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

Enhanced invasiveness promotes the dominance of a widely-distributed carbapenem-resistant virulence-plasmid-carrying Klebsiella pneumoniae sublineage.

Nature communications·2026
Same author

Targeted delivery of kaempferol via mannose-modified PLGA nanoparticles reprograms macrophages and ameliorates rheumatoid arthritis.

International journal of pharmaceutics·2026
Same author

Mediastinal pseudocyst presented as oesophageal mass and melena: a case report.

Journal of surgical case reports·2026
Same author

Efficacy and pharmacokinetics of a stepwise isolated phage cocktail against carbapenem-resistant Klebsiella pneumoniae in a murine pneumonia model.

International journal of antimicrobial agents·2026
Same author

Genome sequence of lytic phage phi1_164023 targeting ST23 KL1-type carbapenem-resistant <i>Klebsiella pneumoniae</i>.

Microbiology resource announcements·2026
Same author

An interpretable model combining brain imaging and clinical indicators for predicting overt hepatic encephalopathy.

Analytical biochemistry·2026

Related Experiment Video

Updated: Jul 19, 2025

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

2.8K

3D PET/CT tumor segmentation based on nnU-Net with GCN refinement.

Hengzhi Xue1, Qingqing Fang1, Yudong Yao1,2

  • 1College of Medicine and Biological Information Engineering, Northeastern University, Shenyang 110004, People's Republic of China.

Physics in Medicine and Biology
|August 7, 2023
PubMed
Summary

This study introduces a graph convolutional network (GCN) postprocessing method to improve tumor segmentation accuracy in positron emission tomography/computed tomography (PET/CT) scans. The novel nnU-Net + GCN framework refines initial segmentations, significantly reducing errors and enhancing diagnostic precision for various cancers.

Keywords:
graph convolutional network (GCN)image segmentationpositron emission tomography/computed tomography (PET/CT)

More Related Videos

Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer
07:53

Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer

Published on: October 13, 2023

1.5K
Author Spotlight: A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules
10:26

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

Published on: May 19, 2023

1.9K

Related Experiment Videos

Last Updated: Jul 19, 2025

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

2.8K
Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer
07:53

Author Spotlight: Advancing 3D Modeling for Enhanced Diagnosis and Treatment of Pulmonary Nodules in Early-Stage Lung Cancer

Published on: October 13, 2023

1.5K
Author Spotlight: A 3D Digital Model for the Diagnosis and Treatment of Pulmonary Nodules
10:26

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

Published on: May 19, 2023

1.9K

Area of Science:

  • Medical Imaging and Radiation Oncology
  • Artificial Intelligence in Healthcare
  • Computational Pathology

Background:

  • Whole-body positron emission tomography/computed tomography (PET/CT) scans are crucial for diagnosing malignancies like melanoma, lymphoma, and lung cancer.
  • Accurate tumor segmentation is essential for effective cancer treatment planning.
  • Existing convolutional neural network (CNN) segmentation methods often yield inaccurate results, including oversegmentation and undersegmentation.

Purpose of the Study:

  • To propose a novel postprocessing method using a graph convolutional network (GCN) to refine inaccurate tumor segmentations from PET/CT scans.
  • To improve the overall accuracy of tumor segmentation by addressing oversegmentation and undersegmentation issues.
  • To develop an enhanced segmentation framework by combining nnU-Net with GCN.

Main Methods:

  • Utilized nnU-Net for initial tumor segmentation on PET/CT datasets.
  • Analyzed segmentation uncertainty to establish graph nodes (certain and uncertain pixels).
  • Employed a semisupervised graph network approach, training the GCN with nnU-Net's accurate segmentations as labels to optimize uncertain regions.

Main Results:

  • The proposed nnU-Net + GCN framework effectively reduced the false-positive rate in tumor segmentation.
  • Demonstrated significant quantitative improvements: a 2.1% increase in average Dice score.
  • Showcased enhanced accuracy with improvements of 6.4 in 95% Hausdorff distance (HD95) and 1.7 in average symmetric surface distance.

Conclusions:

  • Graph convolutional network (GCN) based postprocessing is a highly effective method for refining tumor segmentation in PET/CT imaging.
  • The nnU-Net + GCN framework offers a significant advancement in improving segmentation accuracy and reliability for oncological applications.
  • This approach holds promise for enhancing diagnostic precision and treatment planning in cancer management.