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

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Bioresponsive microneedle stent provides anastomosis and postoperative adjuvant therapy in preclinical resectable intestinal diseases.

Science translational medicine·2026
Same author

Astragalus polysaccharides alleviate oxidative damage by activating the Keap1-Nrf2 antioxidant pathway through miR-183-5p in a fish cell model.

Fish & shellfish immunology·2026
Same author

Mechanistic insights into aroma aging of ripened Pu-erh tea during long-term storage.

Food chemistry·2026
Same author

Influence of Scenting Time on the Volatile Compounds and Sensory Attributes of Jasmine Yellow Tea.

Foods (Basel, Switzerland)·2026
Same author

MPAExpo-LM: Fine-Tuned Large Language Model for Mycophenolic Acid Exposure Estimation After Renal Transplantation.

IEEE journal of biomedical and health informatics·2026
Same author

Decoding long-term aging of ripened Pu-erh tea: sensory equilibrium, metabolic pathways, and chemical markers.

Food chemistry: X·2026

Related Experiment Video

Updated: May 29, 2025

Author Spotlight: Investigating Immune Cell Dynamics in the Tumor Microenvironment — Challenges and Innovations in Cancer Prognosis
07:32

Author Spotlight: Investigating Immune Cell Dynamics in the Tumor Microenvironment — Challenges and Innovations in Cancer Prognosis

Published on: April 12, 2024

1.2K

Evaluating ChatGPT's diagnostic potential for pathology images.

Liya Ding1, Lei Fan1,2, Miao Shen1,3

  • 1Department of Pathology, Sir Run Run Shaw Hospital, Zhejiang University School of Medicine, Hangzhou, China.

Frontiers in Medicine
|February 7, 2025
PubMed
Summary

The large language model GPT-4 shows promise in diagnosing pathological images, achieving accuracy comparable to pathology residents in cancer detection. This AI tool could support pathologists in daily diagnostic tasks.

Keywords:
ChatGPTcolon polypdiagnosislarge language modelpathology images

More Related Videos

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
04:09

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Published on: October 10, 2018

8.2K
Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
14:08

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Published on: April 13, 2013

42.5K

Related Experiment Videos

Last Updated: May 29, 2025

Author Spotlight: Investigating Immune Cell Dynamics in the Tumor Microenvironment — Challenges and Innovations in Cancer Prognosis
07:32

Author Spotlight: Investigating Immune Cell Dynamics in the Tumor Microenvironment — Challenges and Innovations in Cancer Prognosis

Published on: April 12, 2024

1.2K
Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
04:09

Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma

Published on: October 10, 2018

8.2K
Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
14:08

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Published on: April 13, 2013

42.5K

Area of Science:

  • Artificial Intelligence in Medicine
  • Digital Pathology
  • Medical Diagnostics

Background:

  • Large language models (LLMs) like GPT-4 offer potential in healthcare for tasks including clinical note drafting and report generation.
  • Ensuring accuracy in medical applications is a critical challenge for LLMs.
  • This study specifically evaluates GPT-4's accuracy in diagnosing from pathological images.

Purpose of the Study:

  • To investigate the diagnostic accuracy of GPT-4 using both text and image inputs.
  • To assess GPT-4's performance in identifying tumor imaging and tissue origins.
  • To compare GPT-4's diagnostic capabilities with those of human pathologists.

Main Methods:

  • Analysis of 44 histopathological images from 16 organs and 100 colorectal biopsy photomicrographs.
  • GPT-4 (standard and re-evaluation) and four independent pathologists assessed images.
  • Diagnostic accuracy was measured against a reference standard, with evaluation of scanned and photographed images.

Main Results:

  • GPT-4 achieved an overall accuracy of 0.64 for tumor imaging and tissue origin identification.
  • Colon polyp classification accuracy ranged from 0.57 to 0.75; dysplasia grading accuracy reached 0.88.
  • The model demonstrated high sensitivity for adenocarcinoma detection, with slight to moderate consistency between evaluations (Kappa 0.204–0.375).

Conclusions:

  • GPT-4 can diagnose pathological images with accuracy comparable to pathology residents.
  • The model shows potential as a supportive tool in pathology, aiding routine diagnostic workflows.
  • GPT-4's performance indicates advancements in AI-driven pathology diagnostics.