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

Physics Insights into Liver MRI: Educational Guidance for Protocol Optimization.

Radiographics : a review publication of the Radiological Society of North America, Inc·2026
Same author

Multimodal PET/CT-based PD-L1 status prediction in lung cancer via semi-supervised and unsupervised deep learning.

Scientific reports·2026
Same author

Radiomic prediction of substantial LVSI in endometrial cancer using reduced field of view DWI- a feasibility study.

European journal of radiology·2026
Same author

AI-driven volumetric approach for automatic chemotherapy response assessment in colorectal liver metastases.

European radiology·2026
Same author

CT-based deep learning prediction of complete response in intermediate-stage hepatocellular carcinoma treated with drug-eluting beads transarterial chemoembolization.

BJR artificial intelligence·2026
Same author

Radiotherapy and brachytherapy in the management of adult soft tissue sarcomas.

Cancer radiotherapie : journal de la Societe francaise de radiotherapie oncologique·2026

Related Experiment Video

Updated: Nov 11, 2025

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

638

Convolutional neural networks for PET functional volume fully automatic segmentation: development and validation in a

Andrei Iantsen1, Marta Ferreira2, Francois Lucia3

  • 1LaTIM, INSERM, UMR 1101, University Brest, Brest, France. andrei.iantsen@inserm.fr.

European Journal of Nuclear Medicine and Molecular Imaging
|March 27, 2021
PubMed
Summary

This study presents a fully automatic method for determining tumor uptake from PET images, crucial for cervical cancer treatment planning. The developed U-Net model accurately delineates tumors, even near the bladder, outperforming existing methods.

Keywords:
Cervical cancerConvolutional neural networkPETSegmentationU-Net

More Related Videos

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

3.1K
Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
04:25

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Published on: December 15, 2023

3.3K

Related Experiment Videos

Last Updated: Nov 11, 2025

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

638
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

3.1K
Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
04:25

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Published on: December 15, 2023

3.3K

Area of Science:

  • Medical Imaging
  • Radiomics
  • Artificial Intelligence in Oncology

Background:

  • Accurate tumor delineation in Positron Emission Tomography (PET) is vital for cancer treatment, especially in cervical cancer where tumors are often near the bladder.
  • Multicenter studies present challenges due to heterogeneous image characteristics, necessitating robust automated methods.
  • Manual or semi-automated segmentation can be time-consuming and prone to inter-observer variability.

Purpose of the Study:

  • To develop a fully automatic method for determining tumor functional uptake from PET images without external constraints.
  • To address the challenge of tumor delineation near the bladder in cervical cancer PET imaging.
  • To create a robust segmentation model applicable to multicenter, heterogeneous PET datasets.

Main Methods:

  • A novel U-Net based model incorporating residual blocks, spatial squeeze and excitation modules, and learnable down/upsampling was developed.
  • Ground truth for training was generated using a semi-automated approach combining manual tumor/bladder exclusion with the Fuzzy locally Adaptive Bayesian (FLAB) algorithm.
  • The model was trained and validated using cross-validation on PET datasets from 232 patients across five institutions.

Main Results:

  • The proposed model achieved a high Dice Similarity Coefficient (DSC) of 0.80 ± 0.03, demonstrating minimal variability across institutions.
  • The model exhibited superior performance compared to the standard U-Net (DSC 0.77 ± 0.05) and significantly outperformed a fixed 40% SUVmax threshold (DSC 0.33 ± 0.15).
  • Tumor uptake was accurately determined without including the bladder, and the model did not require shape priors or anatomical information.

Conclusions:

  • The developed fully automated method effectively delineates tumor functional uptake in challenging multicenter PET data.
  • This approach can significantly streamline the deployment of automated radiomics pipelines in clinical settings.
  • The model's ability to handle heterogeneous data and avoid bladder inclusion offers a significant advancement for cervical cancer imaging analysis.