PoCo: A Self-Supervised Approach via Polar Transformation Based Progressive Contrastive Learning for Ophthalmic Disease Diagnosis

TL;DR

PoCo introduces a self-supervised learning method using polar transformation and progressive contrastive learning to improve ophthalmic disease diagnosis from fundus images, reducing annotation needs and enhancing accuracy.

Contribution

The paper proposes a novel polar transformation integrated into contrastive learning and a progressive hard negative sampling scheme for better ophthalmic disease diagnosis.

Findings

Achieves state-of-the-art performance on three datasets.

Reduces annotation efforts while maintaining high accuracy.

Demonstrates good generalization ability across datasets.

Abstract

Automatic ophthalmic disease diagnosis on fundus images is important in clinical practice. However, due to complex fundus textures and limited annotated data, developing an effective automatic method for this problem is still challenging. In this paper, we present a self-supervised method via polar transformation based progressive contrastive learning, called PoCo, for ophthalmic disease diagnosis. Specifically, we novelly inject the polar transformation into contrastive learning to 1) promote contrastive learning pre-training to be faster and more stable and 2) naturally capture task-free and rotation-related textures, which provides insights into disease recognition on fundus images. Beneficially, simple normal translation-invariant convolution on transformed images can equivalently replace the complex rotation-invariant and sector convolution on raw images. After that, we develop a progressive contrastive learning method to efficiently utilize large unannotated images and a novel progressive hard negative sampling scheme to gradually reduce the negative sample number for efficient training and performance enhancement. Extensive experiments on three public ophthalmic disease datasets show that our PoCo achieves state-of-the-art performance with good generalization ability, validating that our method can reduce annotation efforts and provide reliable diagnosis. Codes are available at \url{https://github.com/wjh892521292/PoCo}.