Regression Analysis for Blood Oxygen Saturation Uncertainty
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Pulse oximetry measurements are affected by uncertainties and noise, leading to inaccurate blood flow characteristic calculations due to variations in light absorption and interference from external and physiological sources.
Innovation Solution
The use of regression techniques, such as Principal Component Regression or orthogonal-fit, to accommodate uncertainty in both red and infrared signals, allowing for more accurate estimation of blood oxygen saturation by accounting for error in both measurement dimensions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional pulse oximetry measurement is used, then the measurement process is simple and quick, but the measurement precision deteriorates due to uncertainties and noise in light absorption data
Solution Approach 1:
The patent introduces an intermediary regression analysis process between the raw light absorption measurements and the final blood oxygen saturation calculation. This intermediary step models the relationship between red and infrared light absorption ratios and known hemoglobin characteristics, filtering out noise and uncertainties while preserving the essential measurement information.
Solution Approach 2:
The patent employs feedback mechanisms where the regression model continuously refines the modulation ratio determination by comparing measured values against expected physiological ranges and adjusting for detected uncertainties. This feedback loop improves measurement precision by compensating for noise and variations in real-time.
2Measurement precision
If regression analysis is applied to accommodate uncertainty in both red and infrared signals, then the blood oxygen saturation estimation accuracy improves, but the computational complexity increases
Solution Approach 1:
The patent transforms the raw light absorption signals into modified parameters (ratios of absorption at different wavelengths) that are then subjected to regression analysis. By changing the parameter representation from absolute absorption values to normalized ratios, the regression process becomes more robust to uncertainties while maintaining computational feasibility.
Solution Approach 2:
The patent applies regression analysis selectively to the most critical portions of the signal processing where uncertainty has the greatest impact on accuracy. Rather than over-processing all signal aspects, the method focuses computational resources on the modulation ratio determination, achieving improved accuracy without excessive overall 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 improves the reliability and accuracy of blood oxygen saturation measurements by considering uncertainty in both signals, reducing the impact of noise and variations, and providing a more precise estimation of arterial oxygen saturation.
Implementation Method 1
a non-invasive sensor that passes light (e.g., via a light emitting diode) through a portion of a patient's blood perfused tissue and photo-electrically senses (e.g., via a photo-detector) the absorption and scattering of light through the blood perfused tissue
Implementation Method 2
photo-electrically senses (e.g., via a photo-detector) the absorption and scattering of light
Data Source
AI summary
Embodiments of the present invention relate to a system and method for detecting a blood characteristic in a patient. Embodiments of the present invention may comprise detecting a first modulating signal at a first wavelength, detecting a second modulating signal at a second wavelength, and determining a relative amplitude of the first and second modulating signals. Further, embodiments of the present invention may comprise regressing the first and second modulating signals relative to one another, wherein a first uncertainty value in the first modulating signal and a second uncertainty value in the second modulating signal are accommodated.