Design and Development of PPG Signal Quality Assessment Using Hybrid HVD and PSoC 4 BLE Device for IoT Health Care Monitoring Applications

Abstract

This paper explains a framework for assessing the photoplethysmograpy signal quality that is enabled by the internet of things and is based on the first order predictor co-efficient (PC) of the differenced sensor signal. This framework minimises false alarms brought on by inaccurate measurements of PPG signals that are distorted by various motion and noise artifacts. Heart rate monitoring is frequently carried out using photoplethysmography, an easy and affordable optical sensing technique. The method consists of the following steps: 1) Pre-processing stage using Hybrid Hilbert Vibration Decomposition (HVD), which is a combination of EMD and HVD. 2) Feature Extraction stage using Levinson-Durbin recursive algorithm for first order predictor coefficient (FOPC) detection. 3) Classification stage using Support Vector Machine (SVM). 4) Transmission stage using Programmable system on chip (PSoC 4 BLE) Device. The major goals of this work are: realizing a PPG signal quality assessment approach for automatically categorizing the collected PPG signal as acceptable and undesirable, and also to design and develop a PPG monitoring framework using the PSoC 4 BLE Device, sensors, smart phone and cloud server utilizing wifi. Further, it also provides automatic alert system which recognizes various kinds of noises or abnormal conditions. Based on these abnormal conditions, the developed PPG-SQA system sends message to the doctor in case of any unexpected event. The PPG signals used in the suggested method come from the physionet databases. The suggested method achieves good results in classifying PPG as acceptable or unacceptable signals in terms of accuracy and susceptibility, according to the filtering strategies. The suggested IoT enabled architecture which demonstrates the transmission of PPG signals with acceptable quality has great prospective for developing energy-efficiency, accuracy and reliable smart phone based health monitoring applications.