Bayesian Inference for Blood Pressure Trend Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating blood pressure trends using pulse wave velocity (PWV) and other physiological surrogates are hindered by measurement errors, particularly in contactless and short-time window measurements, which affect the reliability of the results.
Innovation Solution
A computer-implemented method using Bayesian inference to analyze blood pressure surrogate measurement values and their error values, correcting for circadian rhythms and steady-state conditions to improve trend estimation, incorporating error correction based on heart rate, body temperature, and physical activity levels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If classical Least-Means-Square (LMS) method is used to infer blood pressure trend from PAT, PTT and/or PWV measurements, then the trend estimation can be obtained, but the measurement errors are not taken into account which reduces the reliability of the results
Solution Approach 1:
The patent changes the analytical approach from classical LMS method to Bayesian inference method. This parameter change in the estimation methodology allows incorporation of measurement error information (standard deviations) into the trend estimation process, thereby improving reliability while accounting for measurement precision limitations
Solution Approach 2:
The Bayesian inference method uses feedback from measurement error data (standard deviations of PAT, PTT and/or PWV measurements) to adjust the trend estimation. The measurement precision information feeds back into the estimation process to weight measurements appropriately, improving overall reliability of the blood pressure trend estimation
2Ease of operation
If contactless measurement methods are used to obtain PAT, PTT and/or PWV measurements, then unobtrusive monitoring is achieved, but measurement errors increase particularly in short time window measurements
Solution Approach 1:
The patent converts the harmful effect of measurement errors into a beneficial factor by explicitly incorporating the error information (standard deviations) into the Bayesian inference process. The measurement errors that result from contactless methods are transformed into useful data for weighting and adjusting the trend estimation, thereby maintaining ease of operation while improving accuracy
Solution Approach 2:
The patent applies parameter changes by modifying the estimation methodology to include error parameters. The standard deviations of measurements are incorporated as additional parameters in the Bayesian framework, allowing the system to adapt to the reduced precision of contactless measurements while maintaining operational ease
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
Enhances the accuracy of blood pressure trend estimation by accounting for measurement errors and circadian variations, providing a more reliable indicator of blood pressure trends.
Implementation Method 1
a photodetector to generate an electrical signal in response to the light
Data Source
Figure 1(a)~1(b)
Figure 2~3
Figure 4
AI summary
According to an aspect, there is provided a computer-implemented method of estimating a trend in a blood pressure surrogate, the method comprising obtaining a set of blood pressure surrogate measurement values of a blood pressure surrogate for a subject; obtaining, for each blood pressure surrogate measurement value, an error value indicating a measurement error for the blood pressure surrogate measurement value; and analysing the set of blood pressure surrogate measurement values and the respective error values using Bayesian inference to determine a trend (4) in the blood pressure surrogate over time.