RET-CLIP: A Retinal Image Foundation Model Pre-trained with Clinical Diagnostic Reports
This work addresses the problem of limited medical data for foundation models in ophthalmology, offering a domain-specific tool for retinal disease diagnosis.
The authors tackled the lack of labeled data for vision-language foundation models in ophthalmology by developing RET-CLIP, a CLIP-style model pre-trained on 193,865 patients' retinal images and clinical reports, which outperformed existing benchmarks across eight datasets in four diagnostic categories.
The Vision-Language Foundation model is increasingly investigated in the fields of computer vision and natural language processing, yet its exploration in ophthalmology and broader medical applications remains limited. The challenge is the lack of labeled data for the training of foundation model. To handle this issue, a CLIP-style retinal image foundation model is developed in this paper. Our foundation model, RET-CLIP, is specifically trained on a dataset of 193,865 patients to extract general features of color fundus photographs (CFPs), employing a tripartite optimization strategy to focus on left eye, right eye, and patient level to reflect real-world clinical scenarios. Extensive experiments demonstrate that RET-CLIP outperforms existing benchmarks across eight diverse datasets spanning four critical diagnostic categories: diabetic retinopathy, glaucoma, multiple disease diagnosis, and multi-label classification of multiple diseases, which demonstrate the performance and generality of our foundation model. The sourse code and pre-trained model are available at https://github.com/sStonemason/RET-CLIP.