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

Imaging Studies for Cardiovascular System VI: Calcium -Scoring CT01:25

Imaging Studies for Cardiovascular System VI: Calcium -Scoring CT

119
Calcium-Scoring CT ScanA calcium-scoring CT scan, also known as coronary artery calcium (CAC) scan, detects calcium deposits in the coronary arteries. This test assesses the risk of coronary artery disease (CAD), which can lead to cardiovascular events such as angina, heart failure, and sudden cardiac arrest.A calcium-scoring CT scan is generally recommended for individuals at intermediate risk of CAD without symptoms. It includes:Men aged 40-75 and women aged 50-75: Especially those with a...
119
Imaging Studies for Cardiovascular System V: CT01:28

Imaging Studies for Cardiovascular System V: CT

66
Cardiac computed tomography (CT) scanning is an advanced cardiac imaging technique that utilizes CT technology, with or without intravenous (IV) contrast, to produce accurate cross-sectional virtual slices of specific areas of the heart, coronary circulation, and major blood vessels such as the aorta, pulmonary veins, and arteries. The computer processes these slices to generate three-dimensional images. Multidetector CT (MDCT) is a rapid form of CT scanning that captures multiple slices...
66

You might also read

Related Articles

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

Sort by
Same author

What It Takes to Reduce Radiation Dose for Coronary Artery Imaging Globally.

Korean journal of radiology·2026
Same author

[The Clinical Utility of Mediastinal MRI for Evaluating Anterior Mediastinal Masses].

Journal of the Korean Society of Radiology·2026
Same author

AI Solution for Identifying Actionable Abnormalities on Chest Radiographs: A Retrospective Case-Control Study.

Journal of the Korean Society of Radiology·2026
Same author

Practical Barriers to Real-World Implementation of Artificial Intelligence-Driven CT-Derived Fractional Flow Reserve.

Korean journal of radiology·2026
Same author

Highlights of the 2026 Korean Society of Hypertension guidelines for the management of hypertension: what's new and what has changed.

Clinical hypertension·2026
Same author

Imaging-based predictors of progression in brain abscess: the role of loculation and ADC gradient.

Acta neurochirurgica·2026
Same journal

Comments on "Association of T2-Weighted Imaging Features in Invasive Breast Cancer With Clinicopathologic Features and Neoadjuvant Treatment Outcomes".

Korean journal of radiology·2026
Same journal

Distinguishing Molecular and Histologic Glioblastomas Using Multiparametric MRI-Based Habitat Analysis.

Korean journal of radiology·2026
Same journal

Transarterial Radioembolization Versus Transarterial Chemoembolization in Elderly Patients (≥75 Years) With Hepatocellular Carcinoma: A Propensity Score-Matched Comparison.

Korean journal of radiology·2026
Same journal

Magnetic Resonance Imaging After Total Neoadjuvant Therapy in Rectal Cancer: Treatment-Oriented Response Assessment and Reporting for Watch-and-Wait Management.

Korean journal of radiology·2026
Same journal

How BI-RADS v2025 Helps Patients: New Audits and Method of Detection.

Korean journal of radiology·2026
Same journal

Comments on "Prognostic Significance of Pretreatment ¹⁸F-FDG PET/CT Parameters in Patients With ER+/HER2- Metastatic Breast Cancer Treated With CDK4/6 Inhibitors Plus Endocrine Therapy".

Korean journal of radiology·2026
See all related articles

Related Experiment Video

Updated: Sep 9, 2025

Author Spotlight: Advancing Cardiovascular Imaging - Introducing the Spatially Weighted Calcium Score for Early Disease Detection
06:57

Author Spotlight: Advancing Cardiovascular Imaging - Introducing the Spatially Weighted Calcium Score for Early Disease Detection

Published on: September 22, 2023

1.1K

Large Language Models for CAD-RADS 2.0 Extraction From Semi-Structured Coronary CT Angiography Reports: A

Dabin Min1,2, Kwang Nam Jin3,4, SangHeum Bang3

  • 1Interdisciplinary Program in Bioengineering, Seoul National University Graduate School, Seoul, Republic of Korea.

