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

Control of spiral breakup by an alternating advective field.

The Journal of chemical physics·2006
Same author

Inhibition of growth and metastasis of human hepatocellular carcinoma by antisense oligonucleotide targeting signal transducer and activator of transcription 3.

Clinical cancer research : an official journal of the American Association for Cancer Research·2006
Same author

New rod-plate anterior instrumentation for thoracolumbar/lumbar scoliosis: biomechanical evaluation compared with dual-rod and single-rod with structural interbody support.

Spine·2006
Same author

Dimeric ansamycins--a new class of antitumor Hsp90 modulators with prolonged inhibitory activity.

International journal of cancer·2006
Same author

A unified mode of epigenetic gene silencing: RNA meets polycomb group proteins.

RNA biology·2006
Same author

Molecular signaling and genetic pathways of senescence: Its role in tumorigenesis and aging.

Journal of cellular physiology·2006

Related Experiment Video

Updated: Jul 9, 2025

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
08:20

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Published on: October 27, 2023

1.5K

Self-supervised representation learning using feature pyramid siamese networks for colorectal polyp detection.

Tianyuan Gan1, Ziyi Jin1, Liangliang Yu2

  • 1Biosensor National Special Laboratory, College of Biomedical Engineering and Instrument Science, Zhejiang University, Hangzhou, 310027, China.

Scientific Reports
|December 8, 2023
PubMed
Summary

A new self-supervised learning method, Feature Pyramid Siamese Networks (FPSiam), improves colorectal polyp detection using unlabeled colonoscopy data. This approach reduces the need for expert annotations, enhancing diagnostic accuracy and efficiency in cancer screening.

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

2.8K
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

406

Related Experiment Videos

Last Updated: Jul 9, 2025

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images
08:20

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Published on: October 27, 2023

1.5K
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
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

406

Area of Science:

  • Medical Imaging
  • Artificial Intelligence in Medicine
  • Gastroenterology

Background:

  • Colorectal cancer is a significant global health concern, with early detection crucial for improving patient outcomes.
  • Computer-aided diagnosis (CAD) systems, particularly those using convolutional neural networks, show promise for real-time polyp detection during colonoscopy.
  • Current CAD methods often require extensive, expert-annotated datasets, posing a barrier to development and widespread adoption.

Purpose of the Study:

  • To introduce a novel self-supervised representation learning method, Feature Pyramid Siamese Networks (FPSiam), for colorectal polyp detection.
  • To leverage large unlabeled colonoscopy datasets to overcome the limitations of manual annotation.
  • To enhance the performance of polyp detection systems in resource-constrained scenarios with limited labeled data.

Main Methods:

  • Developed a feature pyramid encoder to extract and fuse local and global features from colonoscopic images.
  • Employed siamese networks for self-supervised learning to acquire general visual feature representations from unlabeled data.
  • Transferred learned features to downstream colorectal polyp detection tasks using standard detectors.

Main Results:

  • The proposed FPSiam method achieved optimal performance in colorectal polyp detection.
  • FPSiam outperformed other state-of-the-art self-supervised learning methods.
  • Compared to transfer learning methods, FPSiam demonstrated superior results, achieving 2.3 mAP and 3.6 mAP higher with two typical detectors.

Conclusions:

  • FPSiam offers a cost-efficient solution for developing robust colorectal polyp detection systems, especially when labeled data is scarce.
  • The method effectively utilizes unlabeled colonoscopy data, reducing the reliance on laborious expert annotations.
  • FPSiam presents a promising approach for advancing various endoscopic image analysis tasks beyond polyp detection.