Efficient Edge-AI Models for Robust ECG Abnormality Detection on Resource-Constrained Hardware

Zhaojing Huang; Luis Fernando Herbozo Contreras; Wing Hang Leung; Leping Yu; Nhan Duy Truong; Armin Nikpour; Omid Kavehei

doi:10.1007/s12265-024-10504-y

Efficient Edge-AI Models for Robust ECG Abnormality Detection on Resource-Constrained Hardware

J Cardiovasc Transl Res. 2024 Mar 12. doi: 10.1007/s12265-024-10504-y. Online ahead of print.

Authors

Zhaojing Huang^#¹, Luis Fernando Herbozo Contreras^#², Wing Hang Leung², Leping Yu², Nhan Duy Truong², Armin Nikpour³, Omid Kavehei²

Affiliations

¹ School of Biomedical Engineering, The University of Sydney, NSW 2008, Sydney, Australia. zhaojing.huang@sydney.edu.au.
² School of Biomedical Engineering, The University of Sydney, NSW 2008, Sydney, Australia.
³ Department of Neurology, Royal Prince Alfred Hospital, and Central Clinical School, The University of Sydney, NSW 2006, Sydney, Australia.

^# Contributed equally.

PMID: 38472722
DOI: 10.1007/s12265-024-10504-y

Abstract

This study introduces two models, ConvLSTM2D-liquid time-constant network (CLTC) and ConvLSTM2D-closed-form continuous-time neural network (CCfC), designed for abnormality identification using electrocardiogram (ECG) data. Trained on the Telehealth Network of Minas Gerais (TNMG) subset dataset, both models were evaluated for their performance, generalizability capacity, and resilience. They demonstrated comparable results in terms of F1 scores and AUROC values. The CCfC model achieved slightly higher accuracy, while the CLTC model showed better handling of empty channels. Remarkably, the models were successfully deployed on a resource-constrained microcontroller, proving their suitability for edge device applications. Generalization capabilities were confirmed through the evaluation on the China Physiological Signal Challenge 2018 (CPSC) dataset. The models' efficient resource utilization, occupying 70.6% of memory and 9.4% of flash memory, makes them promising candidates for real-world healthcare applications. Overall, this research advances abnormality identification in ECG data, contributing to the progress of AI in healthcare.

Keywords: Abnormality identification; ECG data; Edge devices; Generalization; Performance evaluation; Robustness; Simple network.