Application of Deep Learning for Cardiac Arrhythmia Classification Based on ECG Signals

Gabriela Septiani Simbolon (Information Systems Study Program, Faculty of Science and Technology, Prima Indonesia University, Medan)
Gresia Cesilia Sirait (Information Systems Study Program, Faculty of Science and Technology, Prima Indonesia University, Medan)
Sarah Theresia Aruan (Information Systems Study Program, Faculty of Science and Technology, Prima Indonesia University, Medan)
Rivaldo Robertus Turnip (Information Systems Study Program, Faculty of Science and Technology, Prima Indonesia University, Medan)
Jepri Banjarnahor (Information Systems Study Program, Faculty of Science and Technology, Prima Indonesia University, Medan)
Mardi Turnip (Information Systems Study Program, Faculty of Science and Technology, Prima Indonesia University, Medan)

Journal of Computer Networks, Architecture and High Performance Computing
Vol. 8 No. 3 (2026): Research Paper July 2026

Cardiac arrhythmia is a dangerous heart rhythm disorder, so early detection is crucial for effective treatment. Manual ECG (Electrocardiogram) analysis is less accurate, while deep learning can detect arrhythmias more quickly and precisely. The proposed algorithm uses a deep learning Convolutional Neural Network (CNN) model for arrhythmia classification. The model is trained on labeled normal and arrhythmia ECG datasets to recognize important patterns in sequential data. The ECG data is obtained from PhysioNet, which provides thousands of labeled recordings for training and testing. Additional clinical data from hospitals/clinics can be included for further validation with patient consent according to ethical protocols. The expected result is that this system can detect arrhythmias with high accuracy and optimal sensitivity. The benefits are to improve the quality of healthcare services and reduce the risk of serious complications.