Adaptive Noise Reference Selection for PPG Motion Artifact Removal
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for removing motion artifacts from physiological measurements, such as photoplethysmography (PPG) signals, using tri-axial accelerometer (ACC) sensors are ineffective as they often fail to provide a well-correlated noise reference, leading to deteriorated signal quality due to mismatched correlations between ACC signals and PPG distortions under different motions.
Innovation Solution
A method and device that identify a noise reference from tri-axial ACC signals by determining the axis with maximal amplitude or PSD correlation to PPG distortions, using adaptive filtering to correct PPG signals, and include a signal quality checking module to ensure reliability of the filtered output.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If adaptive filtering with ACC signals is used to remove motion artifacts from PPG signals, then motion artifacts can be reduced, but signal quality deteriorates when ACC signals are not correlated with PPG distortion
Solution Approach 1:
The patent dynamically changes the parameter of noise reference selection by evaluating correlation between ACC signals and PPG distortion across different axes and motion conditions. The system adjusts which ACC axis serves as noise reference based on real-time correlation analysis, ensuring optimal artifact removal while maintaining signal quality.
Solution Approach 2:
The patent implements a feedback mechanism where the system continuously monitors the correlation between ACC signals and PPG distortion, and adjusts the noise reference selection accordingly. This feedback loop ensures that artifact removal is only applied when correlation exceeds a threshold, preventing signal quality deterioration.
2Object-affected harmful factors
If tri-axial ACC signals are combined or a specific axis is specified as noise reference, then artifact removal can be performed, but performance decreases under different motions due to mismatched correlation
Solution Approach 1:
The patent makes the noise reference selection dynamic rather than static. The system continuously evaluates which ACC axis has the highest correlation with PPG distortion and adjusts the noise reference accordingly. This dynamic adaptation enables effective artifact removal across various motion types including linear acceleration, rotation, and vibration.
Solution Approach 2:
The patent changes the parameter of noise reference based on motion conditions by evaluating correlation metrics for each ACC axis and selecting the optimal one. This parameter adjustment ensures adaptability to different motion scenarios while maintaining effective artifact removal performance.
3Object-affected harmful factors
If ACC signals are used as noise reference for filtering, then motion artifacts can be removed, but computational complexity increases
Solution Approach 1:
The patent simplifies the computational process by changing the approach from analyzing all three ACC axes continuously to selecting a single optimal axis based on correlation evaluation. This parameter optimization reduces computational complexity while maintaining effective artifact removal.
Solution Approach 2:
The patent extracts only the most relevant ACC axis signal that shows highest correlation with PPG distortion, rather than processing all three axes. This extraction approach reduces computational load by focusing only on the necessary noise reference signal.
Data Source
AI summary
Device and method for removing artifacts in physiological measurements. The method can comprise the steps of obtaining a physiological signal of a user; obtaining corresponding motion data representative of motion of the user; determining whether the physiological signal is distorted; and if the physiological signal is determined to be distorted, identifying a noise reference and filtering the physiological signal with the noise reference.