EFIQA: Explainable Fundus Image Quality Assessment via Anatomical Priors

2026-06-18

AI Summary

This work addresses the limitations of conventional fundus image quality assessment methods—namely their reliance on annotated labels, poor generalizability, and lack of spatial interpretability—by proposing a novel two-stage unsupervised framework that requires no quality annotations. The approach introduces anatomical priors to learn the expected structural layout of a normal fundus image and leverages masked anatomical structure reconstruction for unsupervised anomaly detection. By integrating a frozen foundation model with lightweight adapters, the framework performs knowledge distillation to produce pixel-level spatial quality maps. Evaluated across multiple external datasets, the method not only surpasses existing supervised approaches but also demonstrates superior cross-domain generalization and enhanced interpretability under diverse quality criteria.

Abstract

Image quality control is vital for a wide range of downstream applications. Deep learning-based image quality assessment methods typically train classifiers on dataset-specific quality labels, inheriting two limitations: (1) generalization is tied to the labeling criteria of the training set and (2) these methods cannot provide spatial feedback on where the quality is degraded, lacking explainability. In this work, we propose EFIQA, a framework that requires no quality-related supervision and produces spatial quality maps by design. Rather than learning ``what is degradation" from human-annotated labels, EFIQA learns ``what should be there" by leveraging anatomical priors. For fundus photography, we instantiate this as a two-stage approach, by first training an unsupervised anomaly detector via masked anatomical inpainting to identify regions of missing vasculature, and then distilling this prior knowledge into a shallow adapter mapping features of a frozen foundation model to precise quality maps. External-dataset evaluation demonstrates that this label-free approach with minimal adaptation achieves better performance and explainability compared with supervised methods across benchmarks with different quality criteria, highlighting its potential for real-world applications.

Problem

• image quality assessment
• explainability
• anatomical priors
• fundus image
• generalization

Innovation

• Explainable AI
• Anatomical Priors
• Unsupervised Anomaly Detection
• Foundation Model Adaptation
• Fundus Image Quality Assessment