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

A survey on deep learning models for detection of COVID-19.

Neural computing & applications·2023
Same author

MedViT: A robust vision transformer for generalized medical image classification.

Computers in biology and medicine·2023
Same author

Automatic breast lesion detection in ultrafast DCE-MRI using deep learning.

Medical physics·2021
Same author

Differentiating benign and malignant mass and non-mass lesions in breast DCE-MRI using normalized frequency-based features.

International journal of computer assisted radiology and surgery·2019
Same author

Localized-atlas-based segmentation of breast MRI in a decision-making framework.

Australasian physical & engineering sciences in medicine·2017
Same author

Fast opposite weight learning rules with application in breast cancer diagnosis.

Computers in biology and medicine·2012

Related Experiment Video

Updated: Jul 5, 2025

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

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

405

ColonGen: an efficient polyp segmentation system for generalization improvement using a new comprehensive dataset.

Javad Mozaffari1, Abdollah Amirkhani2, Shahriar B Shokouhi1

  • 1School of Electrical Engineering, Iran University of Science and Technology, Tehran, 16846-13114, Iran.

Physical and Engineering Sciences in Medicine
|January 15, 2024
PubMed
Summary

This study introduces ColonGen, a novel deep learning model using vision transformer architectures for improved colorectal cancer polyp segmentation during colonoscopy. ColonGen-V2 demonstrates superior performance, enhancing diagnostic accuracy and reducing miss rates.

Keywords:
Colorectal cancerConvolutional neural networksDatasetPolyp segmentationTransformers

More Related Videos

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data
12:08

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data

Published on: August 13, 2014

24.6K
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

2.4K

Related Experiment Videos

Last Updated: Jul 5, 2025

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

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

405
From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data
12:08

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data

Published on: August 13, 2014

24.6K
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

2.4K

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Oncology

Background:

  • Colorectal cancer (CRC) remains a leading cause of cancer mortality.
  • Effective polyp detection during colonoscopy is crucial for CRC diagnosis, but current methods suffer from high miss rates.
  • Deep learning, particularly Convolutional Neural Networks (CNNs), is widely used for image analysis, but Vision Transformers (ViTs) show promise for improved spatial information extraction.

Purpose of the Study:

  • To investigate the generalization capabilities of polyp image segmentation using transformer architectures.
  • To propose a novel approach combining multiple ViT architectures for enhanced feature representation.
  • To develop a more universal dataset to improve model generalization for polyp segmentation.

Main Methods:

  • Proposed a novel polyp segmentation approach utilizing two distinct Vision Transformer (ViT) architectures to learn diverse feature representations.
  • Created a comprehensive dataset by integrating existing research datasets to enhance model generalization.
  • Evaluated model performance through distinct training-testing scenarios and against in- and out-of-domain data.

Main Results:

  • The initial model, ColonGen-V1, outperformed state-of-the-art methods across three different training-testing scenarios.
  • ColonGen-V2, trained on the comprehensive dataset, achieved superior performance, outperforming existing studies by 5.1% on ETIS-Larib, 1.3% on Kvasir-Seg, and 1.1% on CVC-ColonDB.
  • The developed model and dataset demonstrated significant improvements in generalization proficiency for polyp segmentation.

Conclusions:

  • The proposed dual-ViT architecture and comprehensive dataset significantly enhance polyp segmentation accuracy and generalization.
  • This approach offers a promising solution to improve colonoscopy-based colorectal cancer screening by reducing polyp miss rates.
  • The publicly available model and dataset facilitate further research and development in automated polyp detection.