Blood Pressure Estimation Using Normalized Pulsatility Features
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating blood pressure using pulsatility signals are not robust against gain changes, skin colors, and sensor-skin interface issues, leading to inaccurate measurements.
Innovation Solution
Combining time-related and amplitude-related features in a single blood pressure function, with the use of normalized amplitude-related features to enhance independence from electronic drifts and sensor variability, facilitating implementation in low-cost consumer devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If pulsatility signals are used to estimate blood pressure values, then non-invasive measurement is achieved, but measurement precision deteriorates due to amplitude and shape distortions
Solution Approach 1:
The patent transforms the pulsatility signal through mathematical operations (differentiation, normalization) to extract features that correlate with blood pressure. By changing the parameter representation from raw amplitude to derived features like area under curve and signal differentiation, the method achieves accurate blood pressure estimation from non-invasive measurements
Solution Approach 2:
The patent introduces an intermediary processing stage that converts the pulsatility signal into blood pressure estimates through a calibrated model. This intermediary transformation layer bridges the gap between the easily measurable pulsatility signal and the clinically relevant blood pressure values, resolving the contradiction between ease of measurement and precision
2Measurement precision
If conventional pulsatility signal features are used, then blood pressure estimation is achieved, but reliability deteriorates due to gain changes and sensor variability
Solution Approach 1:
The patent implements self-calibration by using the pulsatility signal itself to determine calibration parameters through signal differentiation and area calculation. The system automatically adapts to gain changes and sensor variability by deriving calibration factors from the signal's own characteristics, making the measurement reliable without external reference
Solution Approach 2:
The patent employs dynamic normalization where calibration parameters are not fixed but are continuously adapted based on the incoming pulsatility signal characteristics. This dynamic adjustment allows the system to maintain reliability despite variations in sensor gain, skin properties, and measurement conditions
3Measurement precision
If amplitude-related features are used alone, then blood pressure values can be calculated, but independence from electronic drifts and skin colors is lost
Solution Approach 1:
The patent performs preliminary normalization of the pulsatility signal by calculating the area under the curve and using it to scale the signal before extraction of blood pressure features. This preliminary action removes dependencies on absolute amplitude, skin color, and electronic drifts, allowing subsequent blood pressure calculation to be independent of these variables
Data Source
Figure 1~2
Figure 3~5
Figure 6~9
AI summary
Method for determining a blood pressure value comprising the steps of: providing a pulsatility signal (1), determining a time-related feature and a normalized amplitude-related feature on the basis of the pulsatility signal; and calculating a blood pressure value on the basis of a blood pressure function depending on the time-related feature, the normalized amplitude-related feature and function parameters.