IarCAC: Instance-Aware Representation for Coronary Artery Calcification Segmentation in Cardiac CT Angiography

Weili Jiang, Yiming Li, Zhang Yi, Jianyong Wang, Mao Chen

Abstract: Coronary Artery Calcification (CAC) is a robust indicator of coronary artery disease and a critical determinant of percutaneous coronary intervention outcomes. Our method is inspired by a clinical observation that CAC typically manifests as a sparse distribution of multiple instances. Existing methods focusing solely on spatial correlation overlook the sparse spatial distribution of semantic connections in CAC tasks. Motivated by this, we introduce a novel instance-aware representation method for CAC segmentation, termed IarCAC, which explicitly leverages the sparse connectivity pattern among instances to enhance the model’s instance discrimination capability. The proposed IarCAC first develops an InstanceViT module, which assesses the connection strength between each pair of tokens, enabling the model to learn instance-specific attention patterns. Subsequently, an instance-aware guided module is introduced to learn sparse high-resolution representations over instance-dependent regions in the Fourier domain. To evaluate the effectiveness of the proposed method, we conducted experiments on two challenging CAC datasets and achieved state-of-the-art performance across all datasets. The code is available at https://github.com/WeiliJiang/IarCAC.