Chen Qiong, Jiang Xinyu, Liu Xiangyu, Lu Chunmei, Wang Laishuan, Chen Wei
Annu Int Conf IEEE Eng Med Biol Soc. 2020 Jul;2020:5822-5825. doi: 10.1109/EMBC44109.2020.9175651.
Heart rate (HR) measurement is crucial for newborn infant monitoring in the neonatal intensive care unit (NICU). The widely used contact HR measurement methods based on electrocardiography (ECG) and photoplethysmography (PPG) signals can lead to discomfort and possible skin irritation on neonates, which limit its application in NICU scenarios. In this work, we propose a non-contact HR monitoring method simply using a RGB camera. Eulerian video magnification (EVM) is employed to detect the subtle changes of neonatal faces results from blood circulation. The magnified signal is then transformed to the spectral domain to extract HR information. Compared with the widely investigated independent component analysis (ICA)-based HR measurement method using video recordings, the proposed method can achieve the real time HR measurement, which is a significant superiority in NICU neonatal monitoring. To the best of our knowledge, this is the first study to employ EVM algorithm in real time neonatal HR monitoring.
心率（HR）测量对于新生儿重症监护病房（NICU）中的新生儿监测至关重要。基于心电图（ECG）和光电容积脉搏波描记法（PPG）信号的广泛使用的接触式HR测量方法可能会给新生儿带来不适并可能导致皮肤刺激，这限制了其在NICU场景中的应用。在这项工作中，我们提出了一种仅使用RGB相机的非接触式HR监测方法。采用欧拉视频放大（EVM）来检测由血液循环引起的新生儿面部的细微变化。然后将放大后的信号转换到频域以提取HR信息。与广泛研究的基于独立成分分析（ICA）的使用视频记录的HR测量方法相比，所提出的方法可以实现实时HR测量，这在NICU新生儿监测中是一个显著的优势。据我们所知，这是第一项在实时新生儿HR监测中采用EVM算法的研究。
