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

Opportunistic Promptable Segmentation: Leveraging Routine Radiological Annotations to Guide 3D CT Lesion Segmentation.

Journal of imaging informatics in medicine·2026
Same author

Toward protocol simplification: Deep learning-based image synthesis in three-phase CT urography.

Computers in biology and medicine·2026
Same author

CT-based Opportunistic Screening for Adding Clinical Value: How I Do It.

Radiology·2026
Same author

Opportunistic Screening with Imaging: Actionable Insights from Unused Data.

Radiologic clinics of North America·2026
Same author

ConTEXTual Net 3D: Vision-Language Modeling in PET/CT for Visual Grounding of Positive Findings.

Journal of imaging informatics in medicine·2026
Same author

Portrayal of Medical Students in Artificial Intelligence- Generated Images.

WMJ : official publication of the State Medical Society of Wisconsin·2026

Related Experiment Video

Updated: May 23, 2025

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

Artificial Intelligence for Teaching Case Curation: Evaluating Model Performance on Imaging Report Discrepancies.

Michael Bartley1, Zachary Huemann1, Junjie Hu2

  • 1Department of Radiology, UW-Madison School of Medicine & Public Health, Madison, Wisconsin (M.B., Z.H., X.T., A.B.R., T.K., J.D.W., T.B., E.M.L.).

Academic Radiology
|March 9, 2025
PubMed
Summary
This summary is machine-generated.

Large language models (LLMs) can effectively identify radiology report discrepancies, improving educational case selection and trainee oversight. This technology aids in detecting errors for better resident training.

More Related Videos

Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization
05:49

Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization

Published on: February 23, 2024

754
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
12:18

A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment

Published on: January 11, 2020

7.4K

Related Experiment Videos

Last Updated: May 23, 2025

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
Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization
05:49

Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization

Published on: February 23, 2024

754
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
12:18

A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment

Published on: January 11, 2020

7.4K

Area of Science:

  • Radiology
  • Artificial Intelligence
  • Medical Education

Background:

  • Radiology resident education relies on identifying discrepant reports for learning.
  • Manual curation of discrepant cases is time-consuming and may miss subtle errors.

Purpose of the Study:

  • To assess the feasibility of using a large language model (LLM) for detecting discrepancies in radiology reports.
  • To evaluate the potential of LLM-identified discrepancies for curating educational case sets.

Main Methods:

  • A retrospective study analyzed head CT and musculoskeletal radiograph reports (2017-2021).
  • A fine-tuned LLM (RadBERT) and other models (Mistral, Llama2) were trained to detect report discrepancies on a 5-point scale.
  • Performance was evaluated on a hold-out test set, and LLM-curated cases were compared to random sets.

Main Results:

  • The fine-tuned LLM achieved 90.5% accuracy, 95.5% specificity, and 66.3% sensitivity in discrepancy detection.
  • Sensitivity increased with higher discrepancy scores (81% for scores 4/5).
  • LLM-curated sets showed a higher prevalence of all and major discrepancies compared to random sets (p<0.05).

Conclusions:

  • LLMs can accurately detect trainee report discrepancies, including subtle ones.
  • This technology can enhance radiology resident education by improving case curation.
  • LLMs may also serve as a valuable tool for trainee performance oversight.