ABSTRACT
In this chapter, we present a new noninvasive coronary artery disease (CAD) pre-screening technique using multi-channel phonocardiography (PCG) and electrocardiography (ECG). After a brief overview of the state-of-the-art in PCG, we present a multichannel PCG and ECG instrument. With the aid of a background-noise microphone integrated into each digital stethoscope, adaptive signal processing methods are then proposed to suppress the environmental noise to improve the integrity of the recorded heartbeat signal. Finally, we investigate important factors that can impact the performance of a neural network-based CAD classifier using heart auscultation. Among them, the integrity of the heartbeat signal plays a key role.
