Cuff-less Blood Pressure measurement from Wireless ECG and PPG signals

Abstract

Continuous monitoring of blood pressure (BP) can control hypertension and cardiac diseases. Blood pressure measurement using cuff based technique provides intermittent measurement and inconvenient for long term monitoring. This work is focused on estimation of continuous blood pressure from electrocardiogram (ECG) and photoplethysmogram (PPG). The proposed work extracts ECG and PPG time domain features acquired through wireless hardware system. Using Support Vector Regression of machine learning, a light weight model for Blood Pressure estimation is trained. The proposed work is tested on wireless signals captured from 87 subjects using hardware device. According to the British Hypertension Society (BHS) standard, the proposed method achieves grade A in the estimation of systolic and diastolic pressure for wireless data. The values of mean error and standard deviation by proposed method are within limits of Association for the Advancement of Medical Instrumentation (AAMI) standards. The proposed work is helpful in wireless monitoring of patients to track the physiological conditions without interrupting their routine activities.