Intelligent Phonocardiogram Analysis and Representation Tool

Cardiac auscultation is a highly sensitive, specific, cost effective and comfortable diagnosis technique for many cardiovascular diseases. Unfortunately, it is observed that the art of auscultation is mastered by an increasingly lower number of medical professionals. This paper presents a Matlab tool to support physicians in performing auscultation. This application enables real time signal acquisition using off-the-shelf sensors and performs several automatic annotation functions of heart sounds, such as noise contamination detection, segmentation into S1, S2 and S3, S2-split detection, murmur detection and classification, systolic time intervals measurement, contractility and stroke volume. These are related to the most pertinent clinical applications of this signal. Moreover, it can also be used for auscultation training.