{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,19]],"date-time":"2026-03-19T04:34:26Z","timestamp":1773894866807,"version":"3.50.1"},"reference-count":33,"publisher":"MDPI AG","issue":"5","license":[{"start":{"date-parts":[[2022,2,24]],"date-time":"2022-02-24T00:00:00Z","timestamp":1645660800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"Korea government","award":["No. 1711138362, KMDF_PR_20200901_0174-02"],"award-info":[{"award-number":["No. 1711138362, KMDF_PR_20200901_0174-02"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>The importance of an embedded wearable device with automatic detection and alarming cannot be overstated, given that 15\u201330% of patients with atrial fibrillation are reported to be asymptomatic. These asymptomatic patients do not seek medical care, hence traditional diagnostic tools including Holter are not effective for the further prevention of associated stroke or heart failure. This is likely to be more so in the era of COVID-19, in which patients become more reluctant on hospitalization and checkups. However, little literature is available on this important topic. For this reason, this study developed efficient deep learning with model compression, which is designed to use ECG data and classify arrhythmia in an embedded wearable device. ECG-signal data came from Korea University Anam Hospital in Seoul, Korea, with 28,308 unique patients (15,412 normal and 12,896 arrhythmia). Resnets and Mobilenets with model compression (TensorFlow Lite) were applied and compared for the diagnosis of arrhythmia in an embedded wearable device. The weight size of the compressed model registered a remarkable decrease from 743 MB to 76 KB (1\/10000), whereas its performance was almost the same as its original counterpart. Resnet and Mobilenet were similar in terms of accuracy, i.e., Resnet-50 Hz (97.3) vs. Mo-bilenet-50 Hz (97.2), Resnet-100 Hz (98.2) vs. Mobilenet-100 Hz (97.9). Here, 50 Hz\/100 Hz denotes the down-sampling rate. However, Resnets took more flash memory and longer inference time than did Mobilenets. In conclusion, Mobilenet would be a more efficient model than Resnet to classify arrhythmia in an embedded wearable device.<\/jats:p>","DOI":"10.3390\/s22051776","type":"journal-article","created":{"date-parts":[[2022,2,24]],"date-time":"2022-02-24T21:11:07Z","timestamp":1645737067000},"page":"1776","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":31,"title":["Compressed Deep Learning to Classify Arrhythmia in an Embedded Wearable Device"],"prefix":"10.3390","volume":"22","author":[{"given":"Kwang-Sig","family":"Lee","sequence":"first","affiliation":[{"name":"AI Center, Korea University Anam Hospital, Seoul 02841, Korea"}]},{"given":"Hyun-Joon","family":"Park","sequence":"additional","affiliation":[{"name":"Institute for Health Service Innovation, Korea University College of Medicine, Seoul 02841, Korea"}]},{"given":"Ji Eon","family":"Kim","sequence":"additional","affiliation":[{"name":"Department of Thoracic and Cardiovascular Surgery, Korea University College of Medicine, Korea University Anam Hospital, Seoul 02841, Korea"}]},{"given":"Hee Jung","family":"Kim","sequence":"additional","affiliation":[{"name":"Department of Thoracic and Cardiovascular Surgery, Korea University College of Medicine, Korea University Anam Hospital, Seoul 02841, Korea"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4406-6162","authenticated-orcid":false,"given":"Sangil","family":"Chon","sequence":"additional","affiliation":[{"name":"HUINNO Co., Ltd., Seoul 06011, Korea"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4949-9709","authenticated-orcid":false,"given":"Sangkyu","family":"Kim","sequence":"additional","affiliation":[{"name":"HUINNO Co., Ltd., Seoul 06011, Korea"}]},{"given":"Jaesung","family":"Jang","sequence":"additional","affiliation":[{"name":"HUINNO Co., Ltd., Seoul 06011, Korea"}]},{"given":"Jin-Kook","family":"Kim","sequence":"additional","affiliation":[{"name":"HUINNO Co., Ltd., Seoul 06011, Korea"}]},{"given":"Seongbin","family":"Jang","sequence":"additional","affiliation":[{"name":"HUINNO Co., Ltd., Seoul 06011, Korea"}]},{"given":"Yeongjoon","family":"Gil","sequence":"additional","affiliation":[{"name":"HUINNO Co., Ltd., Seoul 06011, Korea"}]},{"given":"Ho Sung","family":"Son","sequence":"additional","affiliation":[{"name":"Department of Thoracic and Cardiovascular Surgery, Korea University College of Medicine, Korea University Anam Hospital, Seoul 02841, Korea"}]}],"member":"1968","published-online":{"date-parts":[[2022,2,24]]},"reference":[{"key":"ref_1","unstructured":"World Health Organization (2021, December 01). Cardiovascular Diseases (CVDs). Available online: https:\/\/www.who.int\/news-room\/fact-sheets\/detail\/cardiovascular-diseases-(cvds)."