CalcSeg: Confidence-aware 3D Latent Context Curriculum Learning For Myocardial Scar Segmentation From Single-Stack LGE-CMRs
2026-08-20 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors address the difficult task of identifying heart muscle scars from special heart MRI images that usually show only a few thin slices. They develop a new method called CalcSeg which learns step-by-step, starting with easy cases and moving to harder ones, using a system that estimates how confident it is about each prediction. To get around the problem of limited 3D information from these thin slices, their method uses a technique to understand relationships between slices and create a fuller 3D picture. They tested CalcSeg on real clinical data and found it works better than previous methods, especially for tough cases.
Myocardial scar segmentationLate gadolinium-enhanced cardiac magnetic resonance (LGE-CMR)Single-stack 2D imagingCurriculum learningConfidence-aware scoringEpistemic uncertaintyScar burden estimationSelf-attention3D spatial contextSemi-supervised learning
Authors
Nivetha Jayakumar, Hannah Kim, Amit R. Patel, Miaomiao Zhang
Abstract
Myocardial scar segmentation from single-stack late gadolinium-enhanced cardiac magnetic resonance (LGE-CMR) imaging has been a longstanding and clinically important challenge, particularly in the presence of low tissue contrast, diffuse, and small scar regions. These challenges are further intensified by the limited availability of 3D spatial context. This paper presents CalcSeg, a Confidence-aware latent context curriculum learning framework that leverages fused 3D feature representations from single-stack 2D LGE-CMR images for robust scar segmentation. Specifically, we introduce a dynamic semi-supervised curriculum learning strategy that progressively expands training from easier to more challenging scar cases using a learned confidence-aware scoring function. Such a function integrates errors in the predicted scar maps with quantified epistemic uncertainty and scar burden estimation to automatically assess sample difficulty without requiring manual labels. To compensate for the limited spatial context in single-stack acquisitions, we then develop a latent slice-wise self-attention to capture inter-slice dependencies and infer 3D spatial representations from sparse 2D inputs. We evaluate CalcSeg on multi-center clinical LGE-CMR datasets and benchmark against existing scar segmentation networks. Experimental results show that CalcSeg consistently outperforms all competing methods, particularly with substantial improvements on clinically challenging cases. Our code is released on Github.