Abstract

Info

Author(s):: ZHANG Yue1; CHEN Zhencheng1; LIANG Yongbo2; ZHU Jianming2; 1. School of Electronic Engineering and Automation, Guilin University of Electronic Technology, Guilin 541004, China 2. School of Life and Environmental Sciences, Guilin University of Electronic Technology, Guilin 541004, China

Keywords:: Keywords: atrial fibrillation photoplethysmography BP neural network random forest support vector machine

Abstract:: Abstract: A photoplethysmography (PPG)-based method for identifying atrial fibrillation is proposed to detect atrial fibrillation easily and quickly. Firstly, the pulse waves of atrial fibrillation and those in health status are compared and analyzed. Secondly, based on the analysis results, 6 types of characteristic parameters related to atrial fibrillation are extracted from PPG data as the input of classifiers. Finally, 3 classifiers, namely support vector machine, BP neural network and random forest algorithm, are used to establish the model for atrial fibrillation identification. The identification accuracies of the 3 models are 89.1%, 92.3% and 95.2%, respectively. The experimental results show that the atrial fibrillation identification method based on PPG has a high accuracy, especially when using random forest algorithm as the classifier. Meanwhile, the proposed detection method which is more?onvenient?nd?apid can replace the traditional ECG detection to identify atrial fibrillation, having clinical value for the long-term observation and monitoring of patients with atrial fibrillation.