Research On The Method Of Predicting Cardiovascular And Cerebrovascular Health Sudden Risk Based On Pulse Wave

Cerebrovascular disease has high sudden,great harm and short effective treatment time.Accurate prediction of the onset time of cardiovascular and cerebrovascular disease can strive for more time for the treatment of cardiovascular and cerebrovascular disease and reduce the harm caused by cardiovascular and cerebrovascular disease.The traditional medium and long-term prediction methods of cardiovascular and cerebrovascular diseases,whether it is invasive collection or ECG monitoring,need the help of complex clinical instruments.Because of this,the traditional medium and long-term prediction of cardiovascular and cerebrovascular diseases has the disadvantages of cumbersome prediction process and poor real-time performance.Therefore,it is particularly important to construct the risk prediction method of cardiovascular and cerebrovascular health emergencies from the state mutation of cardiovascular and cerebrovascular system before the onset of cardiovascular and cerebrovascular diseases.This paper introduces a method of cardiovascular and cerebrovascular health emergency risk prediction based on pulse wave.The mutation degree of pulse wave was identified by comparing the changes of characteristic parameters of pulse wave before and after treatment,and the risk level of cardiovascular and cerebrovascular health emergencies was predicted according to the mutation degree of pulse wave.Aiming at the problem of inaccurate identification of pulse wave mutation degree caused by different motion states in the process of sudden risk prediction,this paper introduces three-dimensional attitude information and takes the root mean square of six axis acceleration in three-dimensional attitude information as the basis of motion state definition,so as to realize the independent prediction of sudden risk under various motion states.Aiming at the problem of large number and redundancy of pulse wave feature parameters in the process of burst risk prediction,a burst risk prediction model based on Relief F mrmr hybrid feature selection algorithm is constructed.The original subset of pulse wave feature parameters is selected.The original subset contains three feature parameters related to amplitude information and three feature parameters related to time information.The burst risk operator is constructed according to the feature weight of the selected pulse wave characteristic parameters.The burst risk operator is used to identify the mutation degree of pulse wave,and then predict the level of burst risk,which reduces the number and redundancy of characteristic parameters needed in the process of burst risk prediction.According to the Q-T dispersion of ECG,the evaluation standard of emergency risk was established.The effectiveness of the method is verified by the experiment of emergency risk prediction.In the experiment of sudden risk prediction,the accuracy rate of movement state recognition is 100%,and the overall accuracy rate of sudden risk prediction results is88.89%.The accuracy rates of sudden risk prediction results under three kinds of movement states,namely,sitting in,walking and running,are 91.67%,91.67% and83.33% respectively.