Research On Prediction Of Almost Periodic Electrophysiological Signal Based On Improved Markov Model

The data shows that the world's heart disease and mortality are high,which is one of the primary causes of life threatening.The electrical and magnetic signals of the heart are measured data showing the periodic electrophysiological activities of the human heart muscle cells on the body surface.They are rich in massive information.They are important signals for studying the internal physiological status of the human heart mechanism,and are also used to diagnose or evaluate heart disease Therefore,the research and analysis of cardiac electrical and magnetic signals have always been the focus of research at home and abroad.Electrophysiological signals are very complex,and they affect each other.From a strict mathematical definition,it can be seen that this type of signal shows a non-periodic and non-stationary random process,but if we come from the organ's own physiological function Consider,this kind of signal shows more obvious periodicity.Reasonable and effective information feature extraction,analysis,and even development trend prediction of the electrophysiological signals obtained by the measurement are beneficial to make a more comprehensive and accurate judgment on the internal physiological information of the human body.Considering the periodic reproducibility of the almost periodic electrophysiological signal,this paper studies the prediction method to study the changing trend of the measured electrophysiological signal.The method of forecasting is the most widely used method in data mining.The process of forecasting data is to estimate the future data based on the analysis of past data.Forecasting can be used as a reference for practical applications.With the advent of the era of big data,more and more scholars have begun to attach importance to the study of data prediction methods and their applications.Based on the characteristics of the almost periodic electrophysiological signals,this paper takes the typical almost periodic electrophysiological signals: ECG and magnetomagnetic signals as the research objects,analyzes the Markov property of the measured electrophysiological data,and studies and improves its weighted Markov model,A method for predicting almost periodic electrophysiological signals is proposed.The weighted Markov prediction models of different state division methods are used to predict and analyze the time series of ECG signals.The research results show that the weighted Markov prediction model using the ordered clustering methodfor state division is better than the sample mean mean square error method.The divided weighted Markov prediction model has higher accuracy.In addition,for the problem of the limited location of the superconducting quantum interferometer in collecting the magnetic signal of the heart,a prediction method of the weighted Markov prediction model as a data enrichment method for the magnetic heart on the spatial sequence is proposed.The prediction results verify the feasibility of the method.