Using Contactless Facial Image Recognition Technology to Detect Blood Oxygen Saturation

Area of Science:

• Biomedical Engineering
• Optical Sensing
• Health Informatics

Background:

• COVID-19 pandemic highlighted the need for remote patient monitoring.
• Low blood oxygen levels (SpO2) are critical indicators of respiratory distress.
• Existing monitoring methods can be invasive or require specialized equipment.

Purpose of the Study:

• To develop and validate a non-contact method for measuring blood oxygen saturation (SpO2) using a general network camera.
• To assess the accuracy and efficiency of the proposed remote photoplethysmography (RPPG) technique.
• To explore the feasibility of integrating SpO2 monitoring into everyday devices like smartphones.

Main Methods:

• Utilized a network camera to capture facial images (forehead area).
• Applied remote photoplethysmography (RPPG) principles with image signal processing of red and blue light waves.
• Calculated SpO2 by analyzing light reflection, standard deviation, and mean values.

Main Results:

• Achieved a maximum error of 2% when compared to a certified blood oxygen meter.
• Measurement time was significantly reduced to 30 seconds, outperforming other studies.
• Demonstrated the influence of illuminance on experimental SpO2 values.

Conclusions:

• The developed RPPG method offers a cost-effective, convenient, and safe approach for home SpO2 monitoring.
• This technology has potential for integration into smartphones and laptops for widespread personal health management.
• The non-contact SpO2 detection provides a valuable tool for individuals needing regular oxygen level checks.