Improving Robustness using Joint Attention Network For Detecting Retinal Degeneration From Optical Coherence Tomography Images
This work addresses the problem of limited clinical application for automated retinal disease detection systems due to noise and data transferability issues, representing an incremental improvement in domain-specific methods.
The paper tackled the challenge of accurately detecting retinal diseases from optical coherence tomography images despite noisy data and similar ocular appearances across pathologies, proposing a joint attention network that significantly improved accuracy and robustness on unseen datasets.
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.