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

Group C Rotavirus Gastroenteritis Outbreak in a Primary School - Luzhou City, Sichuan Province, China, October 2025.

China CDC weekly·2026
Same author

Transcranial direct current stimulation alleviates depression-like behaviors by activating the ACC-ZI neural circuit.

Brain research bulletin·2026
Same author

Author Correction: The SESAME complex regulates cell senescence through the generation of acetyl-CoA.

Nature metabolism·2026
Same author

Pericentrosomal Redistribution of the Endoplasmic Reticulum Ensures Organelle Symmetric Inheritance and Mitotic Progression.

Advanced science (Weinheim, Baden-Wurttemberg, Germany)·2026
Same author

Association between prognostic nutritional index and mortality in critically ill patients with atrial fibrillation: a retrospective cohort study.

BMC cardiovascular disorders·2026
Same author

Toxicological Evaluation of the Roots of <i>Ficus pandurata</i> Hance var. <i>angustifolia</i> Cheng.

Food science & nutrition·2026

Related Experiment Video

Updated: Jul 20, 2025

Three-Dimensional Phase Resolved Functional Lung Magnetic Resonance Imaging
10:44

Three-Dimensional Phase Resolved Functional Lung Magnetic Resonance Imaging

Published on: June 21, 2024

544

DAWTran: dynamic adaptive windowing transformer network for pneumothorax segmentation with implicit feature

Pengchen Liang1, Jianguo Chen2, Lei Yao1

  • 1The Department of School of Microelectronics, Shanghai University, Shanghai, 201800, People's Republic of China.

Physics in Medicine and Biology
|August 4, 2023
PubMed
Summary

This study introduces the dynamic adaptive windowing transformer (DAWTran) for improved pneumothorax segmentation in CT scans. The novel network enhances accuracy and reduces errors, offering a valuable tool for clinical diagnosis.

Keywords:
dynamic adaptive windowingimplicit feature alignmentmedical image analysispneumothorax segmentationtransformer

More Related Videos

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

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

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

442

Related Experiment Videos

Last Updated: Jul 20, 2025

Three-Dimensional Phase Resolved Functional Lung Magnetic Resonance Imaging
10:44

Three-Dimensional Phase Resolved Functional Lung Magnetic Resonance Imaging

Published on: June 21, 2024

544
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

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

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

442

Area of Science:

  • Medical Imaging
  • Artificial Intelligence in Medicine
  • Computer-Aided Diagnosis

Background:

  • Pneumothorax segmentation in CT images is challenging due to similar appearances of pneumothorax and gas-filled structures.
  • Accurate segmentation is crucial for timely diagnosis and treatment of pneumothorax.

Purpose of the Study:

  • To develop a novel network for precise pneumothorax segmentation in computed tomography (CT) images.
  • To overcome segmentation difficulties caused by the visual similarity between pneumothorax and anatomical structures like the trachea and bronchus.

Main Methods:

  • Introduction of the dynamic adaptive windowing transformer (DAWTran) network with an encoder-decoder architecture.
  • Implementation of a dynamic adaptive windowing strategy for multi-scale feature extraction via multi-head self-attention.
  • Utilization of implicit feature alignment in the decoder to minimize information deviation.
  • Application of a hybrid loss function to address class imbalance.

Main Results:

  • DAWTran achieved a Dice Similarity Coefficient (DSC) of 91.35%, outperforming TransUNet by 2.21%.
  • The network significantly reduced Hausdorff Distance (HD) to 8.06 mm, a 29.92% improvement over TransUNet.
  • The dynamic adaptive windowing (DAW) mechanism increased DSC by 4.53% and reduced HD by 15.85% compared to SwinUnet.
  • Implicit feature alignment (IFA) further boosted accuracy, increasing DSC by 0.11% and reducing HD by 10.01%.

Conclusions:

  • The DAWTran network demonstrates significant potential for accurate pneumothorax segmentation in clinical settings.
  • The proposed DAW and IFA components effectively enhance segmentation performance.
  • This method offers a promising tool for improving the precision of pneumothorax diagnosis and treatment planning.