A clinical decision and support system with automatically ECG classification in telehealthcare

Abstract

Telehealthcare is a global trend affecting clinical practice in the world. With the progress and development of telecommunication technologies, telecom facilities have afforded telehealthcare a new approach for chronic disease management. The electrocardiogram (ECG) is commonly used to detect abnormal heart rhythms and to investigate the cause of heart abnormalities. To reduce the cardiologists' loading and to provide a continuously telehealthcare, we developed a clinical decision and support system (CDSS) with automatic recognition of the ECG in real-time analysis. In addition, we adopted the approach of noise reduction and feature extraction for support vector machine (SVM) implementation with automatic learning algorithms. The automatic interpretation of ECG could provide assistance to physicians in decision-making, especially with large volumes of data. According to the preliminary results of automatic classification models, we acquired 88.4% sensitivity, for noise detection model, 85.9% specificity for sinus classification model and 89.1% sensitivity for disease classification model, respectively. However, it is not reliable enough to obviate the need for physician's diagnosis and confirmation. We should put much effort on enhancing the performance of ECG interpretation in the future.