Few-Shot Left Atrial Wall Segmentation in 3D LGE MRI via Meta-Learning

2026-03-25

AI Summary

Accurate segmentation of the left atrial wall in 3D late gadolinium enhancement MRI remains challenging due to its thin structure, low image contrast, and scarcity of annotated data. To address this, this work proposes a few-shot multi-task segmentation framework that integrates model-agnostic meta-learning (MAML), auxiliary tasks leveraging both left and right atrial cavities, and a boundary-aware composite loss. This approach represents the first integration of multi-task meta-learning with boundary-aware mechanisms, substantially improving generalization and boundary precision for thin anatomical structures under extremely limited annotation. Experimental results demonstrate a Dice score of 0.64 and an HD95 of 5.70 mm in the 5-shot setting—outperforming fine-tuning baselines (Dice: 0.52)—while achieving near fully supervised performance with 20 shots. The method also exhibits robustness under domain shift and on local clinical cohorts.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Computer Vision: Segmentation
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Segmenting the left atrial wall from late gadolinium enhancement magnetic resonance images (MRI) is challenging due to the wall's thin geometry, low contrast, and the scarcity of expert annotations. We propose a Model-Agnostic Meta-Learning (MAML) framework for K-shot (K = 5, 10, 20) 3D left atrial wall segmentation that is meta-trained on the wall task together with auxiliary left atrial and right atrial cavity tasks and uses a boundary-aware composite loss to emphasize thin-structure accuracy. We evaluated MAML segmentation performance on a hold-out test set and assessed robustness under an unseen synthetic shift and on a distinct local cohort. On the hold-out test set, MAML appeared to improve segmentation performance compared to the supervised fine-tuning model, achieving a Dice score (DSC) of 0.64 vs. 0.52 and HD95 of 5.70 vs. 7.60 mm at 5-shot, and approached the fully supervised reference at 20-shot (0.69 vs. 0.71 DSC). Under unseen shift, performance degraded but remained robust: at 5-shot, MAML attained 0.59 DSC and 5.99 mm HD95 on the unseen domain shift and 0.57 DSC and 6.01 mm HD95 on the local cohort, with consistent gains as K increased. These results suggest that more accurate and reliable thin-wall boundaries are achievable in low-shot adaptation, potentially enabling clinical translation with minimal additional labeling for the assessment of atrial remodeling.

Problem

• left atrial wall segmentation
• few-shot learning
• 3D LGE MRI
• thin structure segmentation
• scarce annotations

Innovation

• Meta-Learning
• Few-Shot Segmentation
• Left Atrial Wall
• Boundary-Aware Loss
• 3D LGE MRI