Learning Discriminative Representations for Fine-Grained Diabetic Retinopathy Grading
This work addresses the challenge of early diagnosis for diabetic retinopathy, a leading cause of blindness, by improving automated grading accuracy, though it is incremental as it builds on existing deep learning methods.
The paper tackled the problem of fine-grained diabetic retinopathy grading by developing a bilinear model with ordinal regression and metric loss to identify discriminative areas in fundus images, achieving superior performance on IDRiD and DeepDR datasets.
Diabetic retinopathy (DR) is one of the leading causes of blindness. However, no specific symptoms of early DR lead to a delayed diagnosis, which results in disease progression in patients. To determine the disease severity levels, ophthalmologists need to focus on the discriminative parts of the fundus images. In recent years, deep learning has achieved great success in medical image analysis. However, most works directly employ algorithms based on convolutional neural networks (CNNs), which ignore the fact that the difference among classes is subtle and gradual. Hence, we consider automatic image grading of DR as a fine-grained classification task, and construct a bilinear model to identify the pathologically discriminative areas. In order to leverage the ordinal information among classes, we use an ordinal regression method to obtain the soft labels. In addition, other than only using a categorical loss to train our network, we also introduce the metric loss to learn a more discriminative feature space. Experimental results demonstrate the superior performance of the proposed method on two public IDRiD and DeepDR datasets.