A Foundation Language-Image Model of the Retina (FLAIR): Encoding Expert Knowledge in Text Supervision

TL;DR

FLAIR is a vision-language model for retinal imaging that incorporates expert clinical knowledge through textual prompts, significantly improving generalization and performance in medical imaging tasks, especially under domain shifts and limited data.

Contribution

The paper introduces FLAIR, a novel pre-trained vision-language model that embeds expert knowledge via textual prompts, enhancing medical image understanding beyond existing models.

Findings

FLAIR outperforms dataset-focused models in few-shot scenarios.

Incorporating expert knowledge improves zero-shot generalization.

FLAIR surpasses larger generalist models and self-supervised networks in retinal tasks.

Abstract

Foundation vision-language models are currently transforming computer vision, and are on the rise in medical imaging fueled by their very promising generalization capabilities. However, the initial attempts to transfer this new paradigm to medical imaging have shown less impressive performances than those observed in other domains, due to the significant domain shift and the complex, expert domain knowledge inherent to medical-imaging tasks. Motivated by the need for domain-expert foundation models, we present FLAIR, a pre-trained vision-language model for universal retinal fundus image understanding. To this end, we compiled 38 open-access, mostly categorical fundus imaging datasets from various sources, with up to 101 different target conditions and 288,307 images. We integrate the expert's domain knowledge in the form of descriptive textual prompts, during both pre-training and zero-shot inference, enhancing the less-informative categorical supervision of the data. Such a textual expert's knowledge, which we compiled from the relevant clinical literature and community standards, describes the fine-grained features of the pathologies as well as the hierarchies and dependencies between them. We report comprehensive evaluations, which illustrate the benefit of integrating expert knowledge and the strong generalization capabilities of FLAIR under difficult scenarios with domain shifts or unseen categories. When adapted with a lightweight linear probe, FLAIR outperforms fully-trained, dataset-focused models, more so in the few-shot regimes. Interestingly, FLAIR outperforms by a wide margin larger-scale generalist image-language models and retina domain-specific self-supervised networks, which emphasizes the potential of embedding experts' domain knowledge and the limitations of generalist models in medical imaging.