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 Experiment Video

Updated: Jun 4, 2026

A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules
10:26

A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules

Published on: May 19, 2023

Externally Tested AI Models for Malignancy Classification of Lung Nodules at CT: A Systematic Review and

Oke Dimas Asmara1,2,3,4, Eline G M Steenhuis5, Kim de Jong6

  • 1Department of Pulmonary Medicine, Frisius Medical Center, Henri Dunantweg 2, 8934 AD Leeuwarden, the Netherlands.

Radiology. Artificial Intelligence
|June 3, 2026
PubMed
Summary

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

Clinical urgency of incidental findings in the first year of the 4-IN-THE-LUNG-RUN lung cancer screening program.

European journal of cancer (Oxford, England : 1990)·2026
Same author

Age-Stratified Long-Term Outcomes of Immune Checkpoint Inhibitors for Stage IV Melanoma and NSCLC in The Netherlands: A Population-Based Study.

Cancers·2026
Same author

Immunotherapy-related interstitial lung disease: a call to improve risk stratification and diagnostic consistency.

ERJ open research·2026
Same authorSame journal

Benchmarking of AI and Radiologists for Indeterminate Lung Nodule Malignancy Risk Estimation on Screening CT: The LUNA25 Challenge.

Radiology. Artificial intelligence·2026
Same author

Repeatability of AI-quantified incidental findings on lung cancer screening CT scans in the NELSON trial.

European radiology·2026
Same author

Incidence and lung cancer probability of new nodules in the UK lung screening trial.

European journal of cancer (Oxford, England : 1990)·2026

Externally tested artificial intelligence (AI) models show high sensitivity for classifying lung nodules on chest CT scans but moderate specificity. These AI tools may aid in ruling out malignancy, but heterogeneity and bias require careful consideration.

Area of Science:

  • Radiology and Medical Imaging
  • Artificial Intelligence in Medicine
  • Oncology

Background:

  • Lung nodules are frequently detected on chest CT scans, necessitating accurate classification for malignancy.
  • Artificial intelligence (AI) models are increasingly developed for lung nodule analysis.
  • External validation of AI models is crucial for assessing their real-world diagnostic performance.

Purpose of the Study:

  • To evaluate the pooled diagnostic accuracy of externally tested AI models for lung nodule malignancy classification on chest CT.
  • To synthesize evidence on the sensitivity, specificity, and other performance metrics of AI models in this application.

Main Methods:

  • Systematic literature search of major databases (PubMed, Embase, Web of Science, CINAHL, Cochrane Library) up to January 2025.
Keywords:
Artificial IntelligenceExternal ValidationLung CancerMalignancy ClassificationMeta-Analysis

More Related Videos

Three-Dimensional Reconstruction for the Whole Lung with Early Multiple Pulmonary Nodules
07:53

Three-Dimensional Reconstruction for the Whole Lung with Early Multiple Pulmonary Nodules

Published on: October 13, 2023

CT-guided Preoperative Localization of Pulmonary Nodules Using a Glucose Test and Tissue Adhesive
02:37

CT-guided Preoperative Localization of Pulmonary Nodules Using a Glucose Test and Tissue Adhesive

Published on: January 30, 2026

Related Experiment Videos

Last Updated: Jun 4, 2026

A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules
10:26

A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules

Published on: May 19, 2023

Three-Dimensional Reconstruction for the Whole Lung with Early Multiple Pulmonary Nodules
07:53

Three-Dimensional Reconstruction for the Whole Lung with Early Multiple Pulmonary Nodules

Published on: October 13, 2023

CT-guided Preoperative Localization of Pulmonary Nodules Using a Glucose Test and Tissue Adhesive
02:37

CT-guided Preoperative Localization of Pulmonary Nodules Using a Glucose Test and Tissue Adhesive

Published on: January 30, 2026

  • Inclusion of studies using pathology or 2-year follow-up as reference standards for AI model evaluation.
  • Meta-analysis using bivariate random-effects models to pool diagnostic accuracy metrics; QUADAS-2 for risk of bias assessment.

    • Main Results:

    • Twenty-one studies with 7,454 lung nodules were analyzed.
    • Pooled sensitivity was 88% and specificity was 75%, with an area under the ROC curve of 0.89.
    • High heterogeneity (I² > 90%) was observed; deep learning models, particularly CNN architectures, showed higher specificity.

    Conclusions:

    • Externally validated AI models exhibit high sensitivity but moderate specificity for lung nodule malignancy classification on chest CT.
    • AI models show potential for use in lung nodule "rule-out" strategies.
    • Substantial heterogeneity, inconsistent reporting, and risk of bias necessitate cautious interpretation of current AI model performance.