---
license: apache-2.0
tags:
  - multimodal
  - ophthalmology
  - OCT
  - benchmark
  - medical
  - visual question answering
---

# 🤗 OCT-Bench

![Overview](overview.png)

To comprehensively evaluate Multimodal Large Language Models (MLLMs) on optical coherence tomography (OCT) image understanding, we introduce **OCT-Bench**, a comprehensive benchmark that follows the real-world clinical interpretation workflow from **visual perception**, **medical cognition**, to **clinical reasoning**.

OCT-Bench contains **10,076 expert-verified multiple-choice questions** constructed from **4,137 OCT images** collected from **seven public datasets**. The benchmark establishes a hierarchical capability taxonomy consisting of **3 capability dimensions**, **9 capability groups**, and **20 fine-grained tasks**, covering imaging attributes, retinal anatomy, lesion characteristics, disease diagnosis, treatment planning, and prognostic management.

We benchmark **20 representative MLLMs**, including proprietary models, open-source general-purpose models, and medical-domain models, providing a comprehensive evaluation of OCT understanding capabilities.

For detailed usage and instructions, please refer to the [GitHub page](https://github.com/baochenfu/OCT-Bench).

You can download **OCT-Bench**. The expected directory structure is:

OCT-Bench
├── images
│   ├── OCT5K
│   ├── OCTDL
│   └── ...
└── VQA
    ├── T01_VQA.jsonl
    ├── T02_VQA.jsonl
    └── ...