Enhancing Contrastive Learning for Retinal Imaging via Adjusted Augmentation Scales
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866916552108933120 |
|---|---|
| author | Cheng, Zijie Li, Boxuan Altmann, André Keane, Pearse A Zhou, Yukun |
| author_facet | Cheng, Zijie Li, Boxuan Altmann, André Keane, Pearse A Zhou, Yukun |
| contents | Contrastive learning, a prominent approach within self-supervised learning, has demonstrated significant effectiveness in developing generalizable models for various applications involving natural images. However, recent research indicates that these successes do not necessarily extend to the medical imaging domain. In this paper, we investigate the reasons for this suboptimal performance and hypothesize that the dense distribution of medical images poses challenges to the pretext tasks in contrastive learning, particularly in constructing positive and negative pairs. We explore model performance under different augmentation strategies and compare the results to those achieved with strong augmentations. Our study includes six publicly available datasets covering multiple clinically relevant tasks. We further assess the model's generalizability through external evaluations. The model pre-trained with weak augmentation outperforms those with strong augmentation, improving AUROC from 0.838 to 0.848 and AUPR from 0.523 to 0.597 on MESSIDOR2, and showing similar enhancements across other datasets. Our findings suggest that optimizing the scale of augmentation is critical for enhancing the efficacy of contrastive learning in medical imaging. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_02451 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Enhancing Contrastive Learning for Retinal Imaging via Adjusted Augmentation Scales Cheng, Zijie Li, Boxuan Altmann, André Keane, Pearse A Zhou, Yukun Computer Vision and Pattern Recognition Artificial Intelligence Contrastive learning, a prominent approach within self-supervised learning, has demonstrated significant effectiveness in developing generalizable models for various applications involving natural images. However, recent research indicates that these successes do not necessarily extend to the medical imaging domain. In this paper, we investigate the reasons for this suboptimal performance and hypothesize that the dense distribution of medical images poses challenges to the pretext tasks in contrastive learning, particularly in constructing positive and negative pairs. We explore model performance under different augmentation strategies and compare the results to those achieved with strong augmentations. Our study includes six publicly available datasets covering multiple clinically relevant tasks. We further assess the model's generalizability through external evaluations. The model pre-trained with weak augmentation outperforms those with strong augmentation, improving AUROC from 0.838 to 0.848 and AUPR from 0.523 to 0.597 on MESSIDOR2, and showing similar enhancements across other datasets. Our findings suggest that optimizing the scale of augmentation is critical for enhancing the efficacy of contrastive learning in medical imaging. |
| title | Enhancing Contrastive Learning for Retinal Imaging via Adjusted Augmentation Scales |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2501.02451 |