SSVT: Self-Supervised Vision Transformer For Eye Disease Diagnosis Based On Fundus Images

TL;DR

The paper introduces SSVT, a self-supervised vision transformer that accurately diagnoses four eye diseases from unlabeled fundus images, reducing reliance on labeled data and aiding resource-limited regions.

Contribution

It presents a novel label-free, self-supervised vision transformer approach for eye disease diagnosis from fundus images, achieving high accuracy without requiring annotated datasets.

Findings

Achieved 97.0% accuracy on four eye diseases

Effective on six public and two hospital datasets

Reduces need for manual labeling in medical imaging

Abstract

Machine learning-based fundus image diagnosis technologies trigger worldwide interest owing to their benefits such as reducing medical resource power and providing objective evaluation results. However, current methods are commonly based on supervised methods, bringing in a heavy workload to biomedical staff and hence suffering in expanding effective databases. To address this issue, in this article, we established a label-free method, name 'SSVT',which can automatically analyze un-labeled fundus images and generate high evaluation accuracy of 97.0% of four main eye diseases based on six public datasets and two datasets collected by Beijing Tongren Hospital. The promising results showcased the effectiveness of the proposed unsupervised learning method, and the strong application potential in biomedical resource shortage regions to improve global eye health.