ViT-2SPN: Vision Transformer-based Dual-Stream Self-Supervised Pretraining Networks for Retinal OCT Classification

TL;DR

This paper introduces ViT-2SPN, a self-supervised pretraining framework using Vision Transformers for improved retinal OCT classification, addressing data scarcity and privacy issues in medical imaging.

Contribution

The paper presents a novel dual-stream self-supervised pretraining approach with a Vision Transformer backbone, enhancing feature extraction for OCT diagnosis.

Findings

Achieved a mean AUC of 0.93 on OCT classification

Outperformed existing self-supervised methods in accuracy and F1 score

Demonstrated effectiveness with limited labeled data

Abstract

Optical Coherence Tomography (OCT) is a non-invasive imaging modality essential for diagnosing various eye diseases. Despite its clinical significance, developing OCT-based diagnostic tools faces challenges, such as limited public datasets, sparse annotations, and privacy concerns. Although deep learning has made progress in automating OCT analysis, these challenges remain unresolved. To address these limitations, we introduce the Vision Transformer-based Dual-Stream Self-Supervised Pretraining Network (ViT-2SPN), a novel framework designed to enhance feature extraction and improve diagnostic accuracy. ViT-2SPN employs a three-stage workflow: Supervised Pretraining, Self-Supervised Pretraining (SSP), and Supervised Fine-Tuning. The pretraining phase leverages the OCTMNIST dataset (97,477 unlabeled images across four disease classes) with data augmentation to create dual-augmented views. A Vision Transformer (ViT-Base) backbone extracts features, while a negative cosine similarity loss aligns feature representations. Pretraining is conducted over 50 epochs with a learning rate of 0.0001 and momentum of 0.999. Fine-tuning is performed on a stratified 5.129% subset of OCTMNIST using 10-fold cross-validation. ViT-2SPN achieves a mean AUC of 0.93, accuracy of 0.77, precision of 0.81, recall of 0.75, and an F1 score of 0.76, outperforming existing SSP-based methods.