Key result
HSCM-Net achieves a 0.577 Dice score for myocardial pathology segmentation on unseen multisequence CMR images.
Why the study?
Existing deep learning methods for myocardial pathology segmentation show limited generalizability across domains due to distribution shifts and difficulty learning domain-invariant pathology information.
The proposed HSCM-Net framework improves the generalizability of myocardial pathology segmentation on multisequence CMR images across different domains.
Answers are grounded in this paper and cite their sources.
May aid cross-domain MI severity assessment via CMR; extends DL generalizability beyond single-domain pathology segmentation.
Myocardial pathology segmentation (MyoPS) aims to accurately quantify myocardial scar and edema, which is critical for precise assessment of myocardial infarction (MI) severity. Recent deep learning (DL) methods have shown promising performance. However, due to cross-domain distribution shifts, many methods exhibit limited generalizability because it is challenging to learn domain-invariant pathology information, resulting in degraded performance on unseen data. This work proposes a hierarchical semantic concept segmentation framework, referred to as HSCM-Net. This framework uses concept modeling of multisequence cardiac magnetic resonance (CMR) images and labels to integrate complementary information and enhance the generalizability of MyoPS. Specifically, the CMR images are first decomposed into three independent concept variables, and hierarchical priors are assigned to explicitly force them to model the sequence-irrelevant concept of shape, the semantic concept of pathology, and the domain-specific concept of appearance. Then, the anatomy segmentation and pathology segmentation are modeled as global and local concepts only related to shape and pathology, respectively. Finally, a variational inference (VI) framework is developed to approximate the posteriors of these conceptual variables, and the inference process is implemented with a deep neural network. Results on three-domain multisequence CMR datasets proved the advantageous generalizability of our HSCM-Net, which achieves a MyoPS Dice score of 0.577 on unseen target domains. Our code is released at https://github.com/Dongjinweicn/HSCMNet.
No takes yet. Share an insight, caveat, or question.
Dong et al. (2026) studied Myocardial infarction. HSCM-Net (hierarchical semantic concept segmentation framework) was evaluated on MyoPS Dice score on unseen target domains. The HSCM-Net framework achieved a MyoPS Dice score of 0.577 on unseen target domains, demonstrating generalizability for myocardial pathology segmentation on multisequence CMR images.
Synapse has enriched 10 related papers. Consider them for comparative context: