Context-Aware Vision Language Foundation Models for Ocular Disease Screening in Retinal Images

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Berger, Lucie, Lamard, Mathieu, Zhang, Philippe, Borderie, Laurent, Guilcher, Alexandre Le, Massin, Pascale, Cochener, Béatrice, Quellec, Gwenolé, Matta, Sarah
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866910883795435520
author Berger, Lucie
Lamard, Mathieu
Zhang, Philippe
Borderie, Laurent
Guilcher, Alexandre Le
Massin, Pascale
Cochener, Béatrice
Quellec, Gwenolé
Matta, Sarah
author_facet Berger, Lucie
Lamard, Mathieu
Zhang, Philippe
Borderie, Laurent
Guilcher, Alexandre Le
Massin, Pascale
Cochener, Béatrice
Quellec, Gwenolé
Matta, Sarah
contents Foundation models are large-scale versatile systems trained on vast quantities of diverse data to learn generalizable representations. Their adaptability with minimal fine-tuning makes them particularly promising for medical imaging, where data variability and domain shifts are major challenges. Currently, two types of foundation models dominate the literature: self-supervised models and more recent vision-language models. In this study, we advance the application of vision-language foundation (VLF) models for ocular disease screening using the OPHDIAT dataset, which includes nearly 700,000 fundus photographs from a French diabetic retinopathy (DR) screening network. This dataset provides extensive clinical data (patient-specific information such as diabetic health conditions, and treatments), labeled diagnostics, ophthalmologists text-based findings, and multiple retinal images for each examination. Building on the FLAIR model $\unicode{x2013}$ a VLF model for retinal pathology classification $\unicode{x2013}$ we propose novel context-aware VLF models (e.g jointly analyzing multiple images from the same visit or taking advantage of past diagnoses and contextual data) to fully leverage the richness of the OPHDIAT dataset and enhance robustness to domain shifts. Our approaches were evaluated on both in-domain (a testing subset of OPHDIAT) and out-of-domain data (public datasets) to assess their generalization performance. Our model demonstrated improved in-domain performance for DR grading, achieving an area under the curve (AUC) ranging from 0.851 to 0.9999, and generalized well to ocular disease detection on out-of-domain data (AUC: 0.631-0.913).
format Preprint
id arxiv_https___arxiv_org_abs_2503_15212
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Context-Aware Vision Language Foundation Models for Ocular Disease Screening in Retinal Images
Berger, Lucie
Lamard, Mathieu
Zhang, Philippe
Borderie, Laurent
Guilcher, Alexandre Le
Massin, Pascale
Cochener, Béatrice
Quellec, Gwenolé
Matta, Sarah
Image and Video Processing
Foundation models are large-scale versatile systems trained on vast quantities of diverse data to learn generalizable representations. Their adaptability with minimal fine-tuning makes them particularly promising for medical imaging, where data variability and domain shifts are major challenges. Currently, two types of foundation models dominate the literature: self-supervised models and more recent vision-language models. In this study, we advance the application of vision-language foundation (VLF) models for ocular disease screening using the OPHDIAT dataset, which includes nearly 700,000 fundus photographs from a French diabetic retinopathy (DR) screening network. This dataset provides extensive clinical data (patient-specific information such as diabetic health conditions, and treatments), labeled diagnostics, ophthalmologists text-based findings, and multiple retinal images for each examination. Building on the FLAIR model $\unicode{x2013}$ a VLF model for retinal pathology classification $\unicode{x2013}$ we propose novel context-aware VLF models (e.g jointly analyzing multiple images from the same visit or taking advantage of past diagnoses and contextual data) to fully leverage the richness of the OPHDIAT dataset and enhance robustness to domain shifts. Our approaches were evaluated on both in-domain (a testing subset of OPHDIAT) and out-of-domain data (public datasets) to assess their generalization performance. Our model demonstrated improved in-domain performance for DR grading, achieving an area under the curve (AUC) ranging from 0.851 to 0.9999, and generalized well to ocular disease detection on out-of-domain data (AUC: 0.631-0.913).
title Context-Aware Vision Language Foundation Models for Ocular Disease Screening in Retinal Images
topic Image and Video Processing
url https://arxiv.org/abs/2503.15212