Stratification of High-Risk Hypertensive Patients Using Hybrid Heart Rate Variability Features and Boosting Algorithms

Hypertension is a global challenge to the public health which can easily lead life-threatening vascular diseases unless control measures are adopted. Considering prevalence of and their fatality, early detection high-risk patients an important problem in present world. Heart rate variability (HRV) analysis be effective prognostic tool identify characteristics vulnerable patients, considering its reliability predicting sudden cardiac deaths. However, lies identifying tenuous differences HRV between low-risk at stage. With this motivation, we propose hybrid approach based on dual-tree complex wavelet packet transform (DTCWPT) linear time domain as well nonlinear signal extract multitudinous features. A key issue before such presence marked amount ectopic beats, addressed by using time-varying auto regressive (TVAR) technique. The features extracted from TVAR edited signals shortlisted minimum redundancy maximum relevance algorithm for efficient classifier modeling. Furthermore study, use cost-sensitive RUSBoost (CS-RUSBoost) handling class imbalance data. comparative performance evaluation CS-RUSBoost with RUSBoost, SMOTEBoost, asymmetric AdaBoost shows superior result G-mean 0.9352 F1 score 0.9347.