BPC-Net: Annotation-Free Skin Lesion Segmentation via Boundary Probability Calibration

2026-04-07

AI Summary

This work addresses the challenges of incomplete lesion boundaries in unsupervised skin lesion segmentation, which stem from underconfident boundary probabilities, noisy pseudo-labels, and unstable cross-domain transfer. To tackle these issues, the authors propose BPC-Net, a novel framework that explicitly models and calibrates boundary probabilities through Gaussian probability smoothing to achieve precise boundary recovery. Additionally, BPC-Net incorporates a feature-disentangled decoder and an adaptive transfer strategy that updates only the interaction branch, enhancing segmentation accuracy and deployment stability under fully unsupervised conditions. Evaluated on ISIC-2017, ISIC-2018, and PH2 datasets, BPC-Net achieves state-of-the-art performance among unsupervised methods, with macro-averaged Dice coefficient and Jaccard index reaching 85.80% and 76.97%, respectively—approaching the performance of supervised methods on PH2.

Abstract

Annotation-free skin lesion segmentation is attractive for low-resource dermoscopic deployment. However, its performance remains constrained by three coupled challenges: noisy pseudo-label supervision, unstable transfer under limited target-domain data, and boundary probability under-confidence. Most existing annotation-free methods primarily focus on pseudo-label denoising. In contrast, the effect of compressed boundary probabilities on final mask quality has received less explicit attention, although it directly affects contour completeness and cannot be adequately corrected by global threshold adjustment alone. To address this issue, we propose BPC-Net, a boundary probability calibration framework for annotation-free skin lesion segmentation. The core of the framework is Gaussian Probability Smoothing (GPS), which performs localized probability-space calibration before thresholding to recover under-confident lesion boundaries without inducing indiscriminate foreground expansion. To support this calibration under noisy pseudo-supervision and cross-domain transfer, we further incorporate two auxiliary designs: a feature-decoupled decoder that separately handles context suppression, detail recovery, and boundary refinement, and an interaction-branch adaptation strategy that updates only the pseudo-label interaction branch while preserving the deployed image-only segmentation path. Under a strictly annotation-free protocol, no manual masks are used during training or target-domain adaptation, and validation labels, when available, are used only for final operating-point selection. Experiments on ISIC-2017, ISIC-2018, and PH2 show that the proposed framework achieves state-of-the-art performance among published unsupervised methods, reaching a macro-average Dice coefficient and Jaccard index of 85.80\% and 76.97\%, respectively, while approaching supervised reference performance on PH2.

Problem

• annotation-free
• skin lesion segmentation
• boundary probability
• pseudo-label noise
• domain adaptation

Innovation

• Boundary Probability Calibration
• Gaussian Probability Smoothing
• Annotation-Free Segmentation
• Feature-Decoupled Decoder
• Unsupervised Domain Adaptation