Real-Time Electrogram Dispersion Mapping for Atrial Fibrillation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing technologies for detecting cardiac areas promoting atrial fibrillation, such as Topera, CardioInsight, CARTO, and Ripple Mapping, face challenges in providing real-time, high-definition maps due to catheter insertion difficulties, long computation times, high costs, and complex map interpretations, especially in cases of complex electrical activation.

Innovation Solution

A computer device utilizing a classifier with two classification models, including a gradient boosting based machine learning module and a convolutional neural network, to analyze real-time electrogram signals and generate dispersion maps in real-time, enhancing the detection of cardiac areas associated with atrial fibrillation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If electrograms are analyzed using traditional methods with multiple specialists, then comprehensive analysis can be achieved, but the process is time-consuming and delays treatment

Engineering Contradiction:
Improveanalysis comprehensivenessVSAvoidanalysis time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system divides the complex electrogram analysis task into multiple independent analysis modules, each specializing in detecting specific arrhythmia patterns. These modular analyzers process different aspects of the electrogram simultaneously, enabling comprehensive analysis without requiring sequential review by multiple specialists.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system transforms electrogram signals from their original time-domain representation into frequency-domain representations through Fourier transformation. This parameter transformation enables multiple analysis operations to be performed simultaneously on different frequency components, dramatically reducing analysis time while maintaining comprehensive detection capability.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If Fast Fourier Transform is applied to electrograms, then frequency spectrum analysis is achieved, but computational complexity increases

Engineering Contradiction:
Improvefrequency spectrum resolutionVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system pre-calculates and stores the Fourier transform of the reference electrogram pattern before the actual analysis occurs. During real-time analysis, the system compares incoming electrogram transforms against this pre-computed reference, significantly reducing computational complexity while maintaining high frequency spectrum resolution.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system creates and uses copies of the reference electrogram pattern in the frequency domain, allowing direct comparison with transformed patient electrograms. This copying approach enables efficient pattern recognition without requiring complex real-time calculations, reducing computational burden while preserving spectral analysis precision.

Inventive Principle:
Principle #26Copying

Data Source

PatentEP4291094B1Computer device for real-time analysis of electrograms
Publication Date: 2026.04.29 SUBSTRATE HD
  • EP4291094B1 patent drawingFigure 1
  • EP4291094B1 patent drawingFigure 2

AI summary

A computer device for real-time analysis of electrograms, comprising a memory arranged to receive real-time electrograms signals each originating from one of a plurality of electrodes, a first evaluator comprising an extractor and a gradient boosting based machine learning module, said extractor being arranged to extract a set of features comprising at least one timewise analysis feature and at least one morphological feature from each electrogram signal within a set of electrogram signals, and to feed the resulting sets of features to said gradient boosting based machine learning module trained on data comprising sets of features labelled with a value indicating whether the associated electrogram signal exhibits dispersion and arranged to output for each set of electrogram.