Heart sound segmentation

• Main Author: Peh, Qian Hui
• Other Authors: Soh Cheong Boon
• Format: Final Year Project
• Language: English
• Published: 2018
• Subjects: DRNTU::Engineering::Electrical and electronic engineering
• Online Access: http://hdl.handle.net/10356/75810
• Institution: Nanyang Technological University

Cardiovascular disease is the leading cause of deaths globally. Despite having rapid medical technology advancements, the number of deaths related to cardiovascular disease has also increased over the years. In order to alleviate the problem, it is important to detect possible symptoms of heart diseases before the onset of an attack. Currently, there are various scans in the medical industry that can check the conditions of the heart, but such scans are usually invasive and expensive. Ideally, the use of relatively inexpensive and non-invasive tests is preferred, and the traditional method is via heart auscultation. Heart auscultation is done by listening to the quasi-periodic acoustic sounds that are produced by the heart, using a stethoscope. However, crucial heart sounds might not be heard as accuracy of the diagnosis is dependent on the practitioner’s individual experience, thus, the development of phonocardiography is a more reliable method to acquire and record heart sounds. This objective of this project is to create a MATLAB program that utilizes signal processing enveloping techniques to perform the automatic detection and segmentation of first heart sounds(S1), second heart sounds (S2) and heart cycle boundaries. Different enveloping techniques such as Shannon Energy, Shannon Entropy and Hilbert Transform were used to analyse its accuracies and efficiencies in achieving heart sound segmentation. These techniques primarily create an envelope of the signal to find the start and stop points of both S1 and S2. The most suitable threshold value is determined by finding the lowest Root Mean Square (RMS) Error calculated over a threshold range from 0.01 to 0.2. and used to determine the best technique amongst the other detection techniques. With the studies done for the project, further research or even a possible technology and product can be developed to improve the current techniques used and save more lives. In this report, it details the whole project, from the introduction, literature review, methodology, results, analysis, conclusion and future recommendations.