A Foundation Language-Image Model of the Retina (FLAIR): Encoding expert knowledge in text supervision | Publicación