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

Understanding language model scaling for protein fitness prediction.

Nature computational science·2026
Same author

Lipocalin-2-mediated oligodendrocyte pyroptosis contributes to post-stroke secondary white matter injury.

Brain research bulletin·2026
Same author

A novel score to predict successful endovascular recanalization of non-acute carotid artery occlusion.

Journal of neurointerventional surgery·2026
Same author

Flexible, super-resolution skin with in-sensor hyperdimensional computing for real-time pressure field perception.

Science advances·2026
Same author

A stacked multi-classifier for multi-modal data fusion in transcranial sonography-based Parkinson's disease assessment.

NPJ Parkinson's disease·2026
Same author

Renal IGFBP6 Interacts With THBS1 to Drive Renal Cellular Senescence and Fibrosis.

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

Related Experiment Video

Updated: Sep 5, 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

Simultaneous vessel segmentation and unenhanced prediction using self-supervised dual-task learning in 3D CTA

Wenjian Huang1, Weizheng Gao1, Chao Hou2

  • 1Academy for Advanced Interdisciplinary Studies, Peking University, No.5 Yiheyuan Rd., Beijing, 100871, China.

Computer Methods and Programs in Biomedicine
|July 10, 2022
PubMed
Summary

This study introduces a novel deep learning method for head and neck CT angiography (HNCTA) vessel segmentation and virtual unenhanced (VU) image prediction. The approach significantly improves segmentation accuracy and offers a radiation dose reduction strategy.

Keywords:
All-vessel segmentationDual-task learningHead-neck CTASelf-supervised learningUnenhanced predictionVirtual unenhanced image

More Related Videos

Time-Resolved, Dynamic Computed Tomography Angiography for Characterization of Aortic Endoleaks and Treatment Guidance via 2D-3D Fusion-Imaging
09:32

Time-Resolved, Dynamic Computed Tomography Angiography for Characterization of Aortic Endoleaks and Treatment Guidance via 2D-3D Fusion-Imaging

Published on: December 9, 2021

3.1K
Improved Registration of 3D CT Angiography with X-ray Fluoroscopy for Image Fusion During Transcatheter Aortic Valve Implantation
06:59

Improved Registration of 3D CT Angiography with X-ray Fluoroscopy for Image Fusion During Transcatheter Aortic Valve Implantation

Published on: June 3, 2018

10.7K

Related Experiment Videos

Last Updated: Sep 5, 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
Time-Resolved, Dynamic Computed Tomography Angiography for Characterization of Aortic Endoleaks and Treatment Guidance via 2D-3D Fusion-Imaging
09:32

Time-Resolved, Dynamic Computed Tomography Angiography for Characterization of Aortic Endoleaks and Treatment Guidance via 2D-3D Fusion-Imaging

Published on: December 9, 2021

3.1K
Improved Registration of 3D CT Angiography with X-ray Fluoroscopy for Image Fusion During Transcatheter Aortic Valve Implantation
06:59

Improved Registration of 3D CT Angiography with X-ray Fluoroscopy for Image Fusion During Transcatheter Aortic Valve Implantation

Published on: June 3, 2018

10.7K

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Radiology

Background:

  • Accurate vessel segmentation in CT angiography (CTA) is crucial for diagnosis and hemodynamic analysis.
  • Virtual unenhanced (VU) CT images aid diagnosis and reduce radiation dose but are not available from single-energy CT.
  • Head and neck CTA (HNCTA) vessel segmentation presents unique challenges.

Purpose of the Study:

  • To develop a self-supervised, dual-task deep learning strategy for fully automatic vessel segmentation and unenhanced CT image prediction from single-energy HNCTA.
  • To leverage the correlation between segmentation and image prediction tasks to enhance performance without manual annotation.
  • To provide a potential method for radiation dose reduction in CT imaging.

Main Methods:

  • A self-supervised, dual-task deep learning framework utilizing an iterative residual-sharing scheme was proposed.
  • The model performs simultaneous vessel segmentation and unenhanced CT image prediction from single-energy HNCTA.
  • Annotation-free training was achieved by exploiting the inherent correlation between the two tasks.

Main Results:

  • The proposed model achieved a significantly higher average Dice coefficient (84.83%) for vessel segmentation compared to state-of-the-art methods.
  • The unenhanced image prediction task yielded an average ROI-based error of 6.1±4.5 HU in artery tissue, comparable to existing VU reconstruction methods.
  • The feasibility was validated on data from 24 patients.

Conclusions:

  • The dual-task framework effectively enhances vessel segmentation accuracy in HNCTA.
  • Predicting unenhanced images from single-energy CTA is feasible, offering a potential radiation dose-saving approach.
  • This represents the first reported annotation-free deep learning-based full-image vessel segmentation for HNCTA.