TY - GEN
T1 - Effect of ambient lighting and skin tone on estimation of heart rate and pulse transit time from video plethysmography
AU - Shirbani, F.
AU - Hui, N.
AU - Tan, I.
AU - Butlin, M.
AU - Avolio, A. P.
PY - 2020/7/1
Y1 - 2020/7/1
N2 - Video-based photoplethysmography (vPPG) enables remote and contactless detection of the peripheral pulse of blood flow. This provides a potential mean to extract heart rate (HR) and pulse transit time (PTT) for the purpose of remote health monitoring. The accuracy of average HR and PTT extracted from a two-minute vPPG recording has been investigated at six different lighting conditions among participants with a range of Fitzpatrick skin scores. 12 healthy volunteers (6 females, 27 ± 6 years) were recruited. The video, electrocardiogram and finger PPG were acquired from immobile resting subjects. The vPPG signals from red, green and blue channels, and a combination of those were investigated. The vPPG signals were extracted from two regions of interest (ROIs): one on the forehead and one on the palm of the left hand. The estimated HR error (HR-error) was significantly lower for vPPG from green channels in both ROIs (ROI1 [p<0.001], ROI2 [p<0.05]). The signal from ROI1 demonstrated lower HR-error than ROI2 (p<0.001). HR-error from the darkest lighting conditions (Lumen 1 and 2) were significantly higher than the others (p<0.05). Furthermore, HR-error showed a positive correlation with skin tone scores in every lighting condition. However, at brighter lighting intensity, HR-error was independent of the skin tone score. PTT calculated from vPPG (vPTT) were compared between the 6 levels of lightings and the result was significantly different (p<0.05). In darker lighting conditions, the vPTT increased. Pulse arrival time measured from PPG (PAT-PPG) was calculated, and a positive correlation was found between the ratio of vPTT/PAT-PPG and skin tone score at six different lightings. However, this dependency decreases in brighter lighting intensity. These results suggest that HR-error and the ratio of vPTT/PAT increase with darker skins and at darker backgrounds. However, at brighter lighting conditions, the skin tone score is not a confounder of vPPG accuracy.
AB - Video-based photoplethysmography (vPPG) enables remote and contactless detection of the peripheral pulse of blood flow. This provides a potential mean to extract heart rate (HR) and pulse transit time (PTT) for the purpose of remote health monitoring. The accuracy of average HR and PTT extracted from a two-minute vPPG recording has been investigated at six different lighting conditions among participants with a range of Fitzpatrick skin scores. 12 healthy volunteers (6 females, 27 ± 6 years) were recruited. The video, electrocardiogram and finger PPG were acquired from immobile resting subjects. The vPPG signals from red, green and blue channels, and a combination of those were investigated. The vPPG signals were extracted from two regions of interest (ROIs): one on the forehead and one on the palm of the left hand. The estimated HR error (HR-error) was significantly lower for vPPG from green channels in both ROIs (ROI1 [p<0.001], ROI2 [p<0.05]). The signal from ROI1 demonstrated lower HR-error than ROI2 (p<0.001). HR-error from the darkest lighting conditions (Lumen 1 and 2) were significantly higher than the others (p<0.05). Furthermore, HR-error showed a positive correlation with skin tone scores in every lighting condition. However, at brighter lighting intensity, HR-error was independent of the skin tone score. PTT calculated from vPPG (vPTT) were compared between the 6 levels of lightings and the result was significantly different (p<0.05). In darker lighting conditions, the vPTT increased. Pulse arrival time measured from PPG (PAT-PPG) was calculated, and a positive correlation was found between the ratio of vPTT/PAT-PPG and skin tone score at six different lightings. However, this dependency decreases in brighter lighting intensity. These results suggest that HR-error and the ratio of vPTT/PAT increase with darker skins and at darker backgrounds. However, at brighter lighting conditions, the skin tone score is not a confounder of vPPG accuracy.
UR - http://www.scopus.com/inward/record.url?scp=85091007167&partnerID=8YFLogxK
U2 - 10.1109/EMBC44109.2020.9176731
DO - 10.1109/EMBC44109.2020.9176731
M3 - Conference proceeding contribution
C2 - 33018549
AN - SCOPUS:85091007167
SN - 9781728119915
T3 - Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS
SP - 2642
EP - 2645
BT - 42nd Annual International Conferences of the IEEE Engineering in Medicine and Biology Society
PB - Institute of Electrical and Electronics Engineers (IEEE)
CY - Piscataway, NJ
T2 - 42nd Annual International Conferences of the IEEE Engineering in Medicine and Biology Society, EMBC 2020
Y2 - 20 July 2020 through 24 July 2020
ER -