2025-06-04
2025-04-30
Manuscript received April 21, 2025; revised June 3, 2025; accepted August 11, 2025; published October 17, 2025.
Abstract—The development of technology improves the early detection of disease and diagnosis in the field of medical. Most of the people suffer with heart disease depending on the age factor and is millions of death cases are recorded due to heart attacks. In this paper Heart Rate Variability (HRV) is measured for early detection of Cardio Vascular Disease (CVD). HRV analyse the heartbeats and evaluated the time intervals between the consecutive heart beats. The CPSC 2018 dataset is considered in this paper to evaluate the Electrocardiogram (ECG) signals. Initially the R-R intervals are evaluated. The key features of the HRV like spatial, frequency, Poincare are extracted. Further features are extracted using deep learning model Efficient-Net b0 (ENb0) and classified for identification of heart disease. A Transformer Encoder (TE) layer is added to ENb0 to improve the performance. The Receiver Operating Characteristic (ROC) curve and confusion matrix is achieved using the deep learning classification model. The parameters like accuracy, precision, recall and F1-Score are evaluated using the proposed deep learning model and compared with other techniques. The ENb0 model achieves better efficiency in parameters evaluated and is effective in early detection of CVD and helps to support the research in the field of medical. Keywords—cardiovascular disease, Electrocardiogram (ECG) signals, R-peak detection, Heart Rate Variability (HRV) analysis, Efficient-Net b0 Cite: Munji Gayathri and Suresh Chittineni, "Deep Learning Model for Automated Heart Disease Diagnosis Using ECG Signal," Journal of Image and Graphics, Vol. 13, No. 5, pp. 549-560, 2025. Copyright © 2025 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC-BY-4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.