Diagnosis of Multiple Fundus Disorders Amidst a Scarcity of Medical Experts Via Self-supervised Machine Learning

TL;DR

This paper introduces a self-supervised machine learning framework that accurately diagnoses multiple fundus diseases from unlabeled images, surpassing existing methods and even human experts, especially benefiting regions with limited medical resources.

Contribution

A novel self-supervised learning approach for fundus disease diagnosis that requires no labeled data and generalizes across diverse datasets and imaging sources.

Findings

Achieves 15.7% higher AUC than supervised methods

Outperforms a single human expert in diagnosis accuracy

Adapts effectively to various datasets and imaging conditions

Abstract

Fundus diseases are major causes of visual impairment and blindness worldwide, especially in underdeveloped regions, where the shortage of ophthalmologists hinders timely diagnosis. AI-assisted fundus image analysis has several advantages, such as high accuracy, reduced workload, and improved accessibility, but it requires a large amount of expert-annotated data to build reliable models. To address this dilemma, we propose a general self-supervised machine learning framework that can handle diverse fundus diseases from unlabeled fundus images. Our method's AUC surpasses existing supervised approaches by 15.7%, and even exceeds performance of a single human expert. Furthermore, our model adapts well to various datasets from different regions, races, and heterogeneous image sources or qualities from multiple cameras or devices. Our method offers a label-free general framework to diagnose fundus diseases, which could potentially benefit telehealth programs for early screening of people at risk of vision loss.