New multimodal large language models (LLMs) like OpenAI o3 and Gemini 2.5 Pro demonstrated significant advancements in answering Japanese radiology board exam questions, particularly with image input.
Key Details
- 1Eight LLMs were tested on the Japan Diagnostic Radiology Board Examination (JDRBE).
- 2OpenAI o3 achieved 67% accuracy (text-only) and 72% with image input.
- 3Gemini 2.5 Pro also showed notable accuracy improvements with image data.
- 4Both OpenAI o3 and Gemini 2.5 Pro received higher legitimacy scores from radiologist raters than some competitors.
- 5The test set included 233 questions and 477 images (184 CT, 159 MRI, 15 x-ray, 90 nuclear medicine).
- 6Image input statistically improved diagnostic accuracy for several models.
Why It Matters
This study marks the first demonstration of statistically significant improvement in LLM diagnostic accuracy with image input on a radiology board exam, signaling meaningful progress for AI-assisted radiological training and assessment.
Source
AuntMinnie
Related News
•Radiology Business
Lucida Medical Raises $11M for AI-Based Prostate MRI Diagnosis Expansion
Lucida Medical, specializing in AI-assisted prostate cancer diagnosis via MRI, raises $11.4M to drive US FDA approval and platform expansion.
•Radiology Business
NYC Health + Hospitals CEO Considers AI to Replace Radiologists
NYC Health + Hospitals CEO suggests AI could partially replace radiologists, pending regulatory approval.
•Radiology Business
UCLA Appoints Inaugural Associate Dean for Health AI Strategy
UCLA has appointed Katherine P. Andriole as its first associate dean for Health AI Strategy and Innovation, with an initial focus on radiology.