LSTM-Based ECG Classification for Continuous Monitoring on Personal Wearable Devices

By Saeed Saadatnejad, Mohammadhosein Oveisi, Matin Hashemi. IEEE journal of biomedical and health informatics, 2018

368 citations recorded by the source corpus.

<italic>Objective</italic>: A novel electrocardiogram (ECG) classification algorithm is proposed for continuous cardiac monitoring on wearable devices with limited processing capacity. <italic>Methods</italic>: The proposed solution employs a novel architecture consisting of wavelet transform and multiple long short-term memory (LSTM) recurrent neural networks (see <xref ref-type="fig" rid="fig1">Fig. 1</xref>). <italic>Results</italic>: Experimental evaluations show superior ECG classification performance compared to previous works. Measurements on different hardware platforms show the proposed algorithm meets timing requirements for continuous and real-time execution on wearable devices. <italic>Conclusion</italic>: In contrast to many compute-intensive deep-learning based approaches, the proposed algorithm is lightweight, and therefore, brings continuous monitoring with accurate LSTM-based ECG classification to wearable devices. <italic>Significance</italic>: The proposed algorithm is both accurate and lightweight. The source code is available online at <uri>http://lis.ee.sharif.edu</uri>.

Subfields: Biomedical Engineering, Cardiology and Cardiovascular Medicine, Experimental and Cognitive Psychology

View Paper: https://www.semanticscholar.org/paper/4892fa0eb89fb817b9436eee0b568713d05f2a3e