Back

Capability of GPT-4V(ision) in Japanese National Medical Licensing Examination

Nakao, T.; Miki, S.; Nakamura, Y.; Kikuchi, T.; Nomura, Y.; Hanaoka, S.; Yoshikawa, T.; Abe, O.

2023-11-08 medical education
10.1101/2023.11.07.23298133 medRxiv
Show abstract

BackgroundPrevious research applying large language models (LLMs) to medicine was focused on text-based information. Recently, multimodal variants of LLMs acquired the capability of recognizing images. ObjectiveTo evaluate the capability of GPT-4V, a recent multimodal LLM developed by OpenAI, in recognizing images in the medical field by testing its capability to answer questions in the 117th Japanese National Medical Licensing Examination. MethodsWe focused on 108 questions that had one or more images as part of a question and presented GPT-4V with the same questions under two conditions: 1) with both the question text and associated image(s), and 2) with the question text only. We then compared the difference in accuracy between the two conditions using the exact McNemars test. ResultsAmong the 108 questions with images, GPT-4Vs accuracy was 68% when presented with images and 72% when presented without images (P = .36). ConclusionsThe additional information from the images did not significantly improve the performance of GPT-4V in the Japanese Medical Licensing Examination.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.