Identification of high blood pressure using support vector machine and time-domain heart rate variability from photoplethysmography

Abstract

Hypertension is one of the serious threats to human health by accelerating the cardiovascular disease. The way to prevent hypertension complications is to detect and prevent high blood pressure. This study aimed to identify hypertension using photoplethysmography (PPG) records. The method used time-domain Heart Rate Variability (HRV) from PPG. It used a Support Vector Machine (SVM) with Radial Basis Function (RBF). Variations of SVM-C and RBF gamma were conducted to find the good performance of identification. Using clinical data, the identification system performed with a training accuracy of 99.33 % and a testing accuracy of 71.75%. Best performing results occur when using SVM-C 100 with a gamma of 400,000.