 10.","DOI":"10.1371\/journal.pone.0118432"},{"key":"ref_20","first-page":"18970","article-title":"Label-imbalanced and group-sensitive classification under overparameterization","volume":"34","author":"Kini","year":"2021","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_21","doi-asserted-by":"crossref","unstructured":"Bouchouras, G., Sofianidis, G., and Kotis, K. (2025). Predicting freezing of gait in Parkinson\u2019s disease: A machine-learning-based approach in ON and OFF medication states. J. Clin. Med., 14.","DOI":"10.20944\/preprints202503.0205.v1"},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"319","DOI":"10.1109\/JBHI.2022.3215921","article-title":"Fall risk prediction in Parkinson disease using real-world inertial sensor gait data","volume":"27","author":"Ullrich","year":"2023","journal-title":"IEEE J. Biomed. Health Inform."}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/26\/4\/1352\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,2,20]],"date-time":"2026-02-20T10:34:07Z","timestamp":1771583647000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/26\/4\/1352"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,2,20]]},"references-count":22,"journal-issue":{"issue":"4","published-online":{"date-parts":[[2026,2]]}},"alternative-id":["s26041352"],"URL":"https:\/\/doi.org\/10.3390\/s26041352","relation":{},"ISSN":["1424-8220"],"issn-type":[{"value":"1424-8220","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,2,20]]}}}