Feature Extraction and Abnormality Detection in Autonomic Regulation of Cardiovascular System

Abstract

The objective of this study is to develop an efficient methodology for classifying patients suffering any type of blood pressure dysregulation from healthy subjects. Four features of malfunctions in blood pressure regulation are introduced, and a criterion is proposed for each feature to evaluate and distinguish patients from healthy subjects. The evaluated features are based on the analysis of difference between data related to healthy subjects and those collected from patients. The proposed criteria are implemented on a group of healthy and patient subjects by collecting their systolic blood pressure (SBP) and their heart rate (HR) time series. The proposed method is applied on three different groups of subjects each containing four healthy and eleven patients. It is shown that the algorithm properly detects the status of all fifteen subjects in one group and fourteen subjects in two groups. The results obtained indicate that the selected features have remarkable capability in detection of blood pressure dysregulation.