Multi-Model Ensemble Approach for Accurate Bi-Atrial Segmentation in LGE-MRI of Atrial Fibrillation Patients

TL;DR

This study tackles automatic, multiclass bi-atrial segmentation in LGE-MRI for atrial fibrillation by leveraging an ensemble of four CNN encoders (UNet, ResNet34, EfficientNetB0, and VGG16) within a 2D slice-based framework. Trained on the MBAS 2024 dataset of 200 multi-center scans, the authors build 12 sub-models across three data splits and fuse them with a per-class weighted ensemble to optimize Dice similarity and the 95th percentile Hausdorff distance ($HD_{95}$). The ensemble significantly outperforms individual architectures across left/right atrial walls and cavities, with Dice scores of 88.41% (Wall), 98.48% (RA), and 98.45% (LA), and HD95 of 1.07, 0.95, and 0.64. This approach holds potential to improve AF ablation planning by providing more accurate and robust atrial segmentation from LGE-MRI data.

Abstract

Atrial fibrillation (AF) is the most prevalent form of cardiac arrhythmia and is associated with increased morbidity and mortality. The effectiveness of current clinical interventions for AF is often limited by an incomplete understanding of the atrial anatomical structures that sustain this arrhythmia. Late Gadolinium-Enhanced MRI (LGE-MRI) has emerged as a critical imaging modality for assessing atrial fibrosis and scarring, which are essential markers for predicting the success of ablation procedures in AF patients. The Multi-class Bi-Atrial Segmentation (MBAS) challenge at MICCAI 2024 aims to enhance the segmentation of both left and right atria and their walls using a comprehensive dataset of 200 multi-center 3D LGE-MRIs, labelled by experts. This work presents an ensemble approach that integrates multiple machine learning models, including Unet, ResNet, EfficientNet and VGG, to perform automatic bi-atrial segmentation from LGE-MRI data. The ensemble model was evaluated using the Dice Similarity Coefficient (DSC) and 95% Hausdorff distance (HD95) on the left & right atrium wall, right atrium cavity, and left atrium cavity. On the internal testing dataset, the model achieved a DSC of 88.41%, 98.48%, 98.45% and an HD95 of 1.07, 0.95, 0.64 respectively. This demonstrates the effectiveness of the ensemble model in improving segmentation accuracy. The approach contributes to advancing the understanding of AF and supports the development of more targeted and effective ablation strategies.

Figures (1)

• Figure 1: Outputs of each model (UNet, ResNet, EfficientNet, VGG and Ensemble) for three LGE-MRI inputs (MRI Scan), compared to their respective ground truths (Label). The models UNet, ResNet, EfficientNet and VGG were all trained on dataset split A.