Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Positron Emission Tomography01:29

Positron Emission Tomography

4.2K
Positron emission tomography (PET) is a medical imaging technique involving radiopharmaceuticals — substances that emit short-lived radiation. Although the first PET scanner was introduced in 1961, it took 15 more years before radiopharmaceuticals were combined with the technique and revolutionized its potential.
One of the main requirements of a PET scan is a positron-emitting radioisotope, which is produced in a cyclotron and then attached to a substance used by the part of the body...
4.2K
Data Collection III01:05

Data Collection III

2.8K
The physical assessment examines the patient for objective data that defines the patient's condition, and aids in formulating the nursing care plan. The purpose of physical assessment is a health status appraisal, which includes identifying health problems, and establishing a database for nursing intervention.
The principles to begin the physical assessment include conducting a comprehensive or problem-related history in a quiet, well-lit room, emphasizing privacy and comfort for the...
2.8K
  1. Home
  3. Capability Of Gpt-4v(ision) In The Japanese National Medical Licensing Examination: Evaluation Study.
  1. Home
  3. Capability Of Gpt-4v(ision) In The Japanese National Medical Licensing Examination: Evaluation Study.

Related Experiment Video

The 4 Mountains Test: A Short Test of Spatial Memory with High Sensitivity for the Diagnosis of Pre-dementia Alzheimer's Disease
06:23

The 4 Mountains Test: A Short Test of Spatial Memory with High Sensitivity for the Diagnosis of Pre-dementia Alzheimer's Disease

Published on: October 13, 2016

32.3K

Capability of GPT-4V(ision) in the Japanese National Medical Licensing Examination: Evaluation Study.

Takahiro Nakao1, Soichiro Miki1, Yuta Nakamura1

  • 1Department of Computational Diagnostic Radiology and Preventive Medicine, The University of Tokyo Hospital, Bunkyo-ku, Tokyo, Japan.

JMIR Medical Education
|March 12, 2024

View abstract on PubMed

Summary
This summary is machine-generated.

Generative pretrained transformer (GPT)-4V, a multimodal large language model (LLM), did not show improved accuracy on the Japanese National Medical Licensing Examination when provided with medical images. Visual information did not significantly enhance GPT-4V's performance on this medical test.

Keywords:
AIChatGPTGPT-4GPT-4VLLMNLPansweranswersartificial intelligencechatbotchatbotsconversational agentconversational agentsexamexaminationexaminationsexamsgenerative pretrained transformerimageimagesimaginglanguage modellanguage modelslarge language modelmedical educationnatural language processingresponseresponses

More Related Videos

Pupillary Response as Assessment of Effective Seizure Induction by Electroconvulsive Therapy
04:51

Pupillary Response as Assessment of Effective Seizure Induction by Electroconvulsive Therapy

Published on: April 11, 2019

9.4K
3D Scanning Technology Bridging Microcircuits and Macroscale Brain Images in 3D Novel Embedding Overlapping Protocol
10:14

3D Scanning Technology Bridging Microcircuits and Macroscale Brain Images in 3D Novel Embedding Overlapping Protocol

Published on: May 12, 2019

7.3K

Related Experiment Videos

The 4 Mountains Test: A Short Test of Spatial Memory with High Sensitivity for the Diagnosis of Pre-dementia Alzheimer's Disease
06:23

The 4 Mountains Test: A Short Test of Spatial Memory with High Sensitivity for the Diagnosis of Pre-dementia Alzheimer's Disease

Published on: October 13, 2016

32.3K
Pupillary Response as Assessment of Effective Seizure Induction by Electroconvulsive Therapy
04:51

Pupillary Response as Assessment of Effective Seizure Induction by Electroconvulsive Therapy

Published on: April 11, 2019

9.4K
3D Scanning Technology Bridging Microcircuits and Macroscale Brain Images in 3D Novel Embedding Overlapping Protocol
10:14

3D Scanning Technology Bridging Microcircuits and Macroscale Brain Images in 3D Novel Embedding Overlapping Protocol

Published on: May 12, 2019

7.3K

Area of Science:

  • Artificial Intelligence in Medicine
  • Medical Education Technology
  • Large Language Models

Background:

  • Previous medical applications of large language models (LLMs) primarily utilized text-based data.
  • Recent advancements have introduced multimodal LLMs capable of image recognition.

Purpose of the Study:

  • To assess the medical image recognition capabilities of generative pretrained transformer (GPT)-4V, a multimodal LLM.
  • To evaluate the impact of visual information on GPT-4V's performance in a medical examination context.

Main Methods:

  • The study utilized 108 questions from the 117th Japanese National Medical Licensing Examination that included images.
  • GPT-4V was tested under two conditions: with text and images, and with text only.
  • Accuracy differences between the two conditions were analyzed using the exact McNemar test.

Main Results:

  • GPT-4V achieved 68% accuracy with images versus 72% accuracy without images (P=.36).
  • In clinical questions, accuracy was 71% with images and 78% without (P=.21).
  • In general questions, accuracy was 30% with images and 20% without (P≥.99).

Conclusions:

  • Incorporating visual information did not significantly improve GPT-4V's performance on the Japanese National Medical Licensing Examination.
  • The study suggests that current multimodal LLMs may not benefit from image data in certain medical assessment scenarios.