PPG Signal Quality Grading with Amplitude and SNR Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing PPG signal evaluation methods struggle with inaccurate signal classification due to reliance on complex waveform forms, cycle splitting algorithms, and dataset generalization, leading to rough classification results and increased power consumption.
Innovation Solution
A multi-step evaluation method involving preprocessing, high-pass and low-pass filtering, signal-to-noise ratio analysis, autocorrelation function, and Fourier transform to assess PPG signal quality, categorizing signals into three grades based on amplitude difference, noise ratios, and spectral characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If template matching method is used for PPG signal evaluation, then signal classification can be performed, but the method relies greatly on cycle splitting algorithm and template selection criteria which may lead to inaccurate evaluation
Solution Approach 1:
The patent transforms the evaluation from relying on complex waveform form matching to using quantitative parameter analysis. Specifically, it extracts parameters such as amplitude difference, signal-to-noise ratio, and spectral characteristics, and uses these numerical parameters for signal classification. This parameter-based approach eliminates the need for complex cycle splitting algorithms and template selection, thereby improving both accuracy and robustness.
Solution Approach 2:
The patent replaces the mechanical waveform matching process (cycle splitting and template matching) with a computational parameter extraction and analysis system. By substituting the mechanical analogy of fitting waveforms together with mathematical parameter calculation, the system achieves more reliable and accurate signal evaluation without being constrained by the complexities of waveform morphology.
2Measurement precision
If machine learning and deep learning methods are used for PPG signal evaluation, then good classification results can be achieved, but the process is time-consuming and relies on dataset generalization and labeling accuracy
Solution Approach 1:
The patent uses simple, easily computable parameters (amplitude difference, signal-to-noise ratio, spectral characteristics) instead of complex machine learning models. These parameters can be calculated quickly and independently without requiring time-consuming model training or data labeling processes. The evaluation system becomes lightweight and immediate, eliminating the time investment required for ML model development.
Solution Approach 2:
The patent extracts key quantitative features from the PPG signal (amplitude characteristics, noise ratios, spectral properties) and uses these extracted parameters for direct evaluation. By taking out only the essential numerical characteristics and discarding the need for complex pattern recognition systems, the method achieves accurate classification without the time-consuming data processing required by machine learning approaches.
3Ease of operation
If existing evaluation methods are used, then signal classification is performed, but the results are rough (binary classification only) and cannot provide detailed quality grading
Solution Approach 1:
The patent segments the signal quality evaluation into multiple independent parameter assessments: amplitude difference evaluation, signal-to-noise ratio evaluation, and spectral characteristic analysis. Each parameter is evaluated separately and contributes to the overall quality grading. This segmentation allows for fine-grained differentiation between signal qualities, enabling three-grade classification (first-grade, second-grade, third-grade) instead of rough binary classification.
Solution Approach 2:
The patent adds multiple evaluation dimensions (amplitude, noise ratio, spectral characteristics) to the traditional single-dimension binary classification approach. By introducing these additional quantitative dimensions, the system transitions from coarse binary classification to fine-grained multi-level quality grading, providing detailed information about signal quality that enables better decision-making for physiological indicator calculations.
4Productivity
If PPG signals are processed without quality evaluation, then processing can be performed, but power consumption increases and inaccurate calculations occur
Solution Approach 1:
The patent performs quality evaluation as a preliminary action before physiological indicator calculations. By assessing signal quality parameters (amplitude difference, signal-to-noise ratio, spectral characteristics) in advance, the system can identify and discard low-quality signals before they are processed, preventing wasted computational resources on unusable data. This preliminary filtering action reduces overall power consumption while maintaining calculation accuracy.
Solution Approach 2:
The patent converts the potentially harmful effect of signal noise and distortion into a beneficial evaluation mechanism. By quantifying noise and distortion through parameter analysis (signal-to-noise ratio, spectral characteristics), the system identifies problematic signals and excludes them from processing. This transforms the presence of noise from a harmful factor into a detectable feature that guides selective processing, thereby reducing power consumption on unusable signals.
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
Enables multi-level quantitative evaluation of PPG signals, improving accuracy and reducing power consumption by distinguishing signal quality for precise physiological indicator calculations.
Implementation Method 1
The PPG technology mainly uses a photodiode to emit to the skin, and then obtains waveforms capable of reflecting blood flow by receiving the intensity of reflected light
Implementation Method 2
performing high-pass filtering and low-pass filtering on the original PPG signal respectively
Implementation Method 3
performing high-pass filtering and low-pass filtering on the original PPG signal respectively
Implementation Method 4
performing Fourier transform on the autocorrelation function to obtain a power spectrum of the useful signal
Data Source
AI summary
The present invention relates to a PPG signal quality evaluation method and apparatus and a PPG signal processing method and system. The PPG signal quality evaluation method includes: S1: acquiring an original signal; S2: when an average amplitude difference thereof is within a first preset range, executing S3, and otherwise, executing S11; S3: acquiring a useful signal; S4: when the high-frequency signal-to-noise ratio and the low-frequency signal-to-noise ratio thereof are respectively greater than corresponding first threshold values, executing S5, and otherwise, executing S11; S5: when the kurtosis and the skewness of a power spectrum of the useful signal are respectively greater than second threshold values, executing S6, and otherwise, executing S11; S6: when the average peak period of a correlation function is within a second preset range, executing S7, and otherwise, executing S11.