Multiscale Entropy Algorithm And Its Application In Heart Disease Identification

In today’s increasingly fast pace of life,cardiovascular disease has always been an important factor affecting people’s health.Cardiovascular disease has overtaken cancer for many years,ranking first among all diseases.Especially in China,with the aging of the population and the improvement of living conditions,the incidence of heart disease has been on the rise.Electrocardiogram is a powerful tool for diagnosing cardiovascular disease.It records the changes in the electrical signals of the heart during beating and is a comprehensive manifestation of the heart’s electrophysiological activities on the human body.The ECG contains a wealth of cardiac function and pathological information,which can intuitively and accurately reflect the electrical activity characteristics of the heart and show the working state of the heart.But just by manually analyzing the ECG signals to diagnose heart disease,it not only increases the burden on doctors,it is also inconvenient for patients.Therefore,the automatic discrimination algorithm for heart disease has become a hot spot in the current research field of ECG signal processing.The main research contents of this article are as follows1 Aiming at the problems of various types of heart diseases and complex etiology,the research scope is limited to atrial fibrillation,ventricular arrhythmia and congestive heart failure,which are closely related to heart rate variability.The nonlinear heart rate variability analysis method was used to find the difference between the ECG signals of the three heart diseases and the ECG signals of healthy people.Finally,it was found through experiments that all three diseases will lead to a reduction in the complexity of ECG signals.2 Aiming at the characteristics of non-linear and unstable ECG signals,a multi-scale entropy-based algorithm for discriminating heart diseases was proposed.Firstly,the R wave peaks in each beat are pre-labeled by ECG signal pre-processing to form the RR interval sequence of the ECG signal.Then we use the empirical mode decomposition method to decompose the RR interval sequence of the ECG signal.After decomposition,different intrinsic solid-state functions are obtained.Each intrinsic solid-state function contains different information in the time domain of the original signal.Select the appropriate intrinsic solid-state function and calculate its multi-scale entropy.Finally,the multi-scale entropy value is used as the feature vector,and it is compared and distinguished by the support vector machine.And using the MIT-BIH ECG database data,where each data has been desensitization treatment by medical professionals’ comments.The MIT-BIH ECG database is widely used,which also facilitates the comparison of algorithm discrimination effects at the end of the article.Experimental results show that the accuracy of the automatic discrimination algorithm for the recognition of three types of heart diseases is more than 90%,and the accuracy rate is improved compared with other discrimination algorithms.