Detection Of Coronary Artery Disease Based On Combined Analysis Of Multi-Channel Heart Sound Signals

For a long time,coronary artery disease has been the leading cause of death in cardiovascular disease globally and is still increasing at an alarming rate every year.Noninvasive,nondestructive and accurate detection of coronary artery disease is an effective means to prevent and reduce its harm in the early stage.Coronary angiography is widely regarded as the gold standard for detecting coronary artery disease.But it is not suitable as a routine examination method for early screening due to its invasive and high price defects.Medical research has confirmed that when blood in the coronary artery flows through the stenosis of a blood vessel,it will hit the vessel wall and form turbulence,which will lead to noise in the heart sound.Therefore,non-invasive heart sound signal detection and analysis is expected to become a cost-effective screening tool to realize the noninvasive and nondestructive detection of coronary artery disease,and provide more reference basis for patients with coronary artery disease to see a doctor as soon as possible,assist doctors to make preliminary diagnosis before coronary angiography,and dynamically evaluate the treatment plan.Previous studies have shown that the information contained in single channel heart sound signal can be used to detect coronary artery disease,but the accuracy of coronary artery disease classification is not satisfactory.Inspired by heart sound auscultation and electro-mechanical coupling,this study is committed to exploring the beneficial effect of multi-channel heart sound signals on the detection of coronary artery disease.and looks forward to providing reference basis for doctors’ diagnosis.The multi-channel heart sound signal data is complex,and an effective joint analysis method is very important.At the same time,in order to ensure the stability and reliability of the data,the multi-channel heart sound signal synchronous acquisition technology is also essential.However,there are few researches on the analysis method and acquisition technology of multi-channel heart sound signals.In view of the above problems,the main work and innovations of this thesis are as follows:(1)In the clinical environment,collect the 5-channel heart sound signals of patients with suspected coronary artery disease before coronary angiography,and construct a multi-channel heart sound signal data set for coronary artery disease detection.(2)In order to fully mine the disease information in multi-channel heart sound signals,a multi-domain feature joint analysis method of multi-channel heart sound signals is proposed.The multi-domain features of each channel heart sound signal such as time domain,frequency domain and entropy domain are extracted,and the optimal feature set is constructed through feature selection and sent to the classifier to identify coronary artery disease.The results show that the highest accuracy of multi-channel heart sound signal multi-domain feature joint analysis in identifying coronary artery disease is 87.41%,which is higher than the optimal result of single channel heart sound signal 83.02%.It is confirmed that the combined analysis of multi-channel heart sound signals can provide more disease information related to coronary artery disease than single channel heart sound signals.(3)In order to obtain the coupling information between multi-channel heart sound signals,a multi-channel heart sound signal coupling feature fusion analysis method is proposed to explore the influence of coronary artery stenosis.The results show that the recognition effect of coupling feature fusion analysis between two channel heart sound signals on coronary artery disease is similar to that of multi-channel heart sound signal multi-domain feature joint analysis,and its accuracy,sensitivity and specificity are 87.95%,85.70%and 90.06%respectively.The fusion of coupling features and multi-domain features can further improve the recognition effect of coronary artery disease,and its accuracy,sensitivity and specificity can be improved to 92.00%,86.77%and 96.32%respectively.It is confirmed that the coupling characteristics between two channel heart sound signals can provide effective supplementary information different from multi-domain characteristics for the detection of coronary artery disease.(4)Aiming at the problem that the number of acquisition parts and channels is limited in the traditional heart sound signal acquisition technology,which affects the comprehensive capture of disease information,a new non-contact heart sound signal acquisition system is designed by using the principle of optical fiber vibration measurement.The system uses the characteristics of laser focusing to improve the number of heart sound signal acquisition channels.Based on the optical fiber vibration measuring heart sound signal acquisition system,the synchronous acquisition and verification system of optical center vibration,ECG,heart sound and respiratory signals is built,and the verification experiment is designed to verify the relationship between the optical center vibration signal collected by the system and the traditional heart sound signal.The verification results show that the optical center vibration signal is consistent with the traditional heart sound signal in time sequence;When the physiological state changes,the change of optical center vibration signal is consistent with the traditional heart sound signal.The consistency between the signals collected by the designed system and heart sound signals is preliminarily verified,which provides theoretical support and technical basis for the construction of a new sensor array that can accurately collect heart sound signals in more parts of the body surface.This study further explores the potential value of multi-channel heart sound signals in coronary heart disease detection.The proposed algorithm improves the accuracy of coronary heart disease detection using heart sound signals to 92%,which can provide a basis for doctors’diagnosis.At the same time,the heart sound signal acquisition equipment based on the principle of optical fiber vibration measurement is developed to lay the foundation for collecting more channel heart sound signals.