Search papers, labs, and topics across Lattice.
This paper introduces CalcSeg, a novel framework for myocardial scar segmentation that addresses the challenges of low tissue contrast and limited 3D spatial context in single-stack LGE-CMR imaging. By employing a confidence-aware scoring function within a dynamic semi-supervised curriculum learning strategy, CalcSeg effectively prioritizes training on more difficult cases while leveraging inter-slice dependencies through latent slice-wise self-attention. Experimental results demonstrate that CalcSeg significantly outperforms existing segmentation methods, particularly in clinically challenging scenarios, highlighting its potential for improving diagnostic accuracy in cardiac imaging.
CalcSeg achieves remarkable improvements in myocardial scar segmentation by intelligently prioritizing difficult cases, outperforming existing methods even in low-contrast environments.
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.
