Cardiac Signal Processing for Real-Time Atrial Tachycardia Origin Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting the origin of atrial tachycardia (AT) in real-time are inadequate due to reliance on a posteriori processing of P wave analysis, which is not feasible in real-time contexts, limiting their effectiveness in clinical applications.
Innovation Solution
A device utilizing a random convolutional kernel-based extractor and machine-learning locator, employing decision trees, processes electrocardiogram (ECG) and coronary sinus signals to determine the probable area of AT origin in real-time, generating over 1000 features for precise classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a posteriori processing of P wave analysis is used to determine AT origin, then measurement precision is improved, but productivity deteriorates due to inability to operate in real-time
Solution Approach 1:
The patent pre-calculates and stores convolution kernels before real-time processing. These kernels are computed offline and saved for rapid application during clinical procedures, enabling real-time operation without sacrificing analysis depth
Solution Approach 2:
The patent replaces traditional mechanical signal processing methods with machine learning-based classification algorithms. The system uses trained classifiers that automatically identify AT origins from P wave patterns, substituting manual or rule-based analysis with intelligent automated systems that achieve both speed and accuracy
2Measurement precision
If comprehensive P wave analysis is performed to accurately locate AT origin, then measurement precision is improved, but loss of time increases due to complex processing requirements
Solution Approach 1:
The system performs preliminary training of machine learning models using comprehensive P wave datasets before deployment. During actual use, the pre-trained models rapidly classify new signals without requiring time-consuming re-analysis, thus reducing operational time while maintaining high precision
Solution Approach 2:
The patent creates simplified representations of complex P wave patterns through trained classification models. These models capture the essential features of AT origins from comprehensive training data, allowing rapid inference without replicating the full complexity of the original analysis process
Data Source
Figure 1~3
Figure 4
AI summary
A device for processing cardiac signals, comprises: - a data storage (114) arranged to receive input data sets each comprising a plurality of P wave segments each associated with an electrocardiogram track and with an acquisition time window, and a plurality of coronary sinus signals associated with the same acquisition time window and having one or more activation sequence(s), - a random convolutional kernel-based extractor arranged, for a given input data set, to determine an electrocardiogram feature vector, and - a machine-learning based locator using decision trees trained on data comprising sets of electrocardiogram feature vectors labelled with a value indicating a cardiac region identifier, and arranged to receive an electrocardiogram feature vector from the random convolutional kernel-based extractor as input, and to return a cardiac region identifier as output.