AI Cardiac Arrhythmia Source Detection via Optical Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current clinical electrode approaches, such as multi-electrode mapping, struggle to accurately detect transmural conduction within the 3-dimensional structure of the human atria, leading to false positives and negatives in identifying cardiac arrhythmia sources, particularly for atrial fibrillation, hindering effective ablation treatments.
Innovation Solution
A computer-implemented method using a pre-trained artificial intelligence model that combines electrogram signals with co-registered functional and structural imaging data to predict the location of cardiac arrhythmia sources, including atrial fibrillation drivers, by learning characterizing features from explanted human hearts and applying them to patient-specific data for accurate source localization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multi-electrode mapping is used to record electrical signals from the heart surface, then the coverage area is improved, but the measurement precision of transmural conduction detection deteriorates
Solution Approach 1:
The patent introduces optical mapping as an intermediary technique to bridge the gap between surface electrode recordings and deep tissue activity. Optical mapping agents (fluorescent dyes or genetically encoded indicators) serve as mediators that allow non-invasive visualization of transmural conduction patterns, enabling precise detection of arrhythmia sources without requiring invasive intramural electrode placement.
Solution Approach 2:
The patent replaces the mechanical electrode-based detection system with an optical detection system. Instead of relying on physical contact between electrodes and cardiac tissue to detect electrical signals, the system uses optical imaging techniques (fluorescence, bioluminescence) to detect electrical activity through light emission or absorption changes, thereby overcoming the limitations of surface electrode coverage.
2Productivity
If artificial intelligence algorithms are applied to classify ECG recordings, then the classification speed is improved, but the reliability of arrhythmia source detection deteriorates due to lack of gold standard validation
Solution Approach 1:
The patent performs preliminary validation of AI algorithms using optical mapping data obtained from explanted hearts before clinical deployment. By establishing ground truth arrhythmia source locations through optical mapping in a controlled ex-vivo setting, the system creates a validated training dataset that enables reliable AI classification in subsequent clinical applications.
Solution Approach 2:
The patent creates a simplified copy of the clinical detection problem in an ex-vivo experimental setting. By reproducing arrhythmia conditions in explanted hearts and obtaining both electrode recordings and optical mapping data simultaneously, the system creates a controlled environment to validate AI algorithms before applying them to complex clinical scenarios.
3Difficulty of detecting and measuring
If computational simulations are used to identify reentrant AF drivers, then the detection capability is improved, but the adaptability to clinical settings deteriorates due to lack of translational validation
Solution Approach 1:
The patent uses optical mapping data from explanted hearts to automatically validate and refine computational simulation models. The experimental data serves as ground truth that the simulation models must reproduce, enabling self-validation of the computational approaches and improving their accuracy for clinical applications without requiring extensive manual calibration.
Data Source
AI summary
Disclosed are various embodiments of methods, components and systems configured to determine a location of a source of cardiac arrhythmia in a patient's heart. In some embodiments, to determine a source location, electrogram signals are acquired from a region of the patients' heart using a first set of electrodes; and then a pre-trained artificial intelligence (AI) model is applied to predict the location of the cardiac arrhythmia source by using the signals. Importantly, pre-training of the AI model comprises acquiring electrogram signals from explanted human hearts, the signals are generated by a second set of electrodes assembled into an electrode array that covers at least a part of the explanted human heart, and acquiring co-registered functional and/or structural imaging data in the part of the explanted human heart covered with the electrode array, wherein the functional and/or structural imaging data provide location of at least one source of cardiac arrhythmia.