Cuffless Blood Pressure Estimation via Phase-Specific Deep Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cuffless blood pressure estimation methods lack accuracy and reliability, particularly in varying physiological conditions, due to the complexity of human blood pressure dynamics.
Innovation Solution
An apparatus and method utilizing a sensor to measure pulse wave signals and a processor to classify the phase of mean arterial pressure (MAP) using deep learning-based estimation models, allowing for accurate systolic blood pressure (SBP) estimation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single unified estimation model is used for blood pressure measurement across all conditions, then the device complexity is reduced, but the measurement precision deteriorates due to varying physiological conditions
Solution Approach 1:
The patent divides the blood pressure estimation problem into multiple segments by creating separate estimation models for different physiological conditions (resting state, exercise state, recovery state). Each model is specialized for its specific condition, improving measurement precision without requiring a single complex universal model. The segmentation is based on classifying the current physiological state and selecting the appropriate pre-trained model.
Solution Approach 2:
The system changes the parameter of model selection based on the detected physiological state. By monitoring physiological parameters (such as heart rate, pulse wave characteristics) and classifying the current state, the system dynamically selects which pre-trained estimation model to use, thereby adapting to varying conditions without increasing device complexity.
2Measurement precision
If multiple pre-trained estimation models are maintained for different physiological conditions, then the measurement precision improves, but the loss of information increases due to the need to select the appropriate model
Solution Approach 1:
The system performs preliminary classification of the physiological state before selecting the estimation model. By analyzing pulse wave signals and other physiological parameters to determine whether the subject is in resting, exercise, or recovery state, the system prepares the appropriate model in advance, ensuring accurate blood pressure estimation without losing critical physiological information.
Solution Approach 2:
The system uses feedback from physiological signal analysis to continuously monitor and classify the current state. By comparing real-time pulse wave characteristics against known patterns for different states, the system provides feedback that confirms or adjusts the selected estimation model, minimizing information loss and ensuring accurate measurements.
3Productivity
If deep learning-based estimation models are used for each phase of MAP, then the productivity of blood pressure monitoring is improved, but the device complexity increases due to multiple specialized models
Solution Approach 1:
The patent segments the blood pressure monitoring task into distinct physiological phases (resting, exercise, recovery) with dedicated deep learning models for each. This segmentation allows the system to process different types of physiological data efficiently with specialized models, improving overall monitoring productivity while managing complexity through modular architecture.
Solution Approach 2:
The system achieves universality by creating a multi-functional estimation framework where a single platform supports multiple pre-trained models that can handle different physiological conditions. The underlying architecture remains universal, allowing any of the specialized models to be selected and applied based on the current state, thereby improving productivity without proportionally increasing device complexity.
Data Source
AI summary
An apparatus for and a method for estimating blood pressure are provided. The apparatus for estimating blood pressure includes: a sensor configured to measure a pulse wave signal from an object; and a processor configured to obtain a mean arterial pressure (MAP) based on the pulse wave signal, configured to classify a phase of the obtained MAP according to at least one classification criterion, and to obtain a systolic blood pressure (SBP) by using an estimation model corresponding to the classified phase of the MAP among estimation models corresponding to respective phases of the MAP.