PPG Motion Artifact Elimination via Acceleration Spectrum Subtraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for eliminating motion artifacts from PPG signals using wearable devices are inefficient, especially in real-time applications, as they either require high-resolution spectrum estimation or are limited in situations with low motion artifacts, making accurate pulse wave measurement challenging.
Innovation Solution
A method and apparatus that utilize three-axis acceleration values to filter and Fourier-transform PPG signals, arranging power spectra in ascending order to progressively subtract noise components, with coefficient values adjusted to minimize interference while preserving heart rate information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high resolution spectrum estimation is used to eliminate motion artifacts, then motion artifact elimination accuracy is improved, but processing time increases and real-time measurement becomes difficult
Solution Approach 1:
The patent segments the motion artifact elimination process by separating the spectral analysis into distinct stages: initial power spectrum calculation from PPG signal, acceleration signal power spectrum calculation, and sequential subtraction of acceleration spectra from the PPG spectrum. This segmentation allows efficient processing by avoiding full high-resolution spectrum estimation while maintaining effectiveness in removing motion artifacts.
2Productivity
If simple filtering methods are used, then processing speed is improved, but effectiveness in eliminating motion artifacts deteriorates in situations with large motion artifacts
Solution Approach 1:
The patent introduces acceleration signals as an intermediary to mediate the motion artifact elimination process. By calculating the power spectra of acceleration signals and sequentially subtracting them from the PPG signal power spectrum, the method uses the acceleration data as a mediator to identify and remove motion artifact components without requiring complex high-resolution spectral analysis, thus maintaining both speed and effectiveness.
3Measurement precision
If motion artifact elimination techniques are applied, then pulse wave measurement accuracy is improved, but device complexity increases
Solution Approach 1:
The patent applies the self-service principle by using the acceleration signals, which are already being collected by the wearable device for other purposes, to eliminate motion artifacts from the PPG signal. The existing acceleration data serves the dual purpose of motion tracking and artifact removal, eliminating the need for additional specialized sensors or complex processing hardware, thus improving measurement accuracy without significantly increasing device complexity.
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 effectively and efficiently removes motion artifacts, simplifying calculations and improving the accuracy of pulse wave measurement, allowing for stable and real-time heart rate estimation.
Implementation Method 1
Fourier-transforming the PPG signal and the three-axis acceleration signals corresponding to the extracted period to transform into a frequency domain
Data Source
AI summary
The present invention relates to an apparatus for eliminating motion artifacts and a method thereof, and the method for eliminating motion artifacts includes steps of receiving a PPG signal and three-axis (X-axis, Y-axis, and Z-axis) acceleration signals measured from an examinee, extracting the PPG signal and the three-axis acceleration signals by a period of a predetermined unit of time and filtering the signals at a preset bandwidth, Fourier-transforming the PPG signal and the three-axis acceleration signals corresponding to the extracted period, arranging power spectra of the X-axis, Y-axis, and Z-axis in ascending order of value of average power spectrum by calculating an average power of each frequency domain of the three-axis accelerations, and extracting a final power spectrum of the PPG signal by sequentially eliminating power spectra of the X, Y, and Z-axis accelerations from the power spectrum of the PPG signal in an order of arrangement.