The Optimization Of Cross Fuzzy Entropy And Its Application In Heart Failure Detection

At present, cardiovascular disease has become one of the world’s leading cause of death in adults. The prevalence of cardiovascular disease incidence increased year by year in China. In recent years, the cost of our country in the treatment of cardiovascular diseases had rapid growth at an annual rate of18.6%, far more than China’s gross domestic product (GDP) growth rate, which indicates that China will face the huge medical crisis. Therefore, early detection and assessment of cardiovascular system function, so as to realize the effective intervention is an effective approach to solve this crisis.The research results show, the analysis of physiological information contained in ECG, heart sound and pulse wave signal in cardiovascular system can determine the health of the cardiovascular system. However, with regard to the cardiovascular system of this complexity, signal physiological single not enough to fully reflect the functional status of the cardiovascular system. With the development of modern medical equipment and sensor technology, synchronous measurement of multiple physiological signal has become a practical, thus by measuring the coupling relationship between multiple physiological signals and then get the method to system function state of cardiovascular provides a new idea for the research of the researchers in the early detection of cardiovascular disease. This paper mainly do the following work:(1) On the basis of cross fuzzy entropy introducing adjustment factor A, puts forward improving cross fuzzy entropy algorithm. Research shows that, the traditional entropy algorithm similarity criterion results in the actual application process is poor, This paper introduced the adjustment factor, defined more in line with the actual signal analysis similarity criterion-physical fuzzy membership function, and then put forward the improved mutual fuzzy entropy algorithm. By coupling the noise model and the coupled MIX (p) gives the simulation experiment model, λ is most effective between [0.5,1.5], the statistical stability of the improved cross fuzzy entropy algorithm and relatively consistent values were significantly greater than cross sample entropy and cross fuzzy entropy algorithm.(2) The simulation results show that the improved cross fuzzy entropy algorithm are compared to cross sample entropy and cross fuzzy entropy has higher statistical stability. Using cross sample entropy, cross fuzzy entropy and improved cross fuzzy entropy to Analysis coupling of the simulation cardiovascular time sequence. The simulation results show that:The cross entropy algorithm is applicable to analysis of coupling between cardiovascular system physiology signals; The cross entropy results of the improved cross fuzzy entropy algorithm has the minimal standard deviation, thus it has higher statistical stability.(3) The improved cross fuzzy entropy algorithm has a high value in the detection of heart failure through clinical experiments. Get heartbeat cycle-pulse wave propagation time sequence from heart failure patients and healthy people by clinical trials. Using cross sample entropy, cross fuzzy entropy and improved cross fuzzy entropy to Analysis coupling of the simulation cardiovascular time sequence. The experimental results show that:three kinds of entropy algorithm has the ability to distinguish between heart failure patients and healthy people, and the improved cross fuzzy entropy algorithm has the best statistical performance. the area under ROC curve reached0.963; when the threshold is set to1.12, improved cross entropy in heart failure diagnosis in specificity and sensitivity are93.3%and86.7%.