Evaluation Of Autonomic Nervous System Function And Detecting Apnea Based On Heart Rate Variability

Objective: Frequent cessations of respiration can greatly increase the prevalence rate of arrhythmia.It has been confirmed that cardiac activity is regulated by autonomic nervous system(ANS).And heart rate variability(HRV)is widely used as a method to evaluate the function of ANS.Therefore,we analyzed how apnea can affect the balance and normal function of ANS using short-term HRV indices.Clinically,apnea that cannot be corrected in time can lead to imbalance of cardiac function and hypoxia of body organs and tissues,so the prediction of apnea is getting more and more attention.And we intend to use the HRV indices to establish many machine learning models to detect apnea.Method: 1.We collected data of forty-five healthy subjects under normal breathing and 10 times apnea.Meanwhile,the data of thirty-six patients with normally and 10 times apnea were downloaded from PhysioNet database.We calculated short-term HRV indices of subjects in normal respiratory and apneic states,respectively.2.Many machine learning models were established to detect apnea based on the HRV indices calculated in Study1,.Results: Compared with normal respiratory state,respiration cease would lead to the values of the Mean-RR,nLF,LF/HF,and ?1 were significantly increase whereas the values of rMSSD and nHF were significantly decrease.The accuracy of using HRV indices to predict apnea can reach 94.62%,sensitivity can reach 93.68%,and specificity can reach 95.93%.Conclusion: Cessations of respiration would lead to an imbalance in function of ANS,as well as an increase in fractal characteristics of the heart.These changes in physiological state are likely to induce and cause the occurrence of arrhythmia.The results of the machine learning models prove that the HRV indices in detecting apnea has good accuracy and stability,and provides a good guarantee for detection of apnea.