Benchmark RadarAI BENCHMARK PROFILE
OCT-Bench
OCT-Bench evaluates multimodal large language models on OCT image understanding with 10,076 multiple-choice questions from 4,137 images across seven public datasets, covering 20 tasks in Perception, Cognition, and Reasoning.
- Released
- 2026-07-18
- Readiness
- Runnable
- Primary field
- Health & Life Sciences
Why it matters
Offers a comprehensive benchmark for OCT understanding beyond coarse classification, enabling capability bottleneck analysis for clinical deployment.
Motivation
Optical coherence tomography (OCT) imaging is essential for the diagnosis and treatment of retinal diseases.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.