Heart cycles using heart sounds

• Main Author: Ng, Jia Long
• Other Authors: Soh Cheong Boon
• Format: Final Year Project
• Language: English
• Published: 2016
• Subjects: DRNTU::Engineering
• Online Access: http://hdl.handle.net/10356/67659
• Institution: Nanyang Technological University

What happens when Singapore goes towards an ageing population? Have you ever wondered what if one day, Singapore is short of healthcare professionals? With these two questions in hand, there is a demand in improving the medical technology in Singapore. Heart auscultation is an old but gold technique in finding heart valve abnormalities. It involves the listening of acoustic sounds that are produced from our heart. However, differences in individual’s hearing abilities may result in missing crucial acoustic sounds. Hence, the first step is to implement and study a reliable method in finding heart sound segmentation. Then, further diagnostic technology can be researched using these segmentation data as it provides an exact location of the heart sound in each heart cycle. The main feature is the use of Matlab to do signal processing of the heart sounds. Here, different methods were used to study its efficiencies in acquiring the heart sound segmentation. With these studies, further research or even an end product can be in sight, improving the current heart auscultation technique used. In this report, it describes the whole process of the project, from the introduction, implementation, analysis and conclusions with future recommendations. A heart sound is a quasi periodic signal, where each cycle of heart beat can be identified using the lub-dub called first (S1) and second (S2) heart sounds respectively. This can be visually observed using phonocardiography (PCG), one of the simplest heart diagnostic tool. PCG is the method of using the digitally recorded waveform of heart sounds we hear with stethoscope. Many of the diagnostic procedures require separating of the heart cycles. This project involves the development of signal processing algorithms to first detect S1 and S2 heart sounds and then determine the heart cycles.