Title: AI-enhanced ECG monitoring for arrhythmia detection using semantic LinkNet deep neural network

Authors: Soumen Ghosh; Satish Chander

Addresses: Birla Institute of Technology, Mesra, Ranchi, Jharkhand, 835215, India ' Birla Institute of Technology, Mesra, Ranchi, Jharkhand, 835215, India

Abstract: Cardiac disease is a leading cause of death globally, making early detection crucial for improving diagnosis and treatment. An electrocardiogram (ECG) is a key tool for monitoring heart activity, especially in identifying cardiac arrhythmia, which involves irregular heartbeats - either too fast or too slow. Automated detection systems are vital for improving detection accuracy. This work introduces a deep learning-based semantic LinkNet classifier to detect arrhythmias. The process begins by preprocessing the ECG signal including DC drift cancellation, normalisation, and low-pass filtering. QRS detection is then performed followed by feature extraction. The semantic LinkNet classifier is applied to classify normal and abnormal ECG signals. The system's performance is evaluated using the MIT-BIH database, with metrics such as accuracy, sensitivity, specificity, precision, recall, and F-measure. The proposed method's results, after filtering and QRS detection, are compared to existing models. The outcomes demonstrate that the semantic LinkNet-based system outperforms other approaches, showing superior detection rates in identifying arrhythmias.

Keywords: electrocardiogram; ECG; cardiac arrhythmia; QRS detection; DC drift; normalisation; LPF filter; MIT-BIH; Sematic LinkNet classifier.

DOI: 10.1504/IJICA.2025.145033

International Journal of Innovative Computing and Applications, 2025 Vol.15 No.2, pp.93 - 101

Received: 18 Oct 2024
Accepted: 04 Jan 2025
Published online: 17 Mar 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article