Composite Neural Network for Arrhythmia Estimation Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional wearable monitoring devices using artificial neural networks struggle to accurately estimate arrhythmia on a beat segment basis, such as atrial premature contraction, ventricular premature contraction, left bundle branch block, and right bundle branch block, due to limitations in identifying the number or proportion of arrhythmia in ECG signals.
Innovation Solution
A composite artificial neural network system that combines two neural networks: one trained to estimate arrhythmia types on a beat segment basis and another for rhythm changes, with a verification unit to correct incompatible class estimates, improving accuracy by mutual verification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single artificial neural network is used to estimate arrhythmia type from ECG signal sections, then the device complexity is low, but the measurement precision of arrhythmia estimation deteriorates
Solution Approach 1:
The patent divides the arrhythmia estimation task into two separate neural networks: one specialized in beat segment-level classification and another in section-level rhythm classification. This segmentation allows each network to focus on specific features, improving overall estimation accuracy while maintaining manageable complexity through modular design
Solution Approach 2:
The patent combines the outputs of two specialized neural networks through a verification unit that integrates beat segment estimates with section-level rhythm estimates. This merging of multiple estimation sources improves measurement precision by cross-validating results and resolving inconsistencies between different classification levels
2Reliability
If a single artificial neural network estimates arrhythmia on ECG signal sections, then the device complexity is low, but the reliability of beat segment arrhythmia identification deteriorates
Solution Approach 1:
The verification unit implements a feedback mechanism where section-level rhythm estimates are used to validate and correct beat segment-level estimates. When inconsistencies are detected between the two estimation levels, the system adjusts the beat segment classification to ensure reliability, creating a self-correcting system that improves identification accuracy
Solution Approach 2:
The system performs preliminary beat segment classification before final section-level validation. This preliminary action allows the verification unit to compare initial estimates against section-level patterns, ensuring that beat segment arrhythmia identifications are reliable before finalizing the diagnosis
Data Source
AI summary
A method for estimating arrhythmia using a composite artificial neural network includes the steps of estimating a class corresponding to a beat segment included in a first section of an electrocardiogram (ECG) signal, using a first artificial neural network; estimating a class corresponding to the first section of the ECG signal, using a second artificial neural network; and mutually verifying the estimated class corresponding to the beat segment included in the first section of the ECG signal and the estimated class corresponding to the first section of the ECG signal.