{"id":1685,"date":"2026-10-06T19:04:24","date_gmt":"2026-10-06T19:04:24","guid":{"rendered":"https:\/\/www.cs.ubbcluj.ro\/~meco\/development-of-an-efficient-deep-learning-model-for-arrhythmia-detection-from-consumer-smartwatches-2026\/"},"modified":"2026-10-06T19:25:15","modified_gmt":"2026-10-06T19:25:15","slug":"development-of-an-efficient-deep-learning-model-for-arrhythmia-detection-from-consumer-smartwatches-2026","status":"publish","type":"post","link":"https:\/\/www.cs.ubbcluj.ro\/~meco\/development-of-an-efficient-deep-learning-model-for-arrhythmia-detection-from-consumer-smartwatches-2026\/","title":{"rendered":"Development of an efficient deep learning model for arrhythmia detection from consumer smartwatches (2026)"},"content":{"rendered":"<div class=\"entry-content\">\n<p>Frontiers in Artificial Intelligence<\/p>\n<h2>Authors<\/h2>\n<p>Dan-Cristian George, Mihai-Alexandru Hulubeac, \u015etefan-Lucian Gr\u0103mad\u0103, Z. B\u00e1lint, L. Dio\u015fan<\/p>\n<h2>Abstract<\/h2>\n<p>Cardiac arrhythmias often progress without clear symptoms and may remain undetected until severe complications occur. Continuous monitoring through consumer smartwatches offers a scalable and non-invasive opportunity for real-time rhythm screening in free-living conditions. However, wrist-based photoplethysmography (PPG) signals are highly susceptible to motion artifacts and inter-subject variability, limiting traditional signal-processing techniques. In this work, we compare classical machine learning and deep learning approaches for multi-class arrhythmia classification from real-life smartwatch PPG and accelerometer recordings collected in the Pulsewatch clinical trial. Three rhythm categories are considered: normal sinus rhythm (NSR), atrial fibrillation (AF), and premature atrial or ventricular contractions (PAC\/PVC).<\/p>\n<p>The evaluated approaches include a Random Forest classifier based on engineered temporal and morphological features, together with two end-to-end architectures, CNN\u2013GRU and CNN\u2013BiLSTM, designed to learn temporal patterns directly from raw two-channel smartwatch waveforms while addressing motion artifacts, sensor noise, and cross-subject generalization.<\/p>\n<p>Under two-fold subject-independent cross-validation, the Random Forest achieved the highest average accuracy among the proposed methods (91.22%), together with the best macro precision (85.89%), macro F1 score (84.17%), and macro ROC-AUC (97.03%). The CNN\u2013BiLSTM achieved the highest macro recall (87.37%), indicating stronger sensitivity to arrhythmic classes, while the CNN\u2013GRU remained competitive with a macro recall of 86.85% and a macro F1 score of 83.13%.<\/p>\n<p>Compared with the Pulsewatch baselines reported by Han et al., the proposed methods improved the available macro-level metrics, with the Random Forest obtaining the strongest macro precision and macro F1 score and the CNN\u2013BiLSTM the strongest macro recall, indicating improved class-balanced recognition of clinically relevant arrhythmic events. These results show that feature-based models remain highly competitive for smartwatch-based arrhythmia detection, achieving the strongest overall and macro F1 performance among the proposed methods, while hybrid convolutional\u2013recurrent architectures provide increased sensitivity to minority rhythm classes. Deployment-oriented profiling under an identical single-core benchmark showed a mean end-to-end inference time of 33.08 ms for the Random Forest pipeline and 51.87 ms for the CNN\u2013BiLSTM model on a 30-s segment, supporting near-real-time rhythm screening.<\/p>\n<p>Overall, this work highlights the trade-offs between the analyzed approaches in terms of accuracy, class-balanced performance, computational complexity, and deployment feasibility for wearable cardiac monitoring systems.<\/p>\n<h2>Citation<\/h2>\n<pre class=\"wp-block-preformatted\">@Inproceedings{George2026DevelopmentOA,\n author = {Dan-Cristian George and Mihai-Alexandru Hulubeac and \u015etefan-Lucian Gr\u0103mad\u0103 and Z. B\u00e1lint and L. Dio\u015fan},\n booktitle = {Frontiers in Artificial Intelligence},\n title = {Development of an efficient deep learning model for arrhythmia detection from consumer smartwatches},\n year = {2026}\n}<\/pre>\n<\/div>\n","protected":false},"excerpt":{"rendered":"<p>Frontiers in Artificial Intelligence Authors Dan-Cristian George, Mihai-Alexandru Hulubeac, \u015etefan-Lucian Gr\u0103mad\u0103, Z. B\u00e1lint, L. Dio\u015fan Abstract Cardiac arrhythmias often progress without clear symptoms and may remain undetected until severe complications occur. Continuous monitoring through consumer smartwatches offers a scalable and non-invasive opportunity for real-time rhythm screening in free-living conditions. However, wrist-based photoplethysmography (PPG) signals are&hellip;<\/p>\n","protected":false},"author":6,"featured_media":1711,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":[],"categories":[4],"tags":[69,11],"_links":{"self":[{"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/posts\/1685"}],"collection":[{"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/users\/6"}],"replies":[{"embeddable":true,"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/comments?post=1685"}],"version-history":[{"count":2,"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/posts\/1685\/revisions"}],"predecessor-version":[{"id":1712,"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/posts\/1685\/revisions\/1712"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/media\/1711"}],"wp:attachment":[{"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/media?parent=1685"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/categories?post=1685"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.cs.ubbcluj.ro\/~meco\/wp-json\/wp\/v2\/tags?post=1685"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}