Uncertainty-Guided Dual-Domain Learning for Reliable Skin Lesion Segmentation

TL;DR

This paper introduces UGDD-Net, a novel skin lesion segmentation framework that actively uses prediction uncertainty to improve accuracy and interpretability in challenging cases.

Contribution

UGDD-Net employs a new 'Glance-and-Gaze' mechanism and uncertainty-guided modules to enhance multi-domain learning and calibration in skin lesion segmentation.

Findings

Achieves state-of-the-art results on multiple datasets.

Effectively handles hard samples with improved robustness.

Provides uncertainty maps aligned with expert variability.

Abstract

Accurate skin lesion segmentation is vital for dermoscopic Computer-Aided Diagnosis. However, visual ambiguity and morphological irregularity often defeat spatial modeling, necessitating multi-domain architectures. Existing paradigms frequently overlook the active use of prediction uncertainty, leading to deterministic frameworks that suffer from blind cross-domain fusion and overfit to label noise. To address these issues, we propose the Uncertainty-Guided Dual-Domain Network (UGDD-Net). UGDD-Net introduces a novel "Glance-and-Gaze" mechanism to transform uncertainty into an active guiding signal. Specifically, the Uncertainty-Guided Bi-directional Feature Fusion (UGBFF) module uses pixel-level uncertainty to modulate spatial-spectral interactions. The Uncertainty-Guided Graph Refinement (UGGR) module constructs a topology-aware graph to propagate reliable semantic consensus and refine uncertain nodes. Finally, the Uncertainty-Guided Margin-Adaptive Loss (UGML) enforces strict constraints on confident pixels while relaxing penalties on uncertain ones to improve statistical calibration. Extensive experiments on ISIC2017, ISIC2018, PH2, and HAM10000 datasets demonstrate that UGDD-Net achieves state-of-the-art performance, especially on "Hard Samples". Our uncertainty maps align with expert inter-observer variability, providing robust interpretability for human-machine collaborative diagnosis.