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"837","DOI":"10.1161\/CIRCULATIONAHA.113.005119","article-title":"Worldwide epidemiology of atrial fibrillation: A Global Burden of Disease 2010 Study","volume":"129","author":"Chugh","year":"2014","journal-title":"Circulation"},{"key":"ref_3","doi-asserted-by":"crossref","unstructured":"Statistics Korea (2021). Year 2020 Statistics on Causes of Death in Korea.","DOI":"10.1787\/bd274241-en"},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"e75","DOI":"10.3346\/jkms.2019.34.e75","article-title":"Trends and patterns of burden of disease and injuries in Korea using disability-adjusted life years","volume":"34","author":"Kim","year":"2019","journal-title":"J. Korean Med. Sci."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"2010","DOI":"10.1136\/heartjnl-2017-312930","article-title":"Increasing trends in hospital care burden of atrial fibrillation in Korea, 2006 through 2015","volume":"104","author":"Kim","year":"2018","journal-title":"Heart"},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"268","DOI":"10.1016\/j.procs.2017.11.238","article-title":"Cardiac arrhythmia detection using deep learning","volume":"120","author":"Isin","year":"2017","journal-title":"Procedia Comput. Sci."},{"key":"ref_7","unstructured":"Rajpurkar, P., Hannun, A.Y., Haghpanahi, M., Bourn, C., and Ng, A.Y. (2017). Cardiologist-level arrhythmia detection with convolutional neural networks. arXiv Prepr., Available online: https:\/\/arxiv.org\/abs\/1707.01836."},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Li, D., Zhang, J., Zhang, Q., and Wei, Z. (2017, January 12\u201315). Classification of ECG signals based on 1d convolutional neural network. Proceedings of the 2017 IEEE 19th International Conference on e-Health Networking, Applications and Services (Healthcom), Dalian, China.","DOI":"10.1109\/HealthCom.2017.8210784"},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"446","DOI":"10.1016\/j.future.2018.03.057","article-title":"A deep learning approach for ECG-based heartbeat classification for arrhythmia detection","volume":"86","author":"Sannino","year":"2018","journal-title":"Future Gener. Comput. Syst."},{"key":"ref_10","unstructured":"Jun, T.J., Nguyen, H.M., Kang, D., Kim, D., Kim, D., and Kim, Y.H. (2018). ECG arrhythmia classification using a 2-D convolutional neural network. arXiv Prepr., Available online: https:\/\/arxiv.org\/abs\/1804.06812."},{"key":"ref_11","doi-asserted-by":"crossref","unstructured":"Lee, K.S., Jung, S., Gil, Y., and Son, H.S. (2019). Atrial fibrillation classification based on convolutional neural networks. BMC Med. Inform. Decis. Mak., 19.","DOI":"10.1186\/s12911-019-0946-1"},{"key":"ref_12","doi-asserted-by":"crossref","unstructured":"Park, J., Kim, J.-K., Jung, S., Gil, Y., Choi, J.-I., and Son, H.S. (2020). ECG-signal multi-classification model based on squeeze-and-excitation residual neural networks. Appl. Sci., 10.","DOI":"10.21203\/rs.3.rs-34360\/v1"},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Zhang, C., Wang, G., Zhao, J., Gao, P., Lin, J., and Yang, H. (2017, January 20\u201321). Patient-specific ECG classification based on recurrent neural networks and clustering technique. Proceedings of the 2017 13th IASTED International Conference on Biomedical Engineering (BioMed), Innsbruck, Austria.","DOI":"10.2316\/P.2017.852-029"},{"key":"ref_14","unstructured":"Kim, K. (2021, December 01). Arrhythmia Classification in Multi-Channel ECG Signals Using Deep Neural Networks. Available online: http:\/\/www2.eecs.berkeley.edu\/Pubs\/TechRpts\/2018\/EECS-2018-80.html."},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"3028","DOI":"10.1161\/01.CIR.99.23.3028","article-title":"Characterization of different subsets of atrial fibrillation in general practice in France: The ALFA study. The college of French cardiologists","volume":"99","author":"Levy","year":"1999","journal-title":"Circulation"},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"657","DOI":"10.1016\/j.ahj.2004.06.032","article-title":"Asymptomatic atrial fibrillation: Demographic features and prognostic information from the atrial fibrillation follow-up investigation of rhythm management (AFFIRM) study","volume":"149","author":"Flaker","year":"2005","journal-title":"Am. Heart J."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"48","DOI":"10.1093\/eurheartj\/17.suppl_C.48","article-title":"Follow-up of atrial fibrillation: The initial experience of the Canadian registry of atrial fibrillation","volume":"17","author":"Kerr","year":"1996","journal-title":"Eur. Heart J."