Related Experiment Video
Updated: Nov 2, 2025
Real-Time Cardiac Mapping with a Noninvasive Imageless Electrocardiographic Imaging System
Published on: April 11, 2025
Toward Patient-Specific Prediction of Ablation Strategies for Atrial Fibrillation Using Deep Learning
Marica Muffoletto1,2, Ahmed Qureshi1, Aya Zeidan1
1School of Biomedical Engineering and Imaging Sciences, King's College London, London, United Kingdom.
Insights
This study introduces a novel AI approach combining atrial imaging and computational modeling to predict the best catheter ablation strategy for atrial fibrillation (AF). The AI model shows promise in guiding personalized AF treatment selection.
Area of Science:
- Cardiology
- Medical Imaging
- Artificial Intelligence
Background:
- Atrial fibrillation (AF) is a prevalent arrhythmia with significant morbidity and mortality.
- Catheter ablation (CA) is a primary treatment for AF, but success rates, particularly for persistent AF, remain suboptimal.
- Computational modeling and deep learning show potential for improving CA strategy selection.
Purpose of the Study:
- To develop and validate a novel approach combining image-based computational modeling and deep learning for personalized CA therapy selection in AF patients.
- To train deep convolutional neural network (CNN) classifiers to predict specific CA strategies from patient-specific atrial models derived from MRI data.
Main Methods:
- Trained CNN classifiers using 122 patient LGE-MRI derived atrial images, 157 synthetic images, and outcomes from 558 CA simulations.
- Classifiers were trained to predict pulmonary vein isolation (PVI), rotor-based ablation (Rotor), and fibrosis-based ablation (Fibro) strategies.
- Evaluated classifier performance on training, validation, and an unseen test set, comparing predictions to simulation outcomes.
Main Results:
- High training accuracy (96.22–97.69%) and validation accuracy (78.68–86.50%) were achieved for the CNN classifiers.
- The model achieved 100% prediction success for Rotor and Fibro ablation strategies on the test set.
- The PVI strategy was correctly predicted in 33.33% of cases on the test set.
Conclusions:
- This study demonstrates a proof-of-concept for using deep neural networks trained on patient-specific MRI data and derived models to predict AF ablation strategies.
- The developed technology offers a novel tool to assist in tailoring CA therapy for individual AF patients.
- Further research is warranted to refine prediction accuracy, particularly for the PVI strategy.
Abstract:
Atrial fibrillation (AF) is a common cardiac arrhythmia that affects 1% of the population worldwide and is associated with high levels of morbidity and mortality. Catheter ablation (CA) has become one of the first line treatments for AF, but its success rates are suboptimal, especially in the case of persistent AF. Computational approaches have shown promise in predicting the CA strategy using simulations of atrial models, as well as applying deep learning to atrial images. We propose a novel approach that combines image-based computational modelling of the atria with deep learning classifiers trained on patient-specific atrial models, which can be used to assist in CA therapy selection. Therefore, we trained a deep convolutional neural network (CNN) using a combination of (i) 122 atrial tissue images obtained by unfolding patient LGE-MRI datasets, (ii) 157 additional synthetic images derived from the patient data to enhance the training dataset, and (iii) the outcomes of 558 CA simulations to terminate several AF scenarios in the corresponding image-based atrial models. Four CNN classifiers were trained on this patient-specific dataset balanced using several techniques to predict three common CA strategies from the patient atrial images: pulmonary vein isolation (PVI), rotor-based ablation (Rotor) and fibrosis-based ablation (Fibro). The training accuracy for these classifiers ranged from 96.22 to 97.69%, while the validation accuracy was from 78.68 to 86.50%. After training, the classifiers were applied to predict CA strategies for an unseen holdout test set of atrial images, and the results were compared to outcomes of the respective image-based simulations. The highest success rate was observed in the correct prediction of the Rotor and Fibro strategies (100%), whereas the PVI class was predicted in 33.33% of the cases. In conclusion, this study provides a proof-of-concept that deep neural networks can learn from patient-specific MRI datasets and image-derived models of AF, providing a novel technology to assist in tailoring CA therapy to a patient.
