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

InstantGroup: Instant Template Generation for Scalable Group of Brain MRI Registration.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2025
Same author

Depth-Aware Networks for Multi-Organ Lesion Detection in Chest CT Scans.

Bioengineering (Basel, Switzerland)·2024
Same author

EfficientQ: An efficient and accurate post-training neural network quantization method for medical image segmentation.

Medical image analysis·2024
Same author

Retinal Vessel Segmentation by a Transformer-U-Net Hybrid Model With Dual-Path Decoder.

IEEE journal of biomedical and health informatics·2024
Same author

A Dual Coordinate System Vertebra Landmark Detection Network with Sparse-to-Dense Vertebral Line Interpolation.

Bioengineering (Basel, Switzerland)·2024
Same author

Unsupervised Domain Adaptation for Medical Image Segmentation Using Transformer With Meta Attention.

IEEE transactions on medical imaging·2023

Related Experiment Video

Updated: Oct 12, 2025

Patient-Specific Polyvinyl Alcohol Phantom Fabrication with Ultrasound and X-Ray Contrast for Brain Tumor Surgery Planning
08:41

Patient-Specific Polyvinyl Alcohol Phantom Fabrication with Ultrasound and X-Ray Contrast for Brain Tumor Surgery Planning

Published on: July 14, 2020

8.7K

Relax and focus on brain tumor segmentation.

Pei Wang1, Albert C S Chung1

  • 1Lo Kwee-Seong Medical Image Analysis Laboratory, Department of Computer Science and Engineering, The Hong Kong University of Science and Technology, Hong Kong.

Medical Image Analysis
|November 20, 2021
PubMed
Summary

This study introduces a hybrid Deep Convolutional Neural Network (CNN) model for accurate brain tumor segmentation in MRI scans. The model enhances segmentation of challenging tumor sub-regions and improves overall performance for gliomas.

Keywords:
Attention networkBrain tumor segmentationData imbalanceDynamic loss

More Related Videos

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
14:08

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Published on: April 13, 2013

42.9K
Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
04:25

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Published on: December 15, 2023

3.1K

Related Experiment Videos

Last Updated: Oct 12, 2025

Patient-Specific Polyvinyl Alcohol Phantom Fabrication with Ultrasound and X-Ray Contrast for Brain Tumor Surgery Planning
08:41

Patient-Specific Polyvinyl Alcohol Phantom Fabrication with Ultrasound and X-Ray Contrast for Brain Tumor Surgery Planning

Published on: July 14, 2020

8.7K
Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images
14:08

Automated Midline Shift and Intracranial Pressure Estimation based on Brain CT Images

Published on: April 13, 2013

42.9K
Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies
04:25

Author Spotlight: Bridging Gaps in Anatomy and Establishing a Foundation for Algorithmic Studies

Published on: December 15, 2023

3.1K

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Neuroscience

Background:

  • Brain tumors, particularly gliomas, exhibit significant variability in size, location, and morphology, complicating segmentation.
  • Accurate segmentation of brain tumors is crucial for diagnosis, treatment planning, and monitoring disease progression.
  • Challenges in brain tumor segmentation include data imbalance and the need to differentiate various sub-regions like enhancing tumor, necrotic core, and edema.

Purpose of the Study:

  • To develop a fully automatic Deep Convolutional Neural Network (CNN) model for multi-modality, multi-class brain tumor segmentation in MRI images.
  • To address the challenges of high- and low-grade gliomas, including variations in tumor appearance and data imbalance among sub-regions.
  • To improve the segmentation accuracy of challenging tumor sub-regions and the overall structural relationship within brain tumors.

Main Methods:

  • A hybrid Deep Convolutional Neural Network (CNN) model incorporating a dynamic focal Dice loss function for improved focus on smaller, complex tumor sub-regions.
  • Relaxation of inner boundary constraints in a coarse-to-fine manner to better capture overall tumor structure and inter-sub-region relationships.
  • Implementation of a symmetric attention branch to highlight potential tumor locations by analyzing asymmetric features caused by tumor growth.

Main Results:

  • The proposed hybrid CNN model demonstrated improved overall segmentation performance compared to state-of-the-art methods on the BRATS 2019 dataset.
  • Significant advancements were observed in recognizing tumor shape, the structural relationships among different tumor sub-regions, and segmenting difficult areas like the tumor core and enhancing tumor.
  • The dynamic focal Dice loss and attention mechanisms effectively balanced spatial details with high-level morphological features for enhanced segmentation.

Conclusions:

  • The developed hybrid CNN model offers a robust solution for fully automatic brain tumor segmentation, particularly for high- and low-grade gliomas.
  • The novel loss function and architectural components effectively handle data imbalance and complex tumor morphologies, leading to superior segmentation outcomes.
  • This approach advances the field of medical image analysis, providing a valuable tool for neuro-oncology and clinical decision-making.