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: Jul 4, 2026

Pioneering Patient-Specific Approaches for Precision Surgery Using Imaging and Virtual Reality
06:18

Pioneering Patient-Specific Approaches for Precision Surgery Using Imaging and Virtual Reality

Published on: April 5, 2024

Pre-Deployment Evaluation of a Remote Service for Short-Axis Cardiac MRI Segmentation.

Sadat Hasan Chowdhury1, Hinrich Winther2, Steffen Oeltze-Jafra1

  • 1Peter L. Reichertz Institute for Medical Informatics, Hannover Medical School (MHH), Germany.

Studies in Health Technology and Informatics
|July 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

Integrating Anomaly Detection and LLM-Based Explanation Generation in Clinical Data Dashboards.

Studies in health technology and informatics·2026
Same author

Visualization of Clinical Case Similarities in Pediatric Cardiology Using Structured Case Conference Reports.

Studies in health technology and informatics·2026
Same author

Artificial intelligence for patent ductus arteriosus-a systematic review.

Frontiers in pediatrics·2026
Same author

Automatic Bone Removal in CBCT Scans of the Body Trunk: Thorax, Abdomen, and Pelvis.

Cardiovascular and interventional radiology·2026
Same author

The P4D Dashboard: A Platform for Monitoring Clinical Studies.

Studies in health technology and informatics·2025
Same author

Personalized Support in Hereditary Breast and Ovarian Cancer After Genetic Counseling by the Chatbot-Based GENIE Mobile App: Proof-of-Concept Wizard of Oz Study.

JMIR formative research·2025

The nnU-Net ensemble model demonstrated superior robustness for pediatric cardiac segmentation compared to nnSAM, showing better geometric accuracy and fewer ejection fraction (EF) failures in external validation. This highlights the need for comprehensive validation of cardiac segmentation services.

Area of Science:

  • Medical Imaging Analysis
  • Artificial Intelligence in Healthcare
  • Cardiovascular Diagnostics

Background:

  • Cardiac segmentation models deployed as remote services operate in a black-box manner, necessitating evaluation of their robustness against out-of-distribution data shifts.
  • Assessing the reliability of these models is crucial for clinical applications, particularly when they influence critical diagnostic biomarkers.

Purpose of the Study:

  • To evaluate and compare the robustness of two cardiac segmentation model configurations, a nnU-Net ensemble and nnSAM, for biventricular segmentation.
  • To assess their performance on external datasets, including adult and pediatric cohorts, focusing on geometric accuracy and biomarker-based endpoints like ejection fraction (EF) failure rates.

Main Methods:

  • Two configurations, nnU-Net ensemble and nnSAM, were trained on a public benchmark dataset for short-axis cine cardiac MRI biventricular segmentation.
Keywords:
Cardiac segmentationOMISAX CMRremote deployment

More Related Videos

Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images
06:48

Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images

Published on: January 7, 2019

Related Experiment Videos

Last Updated: Jul 4, 2026

Pioneering Patient-Specific Approaches for Precision Surgery Using Imaging and Virtual Reality
06:18

Pioneering Patient-Specific Approaches for Precision Surgery Using Imaging and Virtual Reality

Published on: April 5, 2024

Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images
06:48

Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images

Published on: January 7, 2019

  • Models were evaluated using a deployment-oriented framework on two distinct external cohorts: a multi-site, multi-vendor adult dataset and a single-site, multi-scanner pediatric congenital heart disease dataset.
  • Performance metrics included geometric measures and biomarker-based endpoints, specifically ejection fraction (EF) failure rates.

    • Main Results:

    • Both models performed similarly on the adult cohort.
    • The nnU-Net ensemble exhibited greater robustness on the pediatric cohort, demonstrating superior geometric preservation and significantly lower ejection fraction (EF) failure rates.
    • This suggests differential performance based on patient population and data characteristics.

    Conclusions:

    • Deployment-oriented validation of remote cardiac segmentation services is essential and should encompass both geometric accuracy and the reliability of derived clinical biomarkers.
    • The nnU-Net ensemble appears more suitable for robust pediatric cardiac segmentation in a remote inference setting compared to nnSAM.
    • Future validation frameworks must consider diverse datasets and biomarker-based outcomes to ensure clinical utility.