A Multi-stage Transfer Learning Framework for Diabetic Retinopathy Grading on Small Data

TL;DR

This paper introduces a multi-stage transfer learning framework combined with a class-balanced loss function to improve diabetic retinopathy grading accuracy on small datasets, addressing data scarcity and class imbalance issues.

Contribution

It proposes a novel multi-stage transfer learning approach and a class-balanced loss function specifically designed for small, imbalanced DR datasets.

Findings

Achieved 0.7961 accuracy on IDRiD dataset

Obtained 0.8763 quadratic weighted kappa score

Outperformed several state-of-the-art methods

Abstract

Diabetic retinopathy (DR) is one of the major blindness-causing diseases currently known. Automatic grading of DR using deep learning methods not only speeds up the diagnosis of the disease but also reduces the rate of misdiagnosis. However,problems such as insufficient samples and imbalanced class distribution in small DR datasets have constrained the improvement of grading performance. In this paper, we apply the idea of multi-stage transfer learning into the grading task of DR. The new transfer learning technique utilizes multiple datasets with different scales to enable the model to learn more feature representation information. Meanwhile, to cope with the imbalanced problem of small DR datasets, we present a class-balanced loss function in our work and adopt a simple and easy-to-implement training method for it. The experimental results on IDRiD dataset show that our method can effectively improve the grading performance on small data, obtaining scores of 0.7961 and 0.8763 in terms of accuracy and quadratic weighted kappa, respectively. Our method also outperforms several state-of-the-art methods.