Dual Knowledge-Aware Guidance for Source-Free Domain Adaptive Fundus Image Segmentation

Yu Chen, Hailing Wang, Chunwei Wu, Guitao Cao

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Source-free domain adaptation (SFDA), where only a pre-trained source model is available to adapt to the target domain, has gained widespread application in the medical field. Most existing methods overlook low-quality pseudo-labels, i.e., pseudo-labels with boundary semantic confusion, when learning target domain-specific knowledge, leading to the loss of crucial boundary information. Furthermore, focusing solely on the specific knowledge can drive the model shifts in an uncontrollable direction, resulting in model degradation. To address these issues, we propose Dual Knowledge-aware Guidance (DKG), a novel SFDA method that integrates domain-specific knowledge with domain-invariant knowledge to improve transfer performance. Specifically, the pseudo-label calibration scheme is proposed to reduce semantic bias in high-uncertainty pixels, preserving the boundary information of target domain-specific knowledge. To ensure stable training, we propose a domain-invariant knowledge-based loss strategy, leveraging a confidence-guided mechanism and a consistency constraint. Additionally, we also introduce a dynamic balancing loss to address class imbalance. Extensive experiments on cross-domain fundus image segmentation show that DKG achieves state-of-the-art performance. Code is available at https://github.com/Hanshuqian/DKG

Original languageEnglish
Title of host publicationMedical Image Computing and Computer Assisted Intervention, MICCAI 2025 - 28th International Conference, Proceedings
EditorsJames C. Gee, Jaesung Hong, Carole H. Sudre, Polina Golland, Daniel C. Alexander, Juan Eugenio Iglesias, Archana Venkataraman, Jong Hyo Kim
PublisherSpringer Science and Business Media Deutschland GmbH
Pages183-193
Number of pages11
ISBN (Print)9783032049773
DOIs
StatePublished - 2026
Event28th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2025 - Daejeon, Korea, Republic of
Duration: 23 Sep 202527 Sep 2025

Publication series

NameLecture Notes in Computer Science
Volume15965 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference28th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2025
Country/TerritoryKorea, Republic of
CityDaejeon
Period23/09/2527/09/25

Keywords

  • Domain-invariant knowledge
  • Fundus image
  • Pseudo-label calibration
  • Source-free domain adaptation

Fingerprint

Dive into the research topics of 'Dual Knowledge-Aware Guidance for Source-Free Domain Adaptive Fundus Image Segmentation'. Together they form a unique fingerprint.

Cite this