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

Correction: Development and validation of a deep learning-based automatic detection and classification model for femoral neck fractures using hip imaging: a retrospective multicenter diagnostic study.

Frontiers in medicine·2026
Same author

Development and validation of a deep learning-based automatic detection and classification model for femoral neck fractures using hip imaging: a retrospective multicenter diagnostic study.

Frontiers in medicine·2026
Same author

Brain tumor classification model guided by class activation mapping.

Computerized medical imaging and graphics : the official journal of the Computerized Medical Imaging Society·2026
Same author

Dataset for Single Character Detection in Dongba Manuscripts.

Scientific data·2025
Same author

Mechanisms of tumor heterogeneity in TACE-resistant liver cancer: Insights from single-cell and whole-exome sequencing.

Hepatology communications·2025
Same author

Dynamically Tunable Optofluidic Multifocal Microlens Arrays by 3D Printing.

ACS sensors·2025

Related Experiment Video

Updated: Jul 3, 2025

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

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

402

ETUNet:Exploring efficient transformer enhanced UNet for 3D brain tumor segmentation.

Wang Zhang1, Shanxiong Chen1, Yuqi Ma1

  • 1School of Computer and Information Science, SouthWest University, China.

Computers in Biology and Medicine
|February 10, 2024
PubMed
Summary

This study introduces an enhanced UNet model integrating Transformer capabilities for improved brain tumor segmentation in MRI scans, achieving superior accuracy and efficiency in clinical applications.

Keywords:
Brain tumor segmentationCross-attentionSpatial-channel attentionTransformerUNet

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
Multicolor 3D Printing of Complex Intracranial Tumors in Neurosurgery
14:15

Multicolor 3D Printing of Complex Intracranial Tumors in Neurosurgery

Published on: January 11, 2020

7.1K

Related Experiment Videos

Last Updated: Jul 3, 2025

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

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

402
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
Multicolor 3D Printing of Complex Intracranial Tumors in Neurosurgery
14:15

Multicolor 3D Printing of Complex Intracranial Tumors in Neurosurgery

Published on: January 11, 2020

7.1K

Area of Science:

  • Medical Image Analysis
  • Artificial Intelligence in Medicine
  • Neuro-oncology Imaging

Background:

  • Accurate medical image segmentation, particularly for brain tumors in MRI, is vital for diagnosis and prognosis.
  • Traditional Convolutional Neural Network (CNN) based UNet models face limitations in capturing long-range dependencies for complex tumor structures.
  • Existing Transformer-UNet hybrids often suffer from high computational costs and overlook multi-scale boundary information.

Purpose of the Study:

  • To develop an advanced fusion of Transformer and UNet architectures for more precise and efficient brain tumor segmentation.
  • To address the limitations of existing methods by improving global information modeling and multi-scale feature utilization.
  • To enhance the capture of deep spatial dependencies and boundary localization for irregular tumor shapes.

Main Methods:

  • Introduced a CNN-Transformer module in the encoder to capture deep spatial dependencies.
  • Incorporated a spatial-channel attention layer in the bottleneck for efficient global semantic feature interaction.
  • Utilized cross-attention in skip connections to fuse multi-scale encoder and decoder features for improved boundary localization.

Main Results:

  • The proposed model demonstrated comparable or superior performance to existing CNN and Transformer-based methods on the BraTS2018 and BraTS2020 datasets.
  • Achieved high segmentation accuracy with average Dice Similarity Coefficient (DSC) of 0.854 and 0.862, and Hausdorff 95% (HD95) of 6.688 and 5.455, respectively.
  • Showcased optimal segmentation of Enhancing tumors, highlighting the model's effectiveness in capturing intricate tumor subregions.

Conclusions:

  • The advanced fusion of Transformer and UNet effectively enhances learning capacity and segmentation accuracy for brain tumors while maintaining low computational complexity.
  • The model's ability to leverage multi-scale features and global information improves boundary delineation and overall segmentation performance.
  • This approach offers a promising tool for clinical applications requiring precise brain tumor segmentation from multimodal MRI.