Arrhythmia warning algorithm

• Main Author: Tan, Calvin Jun Hao
• Other Authors: Vidya Sudarshan
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Computer science and engineering
• Online Access: https://hdl.handle.net/10356/166715
• Institution: Nanyang Technological University

Arrhythmias are cardiac conduction dysfunctions that cause abnormal heartbeats, which can be detected through electrocardiographic (ECG) signals. However, visually analyzing ECG signals can be challenging and time-consuming due to the low amplitudes involved. To address this issue, an automated system can be utilized in the clinical context to improve the speed and accuracy of arrhythmia diagnosis. This report presents a proposed ensemble model comprising convolutional neural network (CNN) and Support Vector Machine (SVM) to automate the identification of arrhythmias in ECG data. The model's hyperparameters will be optimized to achieve the best performance, with solutions presented for any problems encountered in the methods chapter. The report also includes the demonstration and assessment of the model's performance on several datasets.