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"2933","DOI":"10.1161\/CIRCULATIONAHA.111.069450","article-title":"Symptoms and functional status of patients with atrial fibrillation: State of the art and future research opportunities","volume":"125","author":"Rienstra","year":"2012","journal-title":"Circulation"},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"1250","DOI":"10.15585\/mmwr.mm6936a4","article-title":"Delay or avoidance of medical care because of COVID-19-related concerns\u2014United States, June 2020","volume":"69","author":"Czeisler","year":"2020","journal-title":"Morb. Mortal. Wkly. Rep."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"886","DOI":"10.1377\/hlthaff.2021.00250","article-title":"Worse cardiac arrest outcomes during The COVID-19 pandemic in Boston can be attributed to patient reluctance to seek care","volume":"40","author":"Sun","year":"2021","journal-title":"Health Aff."},{"key":"ref_21","doi-asserted-by":"crossref","unstructured":"He, K., Zhang, X., Ren, S., and Sun, J. (2017). Deep residual learning for image recognition. arXiv Prepr., Available online: https:\/\/arxiv.org\/abs\/1512.03385.","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref_22","unstructured":"Howard, A.G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T., Andreetto, M., and Adam, H. (2017). MobileNets: Efficient convolutional neural networks for mobile vision applications. arXiv Prepr., Available online: https:\/\/arxiv.org\/abs\/1704.04861."},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"He, Z., Zhang, X., Cao, Y., Liu, Z., Zhang, B., and Wang, X. (2018). LiteNet: Lightweight neural network for detecting arrhythmias at resource-constrained mobile devices. Sensors, 18.","DOI":"10.3390\/s18041229"},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"515","DOI":"10.1109\/JBHI.2019.2911367","article-title":"LSTM-based ECG classification for continuous monitoring on personal wearable devices","volume":"24","author":"Saadatnejad","year":"2020","journal-title":"IEEE J. Biomed. Health Inform."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"e17037","DOI":"10.2196\/17037","article-title":"A lightweight deep learning model for fast electrocardiographic beats classification with a wearable cardiac monitor: Development and validation study","volume":"8","author":"Jeon","year":"2020","journal-title":"JMIR Med. Inform."},{"key":"ref_26","unstructured":"TensorFlow Lite (2021, October 21). Model Optimization. Available online: https:\/\/www.tensorflow.org\/lite\/performance\/model_optimization."},{"key":"ref_27","unstructured":"Cheng, Y., Wang, D., Zhou, P., and Zhang, T. (2017). A Survey of Model Compression and Acceleration for Deep Neural Networks. arXiv Prepr., Available online: https:\/\/arxiv.org\/abs\/1710.09282."},{"key":"ref_28","first-page":"40","article-title":"Recent R&D trends for lightweight deep learning","volume":"34","author":"Lee","year":"2019","journal-title":"Electron. Telecommun. Trends"},{"key":"ref_29","unstructured":"Han, S., Mao, H., and Dally, W.J. (2015). Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding. arXiv Prepr., Available online: https:\/\/arxiv.org\/abs\/1510.00149."},{"key":"ref_30","unstructured":"Rastegari, M., Ordonez, V., Redmon, J., and Farhadi, A. (2021, December 01). XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks. Available online: https:\/\/arxiv.org\/abs\/1603.05279."},{"key":"ref_31","unstructured":"Ullrich, K., Meeds, E., and Welling, M. (2017). Soft Weight-Sharing for Neural Network Compression. arXiv Prepr., Available online: https:\/\/arxiv.org\/abs\/1702.04008."},{"key":"ref_32","unstructured":"Hinton, G., Vinyals, O., and Dean, J. (2015). Distilling the Knowledge in a Neural Network. arXiv Prepr., Available online: https:\/\/arxiv.org\/abs\/1503.02531."},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"711","DOI":"10.1111\/ggi.13716","article-title":"Social determinants of the association among cerebrovascular disease, hearing loss and cognitive impairment in a middle-aged or older population: Recurrent neural network analysis of the Korean Longitudinal Study of Aging (2014\u20132016)","volume":"19","author":"Lee","year":"2019","journal-title":"Geriatr. Gerontol. Int."}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/22\/5\/1776\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,10]],"date-time":"2025-10-10T22:26:35Z","timestamp":1760135195000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/22\/5\/1776"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,2,24]]},"references-count":33,"journal-issue":{"issue":"5","published-online":{"date-parts":[[2022,3]]}},"alternative-id":["s22051776"],"URL":"https:\/\/doi.org\/10.3390\/s22051776","relation":{},"ISSN":["1424-8220"],"issn-type":[{"value":"1424-8220","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,2,24]]}}}