PSScreen V2: Partially Supervised Multiple Retinal Disease Screening

2025-10-26

AI Summary

To address the joint screening of retinal diseases under conditions of multi-source heterogeneous fundus image data with both label scarcity and domain shift, this paper proposes a partially supervised self-training framework. Methodologically, we design a three-branch teacher-student architecture incorporating low-frequency feature enhancement: (i) Low-Frequency Dropout (LF-Dropout) to suppress high-frequency noise interference, and (ii) Low-Frequency Uncertainty estimation (LF-Uncert) via adversarial Gaussian perturbations to improve cross-domain robustness—compatible with vision foundation models such as DINOv2. Extensive experiments on multiple fundus datasets demonstrate substantial improvements over state-of-the-art partially supervised and domain-adaptive methods, achieving SOTA performance both in-domain and out-of-domain. Further transfer validation on chest X-ray data confirms the framework’s strong generalizability and broad applicability across medical imaging modalities.

Technology Category

• Computer Vision: Other Foundations of Computer Vision
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Security and Privacy: Data transparency and provenance
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

In this work, we propose PSScreen V2, a partially supervised self-training framework for multiple retinal disease screening. Unlike previous methods that rely on fully labelled or single-domain datasets, PSScreen V2 is designed to learn from multiple partially labelled datasets with different distributions, addressing both label absence and domain shift challenges. To this end, PSScreen V2 adopts a three-branch architecture with one teacher and two student networks. The teacher branch generates pseudo labels from weakly augmented images to address missing labels, while the two student branches introduce novel feature augmentation strategies: Low-Frequency Dropout (LF-Dropout), which enhances domain robustness by randomly discarding domain-related low-frequency components, and Low-Frequency Uncertainty (LF-Uncert), which estimates uncertain domain variability via adversarially learned Gaussian perturbations of low-frequency statistics. Extensive experiments on multiple in-domain and out-of-domain fundus datasets demonstrate that PSScreen V2 achieves state-of-the-art performance and superior domain generalization ability. Furthermore, compatibility tests with diverse backbones, including the vision foundation model DINOv2, as well as evaluations on chest X-ray datasets, highlight the universality and adaptability of the proposed framework. The codes are available at https://github.com/boyiZheng99/PSScreen_V2.

Problem

• Screens multiple retinal diseases using partially labeled datasets
• Addresses label absence and domain shift in medical imaging
• Enhances domain generalization through novel feature augmentation strategies

Innovation

• Partially supervised self-training framework for disease screening
• Three-branch architecture with teacher and student networks
• Low-frequency feature augmentation for domain generalization