Computer-aided diagnostic system for the prediction of ventricular tachycardia using ECG signals
Ventricular Tachycardia (VT) is a type of arrhythmia caused by disturbances in the electrical activities originating in the ventricles of the heart. Although it is usually harmless when a VT episode occurs for a short period of time, it is often linked to various underlying cardiac conditions such a...
Saved in:
主要作者: | |
---|---|
其他作者: | |
格式: | Final Year Project |
語言: | English |
出版: |
Nanyang Technological University
2024
|
主題: | |
在線閱讀: | https://hdl.handle.net/10356/175032 |
標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
機構: | Nanyang Technological University |
語言: | English |
總結: | Ventricular Tachycardia (VT) is a type of arrhythmia caused by disturbances in the electrical activities originating in the ventricles of the heart. Although it is usually harmless when a VT episode occurs for a short period of time, it is often linked to various underlying cardiac conditions such as Coronary Artery Disease (CAD) or Valvular Heart Diseases. Prolonged and sustained VT episodes could also potentially lead to life threatening conditions such as Ventricular Fibrillation (VF).
Anomalies in cardiac rhythms during VT or other forms of arrhythmias can be detected by a medical professional through the use of an electrocardiogram (ECG). However, in today’s clinical environment, it can be time-consuming and challenging due to its intermittent nature and variability in morphology. Therefore, the aim of this study is to develop a Computer-Aided Diagnostic System (CADS) to capture ECG changes and identify particular VT rhythms in order for timely intervention and necessary treatments.
This report presents a comparison of the use of Convolutional Neural Network (CNN) and an ensemble approach of Discrete Wavelet Transform (DWT), CNN and Support Vector Machine (SVM) for the identification of abnormal beats occurring in VT episodes. The proposed ensemble approach yielded promising results of an accuracy of 92.3% and precision of 89.1% compared to an accuracy of 85.7% and precision of 87.3% for the CNN model. |
---|