Korean Journal of Radiology
|August 28, 2025
PubMed
Summary
This summary is machine-generated.

Large language models (LLMs) accurately extract Coronary Artery Disease-Reporting and Data System (CAD-RADS) components from coronary CT angiography (CCTA) reports. Chain-of-thought prompting significantly improved LLM accuracy for CAD-RADS 2.0 data extraction.

Keywords:
CAD-RADS 2.0Coronary CT angiographyInformation extractionLarge language modelPrompting strategy

More Related Videos

Improved Registration of 3D CT Angiography with X-ray Fluoroscopy for Image Fusion During Transcatheter Aortic Valve Implantation
06:59

Improved Registration of 3D CT Angiography with X-ray Fluoroscopy for Image Fusion During Transcatheter Aortic Valve Implantation

Published on: June 3, 2018

10.7K
Development and Evaluation of 3D-Printed Cardiovascular Phantoms for Interventional Planning and Training
09:57

Development and Evaluation of 3D-Printed Cardiovascular Phantoms for Interventional Planning and Training

Published on: January 18, 2021

4.2K

Related Experiment Videos

Last Updated: Sep 9, 2025

Author Spotlight: Advancing Cardiovascular Imaging - Introducing the Spatially Weighted Calcium Score for Early Disease Detection
06:57

Author Spotlight: Advancing Cardiovascular Imaging - Introducing the Spatially Weighted Calcium Score for Early Disease Detection

Published on: September 22, 2023

1.1K
Improved Registration of 3D CT Angiography with X-ray Fluoroscopy for Image Fusion During Transcatheter Aortic Valve Implantation
06:59

Improved Registration of 3D CT Angiography with X-ray Fluoroscopy for Image Fusion During Transcatheter Aortic Valve Implantation

Published on: June 3, 2018

10.7K
Development and Evaluation of 3D-Printed Cardiovascular Phantoms for Interventional Planning and Training
09:57

Development and Evaluation of 3D-Printed Cardiovascular Phantoms for Interventional Planning and Training

Published on: January 18, 2021

4.2K

Area of Science:

  • Medical Informatics
  • Artificial Intelligence in Radiology
  • Cardiovascular Imaging

Background:

  • Coronary CT angiography (CCTA) reports contain critical data for diagnosing Coronary Artery Disease (CAD).
  • The Coronary Artery Disease-Reporting and Data System (CAD-RADS) standardizes reporting for CCTA findings.
  • Automated extraction of CAD-RADS components can streamline workflow and improve data consistency.

Purpose of the Study:

  • To assess the accuracy of various large language models (LLMs) in extracting CAD-RADS 2.0 components from CCTA reports.
  • To evaluate the impact of different prompting strategies, including chain-of-thought (CoT), on LLM performance.
  • To compare the performance of multiple LLMs on a multi-institutional dataset.

Main Methods:

  • A multi-institutional dataset of 319 synthetic, semi-structured CCTA reports was curated.
  • Reference standards for CAD-RADS 2.0 components (stenosis severity, plaque burden, modifiers) were established by board-certified radiologists.
  • Six LLMs were evaluated using zero-shot, few-shot, and CoT prompting strategies.
  • Accuracy was measured using McNemar's test for statistical comparison.

Main Results:

  • LLMs demonstrated high accuracy in extracting all CAD-RADS 2.0 components, with peak stenosis severity accuracies up to 0.980 (internal) and 0.946 (external).
  • Plaque burden extraction achieved near-perfect accuracy (up to 0.993 externally), and modifier detection was consistently high (≥0.990).
  • Chain-of-thought (CoT) prompting significantly improved accuracy for several models, notably GPT-4, by up to 0.192 for stenosis severity.

Conclusions:

  • Large language models show significant potential for accurately automating the extraction of CAD-RADS 2.0 components from CCTA reports.
  • The use of chain-of-thought prompting is a key strategy to enhance LLM performance in this task.
  • These findings support the integration of LLMs into radiological workflows for improved efficiency and data standardization.