Contactless rPPG System Combining Face and Hand Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current contactless camera-based methods for estimating blood volume change, particularly for darker skin tones, face significant performance drops due to lower light reflection intensity, leading to reduced signal-to-noise ratios and unreliable measurements.
Innovation Solution
The method combines face and hand plethysmograph signals using a contactless sensor system, leveraging the lower melanin content of the palm to enhance light reflection and improve estimation accuracy, even when facial signals are unreliable due to occlusions or low light conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If facial region imaging is used for blood volume change estimation, then the method is simple and contactless, but measurement precision significantly drops for individuals with darker skin
Solution Approach 1:
The system divides the body into multiple measurement regions (face and hand) and selectively uses different regions based on skin tone characteristics. For darker skin tones, the hand region is prioritized while the face region may be excluded, thereby maintaining measurement precision without compromising the contactless nature of the method
Solution Approach 2:
The system dynamically changes the measurement parameter (selected body region) based on detected skin tone characteristics. When darker skin is detected in the facial region, the system switches to using the hand region for photoplethysmography, thus adapting the measurement approach to maintain accuracy across different skin tones
2Measurement precision
If hand-based rPPG is used to compensate for low light reflection in darker skin, then measurement precision improves, but device complexity increases due to multiple acquisition regions
Solution Approach 1:
The system implements dynamic selection of measurement regions based on real-time skin tone detection and signal quality assessment. The hand region is not continuously monitored but activated dynamically when facial signals are deemed insufficient, thereby improving precision for darker skin while minimizing the operational complexity of multi-region acquisition
Solution Approach 2:
The system autonomously determines which body region to use for measurement based on detected skin characteristics and signal quality, without requiring manual intervention or complex external control. The automated skin tone detection and region selection algorithms enable the system to self-adjust, reducing the operational burden despite the capability for multi-region acquisition
3Reliability
If confidence-based selective acquisition is implemented, then reliability of vital sign measurements improves, but loss of time increases due to conditional signal acquisition
Solution Approach 1:
The system performs preliminary skin tone detection and facial signal quality assessment before deciding whether to acquire hand signals. This preliminary evaluation allows the system to quickly determine if hand-based acquisition is necessary, reducing the time penalty of conditional acquisition while ensuring reliability for darker skin tones where facial signals may be insufficient
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach provides more reliable vital sign measurements, including heart rate, respiratory rate, oxygen saturation, and blood pressure, by compensating for light reflection issues and improving signal quality across various skin tones and conditions.
Implementation Method 1
acquiring a first pleth (plethysmograph) signal based at least in part on capturing ambient light reflected from at least one region of interest (ROI) on a face of the subject
Implementation Method 2
remote photoplethysmography (rPPG) can be beneficial in such cases
Data Source
AI summary
Embodiments of the present disclosure relate to contactless sensor-driven devices, systems, and methods for performing remote photoplethysmography. In one embodiment, a method of assessing at least one vital sign of a subject comprises: acquiring a first pleth signal based at least in part on capturing ambient light reflected from at least one region of interest (ROI) on a face of the subject; acquiring a second pleth signal from a palmar side of a hand of the subject; and computing estimates of the at least one vital sign of the subject from the first and/or second pleth signals.