BiRA-Net: Bilinear Attention Net for Diabetic Retinopathy Grading
This addresses automated diagnosis of diabetic retinopathy, a retinal disease causing blindness, but appears incremental as it builds on existing attention and bilinear methods for medical imaging.
The paper tackles diabetic retinopathy grading by proposing BiRA-Net, a deep learning architecture that combines attention and bilinear models with a new grading loss function, achieving superior performance in experiments.
Diabetic retinopathy (DR) is a common retinal disease that leads to blindness. For diagnosis purposes, DR image grading aims to provide automatic DR grade classification, which is not addressed in conventional research methods of binary DR image classification. Small objects in the eye images, like lesions and microaneurysms, are essential to DR grading in medical imaging, but they could easily be influenced by other objects. To address these challenges, we propose a new deep learning architecture, called BiRA-Net, which combines the attention model for feature extraction and bilinear model for fine-grained classification. Furthermore, in considering the distance between different grades of different DR categories, we propose a new loss function, called grading loss, which leads to improved training convergence of the proposed approach. Experimental results are provided to demonstrate the superior performance of the proposed approach.