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

Radiological Investigation III: Pulmonary Angiogram and PET Scan01:13

Radiological Investigation III: Pulmonary Angiogram and PET Scan

161
Radiological investigations are paramount in the diagnosis and management of various pulmonary diseases. Two essential investigations are the Pulmonary Angiogram and the Positron Emission Tomography (PET) Scan.
Pulmonary Angiogram
A Pulmonary Angiogram is an invasive procedure involving injecting a contrast medium through a catheter threaded into the pulmonary artery or the right side of the heart to visualize the pulmonary vasculature. Computed Tomography (CT) scans have mainly replaced this...
161

You might also read

Related Articles

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

Sort by
Same author

A Whole-Body PSMA-PET/CT dataset with manually annotated tumor lesions.

Scientific data·2026
Same author

AI-Based Longitudinal Scoliosis Monitoring on EOS Whole-Spine Radiographs in a Pediatric to Young Adult Cohort.

Academic radiology·2026
Same author

Enhancing pancreatic imaging in CT - prospective comparison of fixed versus individualized post-trigger delay in bolus tracking.

European journal of radiology·2026
Same author

The effect of medical explanations from large language models on diagnostic accuracy in radiology.

NPJ digital medicine·2026
Same author

Diagnostic performance and reader confidence of MRI-derived CT-like images in patients with femoral head necrosis.

European radiology experimental·2026
Same author

AutoPET Challenge on Fully Automated Lesion Segmentation in Oncologic PET/CT Imaging, Part 2: Domain Generalization.

Journal of nuclear medicine : official publication, Society of Nuclear Medicine·2025

Related Experiment Video

Updated: Sep 3, 2025

Guidelines and Experience Using Imaging Biomarker Explorer IBEX for Radiomics
10:17

Guidelines and Experience Using Imaging Biomarker Explorer IBEX for Radiomics

Published on: January 8, 2018

13.3K

Clinically focused multi-cohort benchmarking as a tool for external validation of artificial intelligence algorithm

Jan Rudolph1, Balthasar Schachtner2,3, Nicola Fink2,3

  • 1Department of Radiology, University Hospital, LMU Munich, Marchioninistr. 15, 81377, Munich, Germany. jan.rudolph@med.uni-muenchen.de.

Scientific Reports
|July 27, 2022
PubMed
Summary

External validation of artificial intelligence (AI) algorithms for chest X-rays is crucial. CheXNet showed varied performance across different patient positions and pathologies, highlighting the need for robust, real-world testing.

More Related Videos

Author Spotlight: Expanding Interventional Pulmonology Research with Robotic-Assisted Bronchoscopy
04:10

Author Spotlight: Expanding Interventional Pulmonology Research with Robotic-Assisted Bronchoscopy

Published on: July 19, 2024

784
Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
13:44

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Published on: August 30, 2013

43.0K

Related Experiment Videos

Last Updated: Sep 3, 2025

Guidelines and Experience Using Imaging Biomarker Explorer IBEX for Radiomics
10:17

Guidelines and Experience Using Imaging Biomarker Explorer IBEX for Radiomics

Published on: January 8, 2018

13.3K
Author Spotlight: Expanding Interventional Pulmonology Research with Robotic-Assisted Bronchoscopy
04:10

Author Spotlight: Expanding Interventional Pulmonology Research with Robotic-Assisted Bronchoscopy

Published on: July 19, 2024

784
Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
13:44

Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns

Published on: August 30, 2013

43.0K

Area of Science:

  • Radiology
  • Artificial Intelligence
  • Medical Imaging

Background:

  • Artificial intelligence (AI) algorithms for evaluating chest radiographs (CXRs) are rapidly advancing.
  • External validation is essential to ensure AI reliability and identify potential training data errors.

Purpose of the Study:

  • To perform a multi-cohort benchmarking of the CheXNet AI algorithm on publicly accessible supine chest radiographs (SCXRs).
  • To evaluate CheXNet's performance across diverse clinical cohorts, reference standards, and patient positioning.
  • To compare AI performance against medical experts with varying qualifications.

Main Methods:

  • Applied a multi-cohort benchmarking approach to the CheXNet AI algorithm using three distinct clinical cohorts.
  • Cohorts varied in patient positioning (CXR vs. SCXR), reference standards (CT-based vs. CXR-based), and reader expertise.
  • Analyzed algorithm performance for common pathologies, pneumothorax detection, and basal lung opacities.

Main Results:

  • CheXNet performance varied significantly based on patient positioning and the specific pathology.
  • The algorithm showed non-significant superiority over radiology residents in detecting lung nodules and readers in detecting basal pneumonia and pleural effusion.
  • Performance was notably poorer for small pneumothoraces and in CXRs lacking thoracic material, likely due to training data confounders.

Conclusions:

  • Multi-cohort external validation is critical for assessing AI algorithm performance in real-world clinical scenarios.
  • Patient positioning and training data characteristics can significantly impact AI diagnostic accuracy.
  • Findings underscore the need for rigorous validation to ensure the safe and effective deployment of AI in radiology.