Remote Estimation of Peripheral Oxygen Saturation and Pulse Rate From Facial Analysis Using a Smartphone Camera

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Physiological Monitoring

Background:

• Vital sign monitoring is crucial in healthcare but traditional devices require physical contact, which can be unreliable.
• Contact-based sensors may fail due to movement, especially during critical events, necessitating contactless solutions.

Purpose of the Study:

• To investigate the feasibility of estimating SpO2 and pulse rate using only smartphone facial videos.
• To develop and validate a contactless method for remote vital sign assessment.

Main Methods:

• Collected facial video data from 10 healthy subjects under ideal lighting conditions.
• Employed advanced image and signal processing techniques to extract remote photoplethysmography (rPPG) signals from forehead videos.
• Utilized rPPG signals to estimate SpO2 and pulse rate.

Main Results:

• Achieved SpO2 estimation with an accuracy RMS of 1.34% and MAE ± STD of 1.26 ± 0.68% within the 92%-99% range.
• Obtained pulse rate estimation with an accuracy RMS of 3.91 bpm and MAE ± STD of 3.24 ± 2.11 bpm within the 60-90 bpm range.
• Demonstrated the possibility of accurate vital sign estimation from smartphone facial videos.

Conclusions:

• Smartphone-based contactless vital sign monitoring is achievable.
• This technology offers a promising, non-invasive alternative for remote patient assessment.
• Further research can explore applications in diverse clinical and home-care settings.