Remote iPPG Vital Sign Estimation Under Motion and Illumination Noise
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing remote photoplethysmography (RPPG) systems face challenges in accurately estimating vital signs in volatile environments due to noise from illumination variations and motion, particularly in NIR frequencies, leading to low signal-to-noise ratio and reduced sensitivity.
Innovation Solution
An imaging photoplethysmography (iPPG) system using a narrow-band near-infrared (NIR) light source and a deep neural network (DNN) with a time-series U-Net architecture incorporating temporal recurrence to process multidimensional time-series data from spatial regions of the skin, jointly estimating vital signs to reduce noise effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If remote PPG systems use NIR frequencies for non-contact vital signs measurement, then ease of operation is improved (no contact required), but measurement precision deteriorates due to low signal-to-noise ratio and reduced sensitivity in volatile environments
Solution Approach 1:
The patent divides the skin region into multiple spatial regions (e.g., facial regions) and processes each region independently through the neural network. This segmentation allows the system to identify and utilize the most informative regions while reducing the impact of noise in other regions, thereby improving measurement precision while maintaining non-contact operation.
Solution Approach 2:
The patent transforms the input from a single 2D image to a multidimensional time-series signal by extracting features across multiple spatial regions and time points. This dimensional expansion provides additional information dimensions that the neural network can process to distinguish signal from noise, improving precision without requiring contact.
2Measurement precision
If the system processes multidimensional time-series data from multiple spatial regions, then measurement precision is improved through joint estimation, but device complexity increases due to the deep neural network architecture
Solution Approach 1:
The deep neural network with U-Net architecture is designed to perform multiple functions: extracting spatial features, temporal patterns, and their interactions simultaneously. This multi-functionality consolidates what would otherwise require separate processing stages into a single unified model, improving precision while managing complexity through integration rather than accumulation of separate components.
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
The system provides accurate estimation of vital signs, such as heart rate and heart rate variability, even in volatile environments by filtering out ambient illumination noise and enhancing signal quality through joint estimation of PPG signals across different skin regions.
Implementation Method 1
PPG is an optical measurement technique that evaluates a time-variant change of light reflectance or transmission of an area or volume of interest, which can be used to detect blood volume changes in microvascular bed of tissue
Implementation Method 2
One embodiment uses a narrow-bandwidth (narrow-band) near-infrared (NIR) light source to illuminate the skin of the person at a narrow frequency band including a near-infrared wavelength of 940 nm
Implementation Method 3
A typical pulse oximeter can include a combination of a green LED, a blue LED, a red LED, and an infrared LED as light sources and one photodiode for detecting light that has been transmitted through patient tissue
Data Source
Figure 1A
Figure 1B
Figure 1C
AI summary
An imaging photoplethysmography (iPPG) system is provided. The iPPG system receives a sequence of images of different regions of the skin of the person, where each region including pixels of different intensities indicative of variation of coloration of the skin. The iPPG system further transforms the sequence of images into a multidimensional time-series signal, each dimension corresponding to a different region from the different regions of the skin. The iPPG system further processes the multidimensional time-series signal with a time-series U-Net neural network wherein the pass-through layers include a recurrent neural network (RNN) to generate a PPG waveform, where the vital sign of the person is estimated based on the PPG waveform, and the iPPG system further renders the estimated vital sign of the person.