Enhancing Contrastive Learning for Retinal Imaging via Adjusted Augmentation Scales
Keywords: contrastive learning, augmentation scales, data distribution, retinal imaging
Abstract: Contrastive learning, a typical self-supervised learning strategy, operates on bringing similar data together while pushing dissimilar data apart in latent space. This approach extracts robust and discriminative representations, thus being widely used in natural computer vision tasks, such as object classification. However, unlike natural images, medical images from the same modality (e.g., color fundus photographs) tend to share substantial similarities in imaging area and anatomical tissues, leading to a denser distribution in latent space. As a result, the default use of strong augmentations in contrastive learning potentially exacerbates this intensive distribution in medical images, making it difficult to distinguish between genuinely similar and dissimilar data, and therefore hindering model pre-training convergence. In this paper, we hypothesise that weaker augmentation better fits contrastive learning for medical imaging and explore model performance under different augmentation strategies. Our study includes six publicly available retinal datasets covering multiple clinically relevant tasks. We assess the models' performance and generalizability via extensive experiments. The model pre-trained with weak augmentation outperforms the one pre-trained with strong augmentation, achieving approximately a 6\% increase in AUPR (P<0.001) and a 12.5\% increase in sensitivity (P<0.001) on MESSIDOR-2. Similar improvements are observed across other datasets. Our findings suggest that optimizing the scale of augmentation is critical for enhancing the efficacy of contrastive learning in medical imaging. The model weights and relevant code are available at: https://github.com/ziijiecheng/MIDL-2025
Primary Subject Area: Application: Ophthalmology
Secondary Subject Area: Foundation Models
Paper Type: Both
Registration Requirement: Yes
Reproducibility: https://github.com/ziijiecheng/MIDL-2025
Visa & Travel: Yes
Submission Number: 139
Loading