Arterial Pressure Wavelet Transform for Hypotension Trend Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional hypotension prediction methods fail to effectively reflect trends in blood pressure changes, which are crucial for identifying the cause of hypotension during surgery, as they primarily rely on monitoring raw arterial blood pressure values.
Innovation Solution
An arterial pressure wavelet transform-based apparatus and method that uses a pre-trained hypotension prediction model to determine hypotension by extracting trend data from compressed arterial blood pressure data through wavelet transformation, shapelet data generation, and calculating similarity features, allowing for the prediction of hypotension based on changes in each measurement interval.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional methods monitor raw arterial blood pressure values, then the monitoring process is simple, but the ability to reflect trends in blood pressure changes is insufficient
Solution Approach 1:
The patent extracts trend information from raw arterial blood pressure signals by applying wavelet transform to decompose the signal into different frequency components. This extraction process isolates the relevant trend data while filtering out noise, thereby preserving important information without requiring overly complex processing systems.
Solution Approach 2:
The patent transforms the raw blood pressure signal from time domain to frequency domain using wavelet transform, changing the representation parameters of the signal. This parameter transformation enables the extraction of trend information that is not apparent in the raw time-series data, improving information retention while managing complexity through mathematical transformation.
2Measurement precision
If wavelet transform is applied to extract trend data, then the prediction accuracy of hypotension is improved, but the computational complexity increases
Solution Approach 1:
The patent segments the continuous blood pressure signal into discrete intervals and applies wavelet transform to each segment. This segmentation approach breaks down the complex computational task into manageable pieces, improving prediction accuracy through localized analysis while reducing overall computational complexity by processing smaller data segments independently.
Solution Approach 2:
The patent performs preliminary wavelet transformation and trend extraction on training data before actual hypotension prediction. This preliminary action pre-computes the trend features and stores them for comparison, thereby improving real-time prediction accuracy while reducing the computational burden during actual monitoring by reusing pre-processed features.
3Reliability
If shapelet data generation is used to identify characteristic patterns, then the detection capability of hypotension trends is improved, but the data processing time is increased
Solution Approach 1:
The patent generates shapelet data for the most significant frequency components and time intervals rather than processing all possible patterns. This partial action approach focuses computational resources on the most relevant features for hypotension detection, improving reliability by concentrating on critical patterns while reducing data processing time by excluding less relevant computations.
Solution Approach 2:
The patent creates shapelet data as simplified representations or copies of complex blood pressure patterns. These shapelet copies capture the essential characteristics of hypotension trends without requiring full reconstruction of the original complex signals, thereby improving detection reliability through pattern matching while reducing processing time by working with compressed representations.
Data Source
AI summary
There is provided an apparatus for predicting hypotension of a subject. The apparatus comprises a memory configured to store one or more instructions and a pre-trained hypotension prediction model; and a processor configured to execute the one or more instructions stored in the memory, wherein the instructions, when executed by the processor, cause the processor to: determine an arterial blood pressure data of the subject, input the arterial blood pressure data of the subject into the hypotension prediction model, and determine whether the subject has hypotension using an output result of the hypotension prediction model.