Wearable PPG Blood Pressure Estimation With Signal Quality Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing single-site PPG-based blood pressure estimation algorithms suffer from high overlap and correlation of features, leading to overfitting and inaccurate predictions due to the capricious nature of PPG signals and challenges in maintaining signal quality.
Innovation Solution
A wearable device using multi-wavelength PPG sensors collects photoplethysmography signals at multiple wavelengths, filters them with a bandpass filter, calculates a Signal Quality Indicator (SQI) to eliminate low-quality signals, and employs a Random Forest model trained with K-Means clustering and regularization techniques to create personalized cardiovascular profiles for accurate blood pressure estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If single-wavelength PPG signals with time-and frequency domain features are used for blood pressure estimation, then the measurement process is simple, but the features have high overlap and correlation making the algorithms prone to overfitting
Solution Approach 1:
The patent segments the PPG signal analysis by separating time-domain features, frequency-domain features, and non-linear features into distinct categories. This segmentation allows the algorithm to process diverse feature types independently, reducing the harmful overlap and correlation between features while maintaining measurement simplicity through systematic feature organization.
Solution Approach 2:
The patent introduces non-linear features as a new dimension beyond traditional time-and frequency domain features. By adding this additional analytical dimension, the algorithm captures more comprehensive signal characteristics, reducing feature correlation and preventing overfitting while maintaining the simplicity of single-wavelength measurement.
2Measurement precision
If multiple-site PPG sensors are used to obtain Pulse-Transit Time feature, then blood pressure measurement accuracy is improved, but sensor placement complexity and signal noise increase
Solution Approach 1:
The patent extracts Pulse-Transit Time features and other cardiovascular parameters from single-site PPG signals by isolating and processing specific waveform characteristics. This extraction approach captures the essential blood pressure information normally requiring multiple sites, while eliminating the complexity of multi-site sensor placement and synchronization.
Solution Approach 2:
The patent uses advanced signal processing algorithms as intermediaries to translate single-site PPG measurements into accurate blood pressure estimates. These intermediary processing steps compensate for the lack of multiple measurement sites, extracting sufficient physiological information through sophisticated analysis of the available signal.
3Extent of automation
If machine learning algorithms are trained on PPG features, then blood pressure estimation is achieved, but the capricious nature of PPG signals causes overfitting on certain characteristics
Solution Approach 1:
The patent transforms the PPG signal representation by extracting multiple types of features (time-domain, frequency-domain, non-linear) and applying various signal processing transformations. These parameter changes create a more robust feature set that captures essential physiological information while reducing the impact of signal variability and preventing overfitting to specific signal characteristics.
Solution Approach 2:
The patent creates a composite feature vector combining diverse feature types from different signal domains. This composite approach integrates multiple independent feature categories, creating a more reliable and generalizable input for machine learning algorithms that is less susceptible to overfitting on any single characteristic.
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 method provides accurate, personalized blood pressure measurements by isolating high-quality PPG signals, reducing overfitting, and adapting to individual physiological variations, ensuring reliable predictions across diverse patient conditions.
Implementation Method 1
Photoplethysmography (PPG) is a non-invasive optical technique that measures blood volume changes in the microvascular bed of tissue. It works by shining a light source, typically an LED, onto the skin and detecting the amount of light that is transmitted or reflected back to a photodetector.
Implementation Method 2
filters them with a bandpass filter
Data Source
AI summary
Methods for non-invasively measuring blood pressure using multi-wavelength photoplethysmography (PPG) signals obtained from peripheral blood vessels. A wearable device is calibrated using PPG signals measured distal to the wearable device and analyzed in order to create a cardiovascular profile of a wearer. After calibration, windows of PPG signals are analyzed to provide a blood pressure estimation, based on the created profile. The current invention utilizes single-site measured PPG signals in order to make a prediction of the wearer's blood pressure. By using a combination of multi-wavelength PPG waveform features and a range of vital signs calculated by the wearable device, the device is able to provide a more accurate and robust prediction of blood pressure compared to existing methods. This has important implications for clinical settings where frequent and remote monitoring of blood pressure is necessary for managing a range of health conditions.