An Itelligent System for Diagnosis of Coronary Artery Disease with BP Neural Networks / 대한의료정보학회지

ABSTRACT

OBJECTIVE: In this paper, an intelligent system using BP neural networks (BPNN) is presented for early detection coronary artery disease (CAD). METHODS: Based on the four features of ECG signals and six basic parameters of patients, BPNN was built and trained. Especially the method which combined feature extraction and classification was discussed. RESULTS: The performance of the intelligent system has been evaluated in 20 samples. The test results showed that this system was effective in detecting CAD. The correct classification rate was about 90% for normal subjects and 100% for abnormal subjects. CONCLUSION: BPNN could quite accurately detect abnormal subjects. Because it is not expensive and noninvasive, it is fit to examine health of the elderly and has good application foreground.