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: Oct 2, 2025

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

2.0K

Self-Supervised Bi-Channel Transformer Networks for Computer-Aided Diagnosis.

Ronglin Gong, Xiangmin Han, Jun Wang

    IEEE Journal of Biomedical and Health Informatics
    |February 24, 2022
    PubMed
    Summary

    This study introduces Self-Supervised Bi-channel Transformer Networks (SSBTN), a novel approach to computer-aided diagnosis (CAD). SSBTN enhances diagnostic accuracy by improving self-supervised learning (SSL) flexibility, especially for small datasets.

    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

    Vorinostat: a histone deacetylases (HDAC) inhibitor ameliorates traumatic brain injury by inducing iNOS/Nrf2/ARE pathway.

    Folia neuropathologica·2018
    Same author

    Synthesis, crystal structure and antimicrobial activity of 2-((2-(4-(1H-1,2,4-triazol-1-yl)phenyl)quinazolin-4-yl)oxy)-N-phenylacetamide derivatives against phytopathogens.

    Molecular diversity·2018
    Same author

    Downregulation of Blimp1 inhibits the maturation of bone marrow-derived dendritic cells.

    International journal of molecular medicine·2018
    Same author

    Multiple Kernel Learning Based Classification of Parkinson's Disease With Multi-Modal Transcranial Sonography.

    Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2018
    Same author

    Transcranial Sonography Based Diagnosis Of Parkinson's Disease Via Cascaded Kernel RVFL.

    Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2018
    Same author

    Speckle Suppression of Ultrasonography Using Maximum Likelihood Estimation and Weighted Nuclear Norm Minimization.

    Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2018

    Area of Science:

    • Artificial Intelligence
    • Medical Imaging
    • Machine Learning

    Background:

    • Self-supervised learning (SSL) is effective for computer-aided diagnosis (CAD) with limited data.
    • Conventional SSL methods often struggle when pretext and downstream tasks differ, hindering pretext network training.
    • This limitation impacts feature representation quality for CAD models.

    Purpose of the Study:

    • To propose a novel task-driven SSL method, Self-Supervised Bi-channel Transformer Networks (SSBTN), to enhance CAD model diagnostic accuracy.
    • To improve the flexibility of SSL in CAD by decoupling pretext and downstream tasks.
    • To develop a more effective feature representation for CAD through a flexible pretext task design.

    Main Methods:

    • Introduced SSBTN, integrating two distinct networks for pretext and downstream tasks within a unified framework.

    More Related Videos

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

    Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

    Published on: July 5, 2024

    538

    Related Experiment Videos

    Last Updated: Oct 2, 2025

    A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
    04:23

    A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

    Published on: April 21, 2023

    2.0K
    Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
    04:48

    Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

    Published on: July 5, 2024

    538
  • Designed a flexible pretext task tailored to data characteristics for effective feature learning.
  • Developed a transformer-based transfer module for efficient knowledge transfer and feature alignment between networks.

    • Main Results:

    • SSBTN demonstrated superior performance compared to existing algorithms on two public datasets (INbreast and CrohnIPI).
    • The task-driven SSL approach led to more effective feature representations.
    • The integrated framework and transformer-based transfer module enhanced knowledge transfer efficiency.

    Conclusions:

    • SSBTN offers a flexible and effective approach to SSL for CAD, improving diagnostic accuracy.
    • The method successfully addresses limitations of conventional SSL by decoupling task-specific networks.
    • SSBTN shows significant potential for advancing medical image analysis and diagnosis.