MTCSNN: Multi-task Clinical Siamese Neural Network for Diabetic Retinopathy Severity Prediction

TL;DR

This paper introduces MTCSNN, a multi-task Siamese neural network that leverages ordinal label information and a regression task to improve diabetic retinopathy severity prediction, outperforming existing models.

Contribution

The novel MTCSNN model incorporates ordinal label information and a regression task, enhancing feature learning for fine-grained DR severity classification.

Findings

MTCSNN outperforms ResNet models in AUC and accuracy.

Utilizing ordinal information improves model discrimination.

Regression task aids in learning more discriminative features.

Abstract

Diabetic Retinopathy (DR) has become one of the leading causes of vision impairment in working-aged people and is a severe problem worldwide. However, most of the works ignored the ordinal information of labels. In this project, we propose a novel design MTCSNN, a Multi-task Clinical Siamese Neural Network for Diabetic Retinopathy severity prediction task. The novelty of this project is to utilize the ordinal information among labels and add a new regression task, which can help the model learn more discriminative feature embedding for fine-grained classification tasks. We perform comprehensive experiments over the RetinaMNIST, comparing MTCSNN with other models like ResNet-18, 34, 50. Our results indicate that MTCSNN outperforms the benchmark models in terms of AUC and accuracy on the test dataset.