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: Mar 21, 2026

Author Spotlight: An Efficient and Robust Software for Automated Fusion of Multiple Preclinical Imaging Modalities
07:13

Author Spotlight: An Efficient and Robust Software for Automated Fusion of Multiple Preclinical Imaging Modalities

Published on: October 27, 2023

1.8K

W-STFNet: A Wavelet Transform-Based Regularized Hybrid Recursive Spatiotemporal Fusion Registration Network.

Xing Chen1, Xinyu Liu1, Zhijia Wang1

  • 1School of Control Science and Engineering, Shandong University, Jinan, 250061, China.

Annals of Biomedical Engineering
|March 19, 2026
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

Association of Smoking with Metabolic Volatile Organic Compounds in Exhaled Breath.

International journal of molecular sciences·2017
Same author

Self-assembly of peptide amphiphiles for drug delivery: the role of peptide primary and secondary structures.

Biomaterials science·2017
Same author

NARRMDA: negative-aware and rating-based recommendation algorithm for miRNA-disease association prediction.

Molecular bioSystems·2017
Same author

Ultrasensitive, high-dynamic-range and broadband strain sensing by time-of-flight detection with femtosecond-laser frequency combs.

Scientific reports·2017
Same author

MicroRNAs and complex diseases: from experimental results to computational models.

Briefings in bioinformatics·2017
Same author

Prediction of microbe-disease association from the integration of neighbor and graph with collaborative recommendation model.

Journal of translational medicine·2017

This study introduces W-STFNet, a novel deep learning network for 4D-CT lung image registration. It accurately aligns images with large respiratory motion, improving radiotherapy planning.

Area of Science:

  • Medical Image Analysis
  • Radiotherapy
  • Deep Learning

Background:

  • Deformable Image Registration (DIR) is vital for 4D-CT lung radiotherapy, but existing methods struggle with large respiratory motion and preserving fine structures.
  • Accurate spatiotemporal alignment and deformation plausibility are critical for tasks like dose accumulation.

Purpose of the Study:

  • To develop a robust learning-based method for 4D-CT lung DIR that addresses limitations in modeling global dependencies and preserving anatomical details.
  • To introduce the wavelet transform-based regularized hybrid recursive spatiotemporal fusion registration network (W-STFNet).

Main Methods:

  • W-STFNet utilizes SwinLSTM for global spatiotemporal dependency modeling and a multi-scale spatiotemporal attention fusion (MSTAF) module for feature integration.
Keywords:
4D-CTDeformable image registrationSpatiotemporal feature fusionSwinLSTMWavelet transform

More Related Videos

Four-Dimensional CT Analysis Using Sequential 3D-3D Registration
05:05

Four-Dimensional CT Analysis Using Sequential 3D-3D Registration

Published on: November 23, 2019

8.6K

Related Experiment Videos

Last Updated: Mar 21, 2026

Author Spotlight: An Efficient and Robust Software for Automated Fusion of Multiple Preclinical Imaging Modalities
07:13

Author Spotlight: An Efficient and Robust Software for Automated Fusion of Multiple Preclinical Imaging Modalities

Published on: October 27, 2023

1.8K
Four-Dimensional CT Analysis Using Sequential 3D-3D Registration
05:05

Four-Dimensional CT Analysis Using Sequential 3D-3D Registration

Published on: November 23, 2019

8.6K
  • A novel frequency-domain loss function based on Discrete Wavelet Transform (DWT) is employed to optimize fine-grained structural matching by aligning high-frequency sub-bands.
  • The network is trained in an unsupervised, patient-specific one-shot setting without anatomical annotations or multi-patient pretraining.

    • Main Results:

    • W-STFNet demonstrated competitive registration accuracy and stable performance on DIR-Lab and POPI-model 4D-CT datasets.
    • Achieved mean target registration error (TRE) of 1.13 ± 0.72 mm on DIR-Lab and 0.87 ± 0.56 mm on POPI-model.
    • Statistical tests confirmed W-STFNet's significant improvement over several learning-based baselines.

    Conclusions:

    • W-STFNet offers an annotation-free, patient-specific one-shot registration framework for 4D-CT lung DIR.
    • The method shows robust and competitive performance, especially in handling large deformations and complex temporal dynamics.
    • This approach enhances accuracy for downstream tasks in lung radiotherapy.