AI Science Tracker
EN

Department of Ophthalmology, Faculdade de Medicina, Universidade de Lisboa, 1649-004 Lisbon, Portugal

1 discovery from 2026

Discoveries

Diagnostic Agreement Between a General-Purpose AI Model and Retinal Specialists in Color Fundus Photography-A Pilot Study.

Source check: Source checked. Not independent scientific confirmation.

Finding: A single-prompt Gemini 2.5 Flash image interpretation was compared with retinal specialists on 66 fundus photographs from 35 patients; agreement with the context-informed reference specialist was low (Gwet AC2 −0.58). The study notes the asymmetry of clinical context and is a small exploratory assessment.

Technology: Google Gemini 2.5 Flash assessed 66 CFPs from 35 patients using one open-ended prompt, image-only; masked specialist also image-only; reference specialist had clinical context and multimodal data. Agreement evaluated with Gwet AC2 and weigh

Family Gemini: profile · models page

Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
Google Gemini 2.5 Flash assessed 66 CFPs from 35 patients using one open-ended prompt, image-only; masked specialist also image-only; reference specialist had clinical context and multimodal data. Agreement evaluated with Gwet AC2 and weighted agreement; AI/reference AC2 −0.58.
Test and result
A single-prompt Gemini 2.5 Flash image interpretation was compared with retinal specialists on 66 fundus photographs from 35 patients; agreement with the context-informed reference specialist was low (Gwet AC2 −0.58). The study notes the asymmetry of clinical context and is a small exploratory assessment. Evaluation: A single-prompt Gemini 2.5 Flash image interpretation was compared with retinal specialists on 66 fundus photographs from 35 patients; agreement with the context-informed reference specialist was low (Gwet AC2 −0.58).
Limitation / still unknown
This is a small exploratory comparison (66 images from 35 patients) with unequal clinical context for the reference specialist; it does not establish diagnostic utility.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://mdpi-res.com/d_attachment/jcm/jcm-15-03430/article_deploy/jcm-15-03430.pdf

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review date: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Open the interactive profile for the timeline, or browse all organizations and model families.