UHR-Net: An Uncertainty-Aware Hypergraph Refinement Network for Medical Image Segmentation
Shuokun Cheng, Jinghao Shi, Kun Sun
TLDR
UHR-Net improves medical image lesion segmentation by using uncertainty-aware pretraining and hypergraph refinement for ambiguous and small lesions.
Key contributions
- Introduces UO-IC pretraining with geometry-aware augmentation and hard-negative mining for small, ambiguous lesions.
- Designs UGHR block using entropy-based uncertainty maps to guide hypergraph refinement.
- Splits hyperedge prototypes in UGHR into foreground/background groups to decouple higher-order interactions.
- Achieves consistent state-of-the-art performance on five public medical image segmentation benchmarks.
Why it matters
Accurate lesion segmentation is vital for clinical diagnosis and treatment. UHR-Net addresses key challenges like ambiguous boundaries and small lesion detection, offering more stable and precise predictions. This advancement can significantly improve medical imaging analysis.
Original Abstract
Accurate lesion segmentation is crucial for clinical diagnosis and treatment planning. However, lesions often resemble surrounding tissues and exhibit ill-defined boundaries, leading to unstable predictions in boundary/transition regions. Moreover, small-lesion cues can be diluted by multi-scale feature extraction, causing under- or over-segmentation. To address these challenges, we propose an Uncertainty-Aware Hypergraph Refinement Network (UHR-Net). First, we introduce an Uncertainty-Oriented Instance Contrastive (UO-IC) pretraining strategy that couples geometry-aware copy-paste augmentation with hard-negative mining of lesion-like background regions to improve instance-level discrimination for small and visually ambiguous lesions. Second, we design an Uncertainty-Guided Hypergraph Refinement (UGHR) block, which derives an entropy-based uncertainty map from a coarse probability map to guide hypergraph refinement. By splitting hyperedge prototypes into foreground and background groups, UGHR decouples higher-order interactions and improves refinement in ambiguous regions. Experiments on five public benchmarks demonstrate consistent gains over strong baselines. Code is available at: https://github.com/CUGfreshman/UHR-Net.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.