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

Imaging foundation model for universal enhancement of non-ideal measurement CT.

Nature communications·2026
Same author

The current and future landscape of AI foundation models for cancer management.

Nature communications·2026
Same author

MILU: a consensus ensemble benchmark for multimodal medical imaging lecture understanding.

Journal of medical imaging (Bellingham, Wash.)·2026
Same author

MEDI-SLATE: medical imaging slide-lecture aligned teaching ensemble.

Visual computing for industry, biomedicine, and art·2026
Same author

Manifold topological deep learning for biomedical data.

Nature communications·2026
Same author

Artificial-intelligence-driven segmentation and analysis of microbial cells.

Journal of biomedical optics·2026

Related Experiment Video

Updated: Aug 28, 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

1.9K

Unsupervised contrastive learning based transformer for lung nodule detection.

Chuang Niu1, Ge Wang1

  • 1Biomedical Imaging Center, Department of Biomedical Engineering, Rensselaer Polytechnic Institute, Troy, New York, United States of America.

Physics in Medicine and Biology
|September 16, 2022
PubMed
Summary
This summary is machine-generated.

A novel self-supervised 3D transformer model significantly improves lung nodule detection accuracy in computed tomography (CT) scans. This advancement enhances computer-aided detection (CAD) systems, reducing false positives and aiding early lung cancer diagnosis.

Keywords:
deep learninglung nodule detectiontransformerunsupervised pretraining

More Related Videos

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

2.0K
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

1.6K

Related Experiment Videos

Last Updated: Aug 28, 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

1.9K
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

2.0K
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

1.6K

Area of Science:

  • Medical imaging analysis
  • Artificial intelligence in healthcare
  • Radiology

Background:

  • Early lung nodule detection via computed tomography (CT) is crucial for improving lung cancer patient survival and quality of life.
  • Computer-aided detection/diagnosis (CAD) systems assist radiologists but struggle with nodule variability and complex lung structures, leading to high false-positive rates.
  • Current CAD systems require improvement for accurate and efficient lung nodule identification.

Purpose of the Study:

  • To develop a self-supervised region-based 3D transformer model for enhanced lung nodule detection in CT images.
  • To address the challenges of nodule variability and complex lung anatomy that hinder CAD system performance.
  • To improve the clinical efficacy of CAD systems by reducing false positives and increasing detection accuracy.

Main Methods:

  • A 3D vision transformer model was developed, processing CT volumes as sequences of non-overlapping cubes.
  • An embedding layer extracted features from each cube, analyzed via a self-attention mechanism for nodule prediction.
  • Region-based contrastive learning was employed for pre-training the transformer on public CT data to enhance performance on smaller datasets.

Main Results:

  • The proposed 3D transformer model demonstrated significantly improved performance in lung nodule screening compared to traditional 3D convolutional neural networks.
  • Experiments confirmed the method's effectiveness in enhancing the accuracy of lung nodule detection.
  • The self-supervised approach proved beneficial for training on limited medical imaging datasets.

Conclusions:

  • The self-supervised region-based 3D transformer model offers a promising advancement for computer-aided lung nodule detection.
  • This approach has the potential to significantly improve the performance and clinical utility of current CAD systems.
  • Further development in AI-driven medical imaging analysis can lead to better early cancer detection and patient outcomes.