使用摄像头作为非接触式传感器的逐搏心率检测方法评估

Assessment of beat-to-beat heart rate detection method using a camera as contactless sensor.

作者信息

Iozzia L, Cerina L, Mainardi L T

出版信息

Annu Int Conf IEEE Eng Med Biol Soc. 2016 Aug;2016:521-524. doi: 10.1109/EMBC.2016.7590754.

DOI:10.1109/EMBC.2016.7590754

Abstract

Video photoplethysmography (videoPPG) has emerged as area of great interest thanks to the possibility of remotely assessment of cardiovascular parameters, as heart rate (HR), respiration rate (RR) and heart rate variability (HRV). The present article proposes a fully automated method based on chrominance model, that selects for each subject the best region of interest (ROI) to detect and evaluate the accuracy of beat detection and interbeat intervals (IBI) measurements. The experimental recordings were conducted on 26 subjects which underwent a rest-to-stand maneuver. The results show that the accuracy of beat detection is slightly better during supine position (95%) compared to the standing one (92%), due to the maintenance of the balance that introduces larger motion artifact in the signal dynamic. The error in the measurement (expressed as mean±sd) of instantaneous heart rate is of +0.04 ±3.29 bpm in rest and +0.01±4.26 bpm in stand.

摘要

视频光电容积脉搏波描记法（videoPPG）由于具备远程评估心血管参数（如心率（HR）、呼吸频率（RR）和心率变异性（HRV））的可能性，已成为一个备受关注的领域。本文提出了一种基于色度模型的全自动方法，该方法为每个受试者选择最佳感兴趣区域（ROI），以检测并评估心跳检测和心动周期（IBI）测量的准确性。实验记录是在26名经历了从静息到站立动作的受试者身上进行的。结果表明，由于站立时维持平衡会在信号动态中引入更大的运动伪影，因此仰卧位时心跳检测的准确率（95%）略高于站立位（92%）。静息时瞬时心率测量的误差（以平均值±标准差表示）为+0.04±3.29次/分钟，站立时为+0.01±4.26次/分钟。