A Novel Approach for Heart Murmurs Detection and Classification

In this paper, a novel approach for an automated algorithm design for cardiac murmurs detection and classification has been proposed. This approach is based on the measurement technique of the normalized time dependent sound energy decay used in room acoustics for characterization. Given a single heart sound recorded from the auscultation area of a patient, the algorithm computes the early time decay which can be used as indicator of possible pathology associated to corresponding patient heart valve. Results obtained from a restricted database of adult patients show that this novel approach demonstrates a potential promising as an alternative diagnostic tool to the traditional stethoscope for heart murmurs detection and classification.