Invention Title:

Adaptable Camera-based Contactless SpO2 Detection

Publication number:

US20250087016

Publication date:
Section:

Physics

Class:

G06V40/15

Inventors:

Applicant:

Smart overview of the Invention

The invention provides a method for contactless detection of blood-oxygen saturation (SpO2) using a camera. It involves recording a video of a user's skin, estimating a ratio of ratios (RoR) from the video, and transforming this RoR using a baseline model. This model, trained on a baseline user's data, predicts the current SpO2 level of the user based on the transformed RoR. This approach enables non-invasive and continuous monitoring of blood oxygen levels without the need for direct contact with the skin.

Technical Field

The application relates to techniques for measuring SpO2 levels using adaptable, camera-based methods. It leverages optical signals captured by a camera to estimate blood oxygen saturation through advanced signal processing and modeling techniques. This innovation aims to provide an alternative to traditional invasive or contact-based methods like arterial blood gas tests or finger pulse oximeters.

Background

Blood volume changes in human vessels are linked to physiological phenomena such as heartbeats and blood pressure. These changes can also indicate oxygen levels in the blood, commonly measured as SpO2. Traditional methods for measuring SpO2, such as arterial blood gas tests, are invasive and require clinical settings. Alternatively, finger pulse oximeters offer non-invasive measurements but require continuous skin contact, posing inconvenience and infection risks.

Innovative Approach

The method employs a ratio-of-ratios technique utilizing light reflection properties to estimate SpO2 levels. When light interacts with skin, it undergoes specular and diffuse reflection. By analyzing these reflections at different wavelengths, the system can determine oxygenated versus deoxygenated hemoglobin levels. The model uses this data to calculate the RoR, which is then transformed based on a baseline model for accurate SpO2 estimation.

Technical Details

The process involves capturing light reflected off the skin at various wavelengths and analyzing its intensity and reflectance characteristics. The mathematical formulation incorporates variables like static and pulsatile blood reflectance. By isolating these factors, the system achieves accurate readings independent of external light intensity variations. This innovative technique allows for continuous monitoring without physical contact, providing convenience and hygiene benefits over existing methods.