基于模糊熵的指尖光电容积脉搏波运动伪影检测与心率估计
Fuzzy entropy based motion artifact detection and pulse rate estimation for fingertip photoplethysmography.
作者信息
Paradkar Neeraj, Chowdhury Shubhajit Roy
出版信息
Annu Int Conf IEEE Eng Med Biol Soc. 2014;2014:58-61. doi: 10.1109/EMBC.2014.6943528.
The paper presents a fingertip photoplethysmography (PPG) based technique to estimate the pulse rate of the subject. The PPG signal obtained from a pulse oximeter is used for the analysis. The input samples are corrupted with motion artifacts due to minor motion of the subjects. Entropy measure of the input samples is used to detect the motion artifacts and estimate the pulse rate. A three step methodology is adapted to identify and classify signal peaks as true systolic peaks or artifact. CapnoBase database and CSL Benchmark database are used to analyze the technique and pulse rate estimation was performed with positive predictive value and sensitivity figures of 99.84% and 99.32% respectively for CapnoBase and 98.83% and 98.84% for CSL database respectively.
本文提出了一种基于指尖光电容积脉搏波描记法（PPG）的技术来估计受试者的脉搏率。从脉搏血氧仪获得的PPG信号用于分析。由于受试者的轻微运动，输入样本会受到运动伪影的干扰。输入样本的熵测量用于检测运动伪影并估计脉搏率。采用三步方法将信号峰值识别并分类为真正的收缩期峰值或伪影。使用CapnoBase数据库和CSL基准数据库对该技术进行分析，CapnoBase数据库的脉搏率估计的阳性预测值和灵敏度分别为99.84%和99.32%，CSL数据库的分别为98.83%和98.84%。
Zotero 插件