Improving Robustness using Joint Attention Network For Detecting Retinal Degeneration From Optical Coherence Tomography Images

TL;DR

This paper introduces a joint attention network architecture that enhances the robustness and accuracy of retinal disease detection from OCT images by leveraging disease-specific features and attention mechanisms.

Contribution

The paper proposes a novel joint network architecture combining supervised disease encoding with unsupervised attention map generation for improved robustness.

Findings

Significant accuracy improvement on unseen datasets.

Enhanced robustness against noisy and similar ocular appearances.

Effective utilization of disease-specific spatial information.

Abstract

Noisy data and the similarity in the ocular appearances caused by different ophthalmic pathologies pose significant challenges for an automated expert system to accurately detect retinal diseases. In addition, the lack of knowledge transferability and the need for unreasonably large datasets limit clinical application of current machine learning systems. To increase robustness, a better understanding of how the retinal subspace deformations lead to various levels of disease severity needs to be utilized for prioritizing disease-specific model details. In this paper we propose the use of disease-specific feature representation as a novel architecture comprised of two joint networks -- one for supervised encoding of disease model and the other for producing attention maps in an unsupervised manner to retain disease specific spatial information. Our experimental results on publicly available datasets show the proposed joint-network significantly improves the accuracy and robustness of state-of-the-art retinal disease classification networks on unseen datasets.