Hierarchical Corpus-View-Category Refinement for Carotid Plaque Risk Grading in Ultrasound

2025-06-29

AI Summary

To address the challenges of small plaque volume, high intra-class variability, and insufficient multi-view collaborative modeling in carotid plaque (CP) risk stratification, this paper proposes the first multi-view ultrasound analysis framework aligned with the Carotid Plaque-RADS guidelines. Methodologically, it introduces a three-stage progressive refinement architecture: (i) a center-memory contrastive loss to enhance inter-class discriminability; (ii) a cascaded downsampling attention module to strengthen feature representation for small targets; and (iii) a parameter-free Mixture-of-Experts weighting strategy to achieve decoupled yet synergistic feature fusion across views, samples, and classes. Extensive experiments demonstrate state-of-the-art performance on CP grading, with significant improvements in global representation capability and generalization robustness.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Computer Vision: Multi-modal Vision
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Accurate carotid plaque grading (CPG) is vital to assess the risk of cardiovascular and cerebrovascular diseases. Due to the small size and high intra-class variability of plaque, CPG is commonly evaluated using a combination of transverse and longitudinal ultrasound views in clinical practice. However, most existing deep learning-based multi-view classification methods focus on feature fusion across different views, neglecting the importance of representation learning and the difference in class features. To address these issues, we propose a novel Corpus-View-Category Refinement Framework (CVC-RF) that processes information from Corpus-, View-, and Category-levels, enhancing model performance. Our contribution is four-fold. First, to the best of our knowledge, we are the foremost deep learning-based method for CPG according to the latest Carotid Plaque-RADS guidelines. Second, we propose a novel center-memory contrastive loss, which enhances the network's global modeling capability by comparing with representative cluster centers and diverse negative samples at the Corpus level. Third, we design a cascaded down-sampling attention module to fuse multi-scale information and achieve implicit feature interaction at the View level. Finally, a parameter-free mixture-of-experts weighting strategy is introduced to leverage class clustering knowledge to weight different experts, enabling feature decoupling at the Category level. Experimental results indicate that CVC-RF effectively models global features via multi-level refinement, achieving state-of-the-art performance in the challenging CPG task.

Problem

• Improves carotid plaque grading accuracy in ultrasound
• Addresses multi-view feature fusion and representation learning
• Enhances global modeling with multi-level refinement framework

Innovation

• Center-memory contrastive loss enhances global modeling
• Cascaded down-sampling attention fuses multi-scale information
• Parameter-free mixture-of-experts weighting enables feature decoupling