An evaluating method for autonomic nerve activity by means of estimating the consistency of heart rate variability...

The imbalance of autonomic nervous system (ANS) has a close relationship to cardiac mortality. The noninvasive assessment of ANS is of great importance and remains challenging. A new method, characterizing the consistency of short-term nonlinear indexes of heart rate variability (HRV) and QT variability (QTV), is proposed and validated. Holter records from two databases in Telemetric and Holter ECG Warehouse were used, 43 records from one database (named ESRD) as typical subjects of ANS dysfunction and 118 records from the other database (named Normal) as normal controls. The consistency of HRV and QTV was characterized by estimating mutual information (MI) of paired short-term recurrence quantification analysis (RQA) indexes in resting state. The influence of physiological differences on MIs of paired RQA indexes in Normal was investigated as well. Results showed that there were significant differences in MI-DET (day: 0.283 ± 0.070 versus 0.133 ± 0.055 and night: 0.258 ± 0.061 versus 0.117 ± 0.055) and MI-LAM (day: 0.439 ± 0.053 versus 0.293 ± 0.073 and night: 0.361 ± 0.079 versus 0.241 ± 0.087) between Normal and ESRD, much reduced consistency in ESRD. For MI-DET in Normal, sex had no influence, and there was age related alternations by day but not at night. There was no influence of sex and age on MI-LAM in Normal. The sensitivity, specificity, and total accuracy for discriminating Normal and ESRD were 88.37%, 95.76%, and 93.79%, respectively. The proposed measures are shown to have the advantage in reducing the influence of physiological differences and highlighting the pathological influence, providing a promising method to find clinical application for noninvasive assessment of ANS state.