Neural Network EP Mapping Basket Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current diagnostic tools for cardiac rhythm disorders, particularly atrial fibrillation, lack the precision needed to accurately identify the sources and locations of AF drivers, leading to inadequate treatment outcomes and increased morbidity due to insufficient spatial and temporal resolution, as well as artifacts in intracardiac signals.
Innovation Solution
A method and system utilizing a neural network architecture to process intracardiac and body surface electrogram signals, isolating ventricular signals, and reconstructing the three-dimensional positions of an EP mapping basket within the heart, allowing for precise localization of cardiac rhythm disorder sources without additional hardware, thus minimizing noise and improving treatment accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional diagnostic tools are used for cardiac rhythm disorders, then the treatment can be performed, but the precision of identifying AF driver sources and locations is insufficient
Solution Approach 1:
The patent replaces traditional hardware-based navigation systems with a neural network-based computational system that processes intracardiac electrogram signals to reconstruct three-dimensional catheter positions and localize AF drivers. This substitution of mechanical/electrical hardware systems with an information-processing system eliminates hardware-induced artifacts while achieving superior localization precision through advanced signal analysis and machine learning algorithms.
Solution Approach 2:
The patent combines multiple signal sources (intracardiac electrogram signals from multiple electrodes) and integrates them through a neural network to create a composite representation of cardiac electrical activity. This composite approach synthesizes spatial and temporal information from distributed sensors to achieve high-resolution localization that exceeds the capabilities of individual measurement points.
2Measurement precision
If additional hardware is added to improve signal resolution, then measurement precision may improve, but device complexity and noise increase
Solution Approach 1:
The patent replaces complex hardware systems (navigation catheters, electromagnetic tracking systems, additional sensors) with a computational neural network that processes existing intracardiac electrogram signals. This substitution achieves enhanced measurement precision without adding physical complexity, as the neural network extracts spatial-temporal patterns from standard electrode recordings to reconstruct three-dimensional catheter positions and localize AF drivers.
Solution Approach 2:
The neural network system utilizes the existing intracardiac electrogram signals already being recorded during the procedure to simultaneously perform multiple functions: localizing the catheter position, reconstructing three-dimensional anatomy, and identifying AF driver sources. This self-service approach extracts maximal information from the available signals without requiring additional hardware inputs.
3Reliability
If traditional mapping systems are used, then the procedure can be completed, but artifacts in intracardiac signals reduce treatment accuracy
Solution Approach 1:
The patent eliminates hardware-induced artifacts by replacing traditional navigation and mapping hardware with a neural network-based computational system. Since the neural network processes electrical signals without introducing electromagnetic fields or mechanical interference, it avoids the artifacts generated by conventional navigation catheters and tracking systems, thereby improving signal fidelity and treatment accuracy.
Solution Approach 2:
The neural network extracts and isolates the relevant cardiac electrical signals from the intracardiac electrogram recordings, separating the useful diagnostic information from background noise and artifacts. Through advanced signal processing and pattern recognition, the system extracts precise localization data and AF driver identification while filtering out harmful artifacts that would otherwise compromise treatment accuracy.
Data Source
Figure 1A
Figure 1B
Figure 2A
AI summary
In some embodiments, there are provided systems, devices, components, and corresponding methods configured to permit navigation and/or positioning of an intra-cardiac electrophysiological (EP) mapping basket or other EP mapping structure of an EP mapping catheter inside or near an atrium or other heart chamber of a patient's heart using biosignals or intra-cardiac signals. In one embodiment, QRS complexes are extracted or isolated from intra-cardiac signals sensed by electrodes mounted on the EP mapping basket. Using the QRS complexes and a statistical shape or other model of the EP mapping basket or other type of EP mapping structure, one or more computing devices then determine the locations of the electrodes inside or near the patient's atrium that are associated with each isolated or extracted QRS complex, and thereby permit accurate navigation within the heart and/or processing of data acquired using the EP mapping basket or other EP mapping structure. The one or more computing devices can also be used to determine changes in the three-dimensional locations and orientations of the basket and the electrodes thereof as the EP mapping basket is moved around, in, or near the patient's atrium, heart chamber, or other portion of the patient's heart, and to display to a user multiple positions of the basket inside or near the patient's heart.