Remote Photoplethysmography with a High-Speed Camera Reveals Temporal and Amplitude Differences between Glabrous and Non-Glabrous Skin.
Meiyun CaoTimothy BurtonGennadi SaikoAlexandre DouplikPublished in: Sensors (Basel, Switzerland) (2023)
Photoplethysmography (PPG) is a noninvasive optical technology with applications including vital sign extraction and patient monitoring. However, its current use is primarily limited to heart rate and oxygenation monitoring. This study aims to demonstrate the utility of PPG for physiological investigations. In particular, we sought to demonstrate the utility of simultaneous data acquisition from several regions of tissue using remote/contactless PPG (rPPG). Specifically, using a high-speed scientific-grade camera, we collected rPPG from the hands (palmar/dorsal) of 22 healthy volunteers. Data collected through the red and green channels of the RGB CMOS sensor were analyzed. We found a statistically significant difference in the amplitude of the glabrous skin signal over the non-glabrous skin signal (1.41 ± 0.85 in the red channel and 2.27 ± 0.88 in the green channel). In addition, we found a statistically significant lead of the red channel over the green channel, which is consistent between glabrous (17.13 ± 10.69 ms) and non-glabrous (19.31 ± 12.66 ms) skin. We also found a statistically significant lead time (32.69 ± 55.26 ms in the red channel and 40.56 ± 26.97 ms in the green channel) of the glabrous PPG signal over the non-glabrous, which cannot be explained by bilateral variability. These results demonstrate the utility of rPPG imaging as a tool for fundamental physiological studies and can be used to inform the development of PPG-based devices.
Keyphrases
- high speed
- heart rate
- atomic force microscopy
- mass spectrometry
- high resolution
- multiple sclerosis
- ms ms
- soft tissue
- heart rate variability
- wound healing
- blood pressure
- spinal cord
- electronic health record
- spinal cord injury
- convolutional neural network
- neuropathic pain
- machine learning
- deep learning
- artificial intelligence
- low cost
- case control