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

Charting a new melanoma risk factor: Satellite Naevus Prevalence in High-Risk and Population-Based Cohorts.

The Journal of investigative dermatology·2026
Same author

Generalised Medical Phrase Grounding.

IEEE transactions on medical imaging·2026
Same author

Interpretable Semantic Medical Image Segmentation with Style and Confidence.

IEEE transactions on pattern analysis and machine intelligence·2026
Same author

It's not your device, it's how you use it - Sensitivity and specificity of PSQA devices in the SEAFARER head and neck study.

Radiotherapy and oncology : journal of the European Society for Therapeutic Radiology and Oncology·2026
Same author

The impact of bladder and rectal dynamics on prostate and seminal vesicles intrafraction motion and deformation in radiotherapy.

Physics and imaging in radiation oncology·2026
Same author

The long and winding road of radiomics: learnings from two meta-analyses of the radiomics quality score.

Physics in medicine and biology·2026

Related Experiment Video

Updated: Aug 30, 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

2.9K

CAN3D: Fast 3D medical image segmentation via compact context aggregation.

Wei Dai1, Boyeong Woo1, Siyu Liu1

  • 1School of Information Technology and Electrical Engineering, The University of Queensland, Australia.

Medical Image Analysis
|September 1, 2022
PubMed
Summary
This summary is machine-generated.

This study introduces CAN3D, a compact deep learning model for real-time 3D medical image segmentation on clinical workstations. It achieves high accuracy with reduced computational costs and memory usage, outperforming existing methods.

Keywords:
3DDeep learningMagnetic resonance imagingProstate and kneeSemantic segmentation

More Related Videos

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data
12:08

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data

Published on: August 13, 2014

24.6K
Author Spotlight: Segmentation and VR for Advanced Neurovascular Interventions
06:18

Author Spotlight: Segmentation and VR for Advanced Neurovascular Interventions

Published on: April 5, 2024

1.2K

Related Experiment Videos

Last Updated: Aug 30, 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

2.9K
From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data
12:08

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data

Published on: August 13, 2014

24.6K
Author Spotlight: Segmentation and VR for Advanced Neurovascular Interventions
06:18

Author Spotlight: Segmentation and VR for Advanced Neurovascular Interventions

Published on: April 5, 2024

1.2K

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Computer Vision

Background:

  • Direct automatic segmentation of 3D medical images (e.g., MRI) is complex due to intricate structures and large data volumes.
  • Current deep learning models often require significant computational resources and memory, limiting their use on standard clinical workstations.
  • Existing methods struggle with real-time processing on low-end hardware, hindering clinical application.

Purpose of the Study:

  • To develop a compact convolutional neural network (CAN3D) for efficient, real-time 3D medical image segmentation.
  • To enable state-of-the-art segmentation performance on clinical workstations with limited computational resources.
  • To improve segmentation accuracy for imbalanced classes in 3D MR images using a novel loss function.

Main Methods:

  • Designed CAN3D, a shallow convolutional neural network with a reduced memory footprint for direct processing of full-size 3D MR image volumes.
  • Developed a novel loss function incorporating shape constraints to enhance segmentation of imbalanced classes.
  • Evaluated CAN3D on knee and pelvis 3D MR datasets, comparing performance against U-Net3D, improved U-Net3D, and V-Net.

Main Results:

  • CAN3D significantly reduced model parameters (up to two orders of magnitude) and achieved up to 5x faster CPU inference compared to state-of-the-art methods.
  • Achieved high Dice coefficients (e.g., 0.87 ± 0.02 for femoral cartilage) and low surface distance errors on the OAI-ZIB knee MR dataset with limited video memory (12G).
  • Demonstrated high accuracy and efficiency on a publicly released pelvis 3D MR imaging dataset for prostate cancer segmentation.

Conclusions:

  • CAN3D offers a computationally efficient and memory-light solution for real-time 3D medical image segmentation on clinical workstations.
  • The proposed architecture and novel loss function enable accurate segmentation of complex structures, even with imbalanced data.
  • CAN3D represents a significant advancement for deploying advanced deep learning segmentation models in routine clinical practice.