AdvBlur: Adversarial Blur for Robust Diabetic Retinopathy Classification and Cross-Domain Generalization

TL;DR

AdvBlur introduces adversarially blurred images and a dual-loss framework to enhance the robustness and cross-domain generalization of diabetic retinopathy classification models across diverse datasets and imaging conditions.

Contribution

The paper proposes AdvBlur, a novel method integrating adversarial blur and dual-loss training to improve DR classification robustness and domain generalization.

Findings

Achieves competitive performance on unseen datasets.

Effectively mitigates distributional variations from different devices and conditions.

Demonstrates robustness against low-quality images and dataset size variations.

Abstract

Diabetic retinopathy (DR) is a leading cause of vision loss worldwide, yet early and accurate detection can significantly improve treatment outcomes. While numerous Deep learning (DL) models have been developed to predict DR from fundus images, many face challenges in maintaining robustness due to distributional variations caused by differences in acquisition devices, demographic disparities, and imaging conditions. This paper addresses this critical limitation by proposing a novel DR classification approach, a method called AdvBlur. Our method integrates adversarial blurred images into the dataset and employs a dual-loss function framework to address domain generalization. This approach effectively mitigates the impact of unseen distributional variations, as evidenced by comprehensive evaluations across multiple datasets. Additionally, we conduct extensive experiments to explore the effects of factors such as camera type, low-quality images, and dataset size. Furthermore, we perform ablation studies on blurred images and the loss function to ensure the validity of our choices. The experimental results demonstrate the effectiveness of our proposed method, achieving competitive performance compared to state-of-the-art domain generalization DR models on unseen external datasets.