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: Nov 29, 2025

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

3.2K

Toward reliable automatic liver and tumor segmentation using convolutional neural network based on 2.5D models.

Girindra Wardhana1, Hamid Naghibi2, Beril Sirmacek2

  • 1Department of Robotics and Mechatronics, The Faculty of Electrical Engineering, Mathematics and Computer Science, Technical Medical Centre, University of Twente, 7522 NB, Enschede, The Netherlands. g.wardhana@utwente.nl.

International Journal of Computer Assisted Radiology and Surgery
|November 21, 2020
PubMed
Summary

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

Modeling age and sex variations of arterial pressure and cardiac hemodynamics in a virtual population using a 0D-1D cardiovascular simulator.

American journal of physiology. Heart and circulatory physiology·2025
Same author

Dynamic abdominal MRI image generation using cGANs: A generalized model for various breathing patterns with extensive evaluation.

Computers in biology and medicine·2025
Same author

Experimental evaluation of virtual needle insertion framework with enhanced haptic feedback.

International journal of computer assisted radiology and surgery·2025
Same author

Virtual Needle Insertion with Enhanced Haptic Feedback for Guidance and Needle-Tissue Interaction Forces.

Sensors (Basel, Switzerland)·2024
Same author

IRE made easy: introducing the robotic grid system for multiple parallel needle insertion in irreversible electroporation treatment.

International journal of computer assisted radiology and surgery·2024
Same author

Liver respiratory-induced motion estimation using abdominal surface displacement as a surrogate: robotic phantom and clinical validation with varied correspondence models.

International journal of computer assisted radiology and surgery·2024

Optimizing 2.5D convolutional neural networks for liver and tumor segmentation requires careful parameter tuning. The study found that three stacked layers yield optimal performance, with more layers leading to overfitting and no significant benefit from contrast enhancement.

Area of Science:

  • Medical Imaging
  • Artificial Intelligence

Background:

  • Automatic liver and tumor segmentation is crucial for clinical diagnosis.
  • 2.5D convolutional neural networks (CNNs) offer a promising approach by integrating 3D information with deeper architectures.
  • Detailed parameter optimization for 2.5D CNNs in this context is lacking.

Purpose of the Study:

  • To investigate the impact of parameter configurations on 2.5D CNN performance for automatic liver and tumor segmentation.
  • To identify optimal parameter settings for improved segmentation accuracy.

Main Methods:

  • Studied parameters include the number of stacked layers, image contrast enhancement, and network depth.
  • Trained and tested 2.5D CNN models using the Liver and Tumor Segmentation (LiTS) dataset.
  • Evaluated segmentation performance against manual segmentations by multiple physicians and a radiologist.
Keywords:
CT imageConvolutional neural networkDeep learningImage segmentationLiver tumor

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

659

Related Experiment Videos

Last Updated: Nov 29, 2025

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

3.2K
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

659

Main Results:

  • Performance improved with multiple stacked layers, peaking at three layers.
  • Exceeding three stacked layers resulted in decreased Dice scores due to overfitting.
  • Contrast enhancement did not significantly alter network performance.
  • Increasing network layers did not consistently improve Dice scores.

Conclusions:

  • Parameter configuration significantly affects 2.5D CNN performance in liver and tumor segmentation.
  • Optimal configuration involves a balance to prevent overfitting and maximize accuracy.
  • Findings guide the selection of best parameters for enhanced automatic segmentation.