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

Generating Findings for Jaw Cysts in Dental Panoramic Radiographs Using a GPT-Based VLM: A Preliminary Study on Building a Two-Stage Self-Correction Loop with a Structured Output (SLSO) Framework.

Diagnostics (Basel, Switzerland)·2026
Same author

DermViT: Diagnosis-Guided Vision Transformer for Robust and Efficient Skin Lesion Classification.

Bioengineering (Basel, Switzerland)·2025
Same author

Unsupervised deep learning-based medical image registration: a survey.

Physics in medicine and biology·2024
Same author

Editorial Topical Collection: "Biomedical Imaging and Data Analytics for Disease Diagnosis and Treatment".

Bioengineering (Basel, Switzerland)·2024
Same author

Machine Learning Algorithms for the Diagnosis of Class III Malocclusions in Children.

Children (Basel, Switzerland)·2024
Same author

Classification of Ameloblastoma, Periapical Cyst, and Chronic Suppurative Osteomyelitis with Semi-Supervised Learning: The WaveletFusion-ViT Model Approach.

Bioengineering (Basel, Switzerland)·2024

Related Experiment Video

Updated: May 10, 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

2.6K

Wavelet-Guided Multi-Scale ConvNeXt for Unsupervised Medical Image Registration.

Xuejun Zhang1, Aobo Xu1, Ganxin Ouyang2

  • 1School of Computer, Electronics and Information, Guangxi University, Nanning 530004, China.

Bioengineering (Basel, Switzerland)
|April 26, 2025
PubMed
Summary

WaveMorph, a new wavelet-guided method, enhances unsupervised medical image registration by preserving anatomical details. This approach achieves state-of-the-art accuracy and real-time performance, overcoming limitations of Transformer models.

Keywords:
ConvNeXtHaar waveletnon-rigid medical image registrationunsupervised deep learning

More Related Videos

Multi-modal Pulmonary Imaging: Using Complementary Information from CT and Hyperpolarized 129Xe MRI to Evaluate Lung Structure-Function
02:09

Multi-modal Pulmonary Imaging: Using Complementary Information from CT and Hyperpolarized 129Xe MRI to Evaluate Lung Structure-Function

Published on: April 12, 2024

497
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

319

Related Experiment Videos

Last Updated: May 10, 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

2.6K
Multi-modal Pulmonary Imaging: Using Complementary Information from CT and Hyperpolarized 129Xe MRI to Evaluate Lung Structure-Function
02:09

Multi-modal Pulmonary Imaging: Using Complementary Information from CT and Hyperpolarized 129Xe MRI to Evaluate Lung Structure-Function

Published on: April 12, 2024

497
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

319

Area of Science:

  • Medical Imaging
  • Computer Vision
  • Artificial Intelligence

Background:

  • Medical image registration is crucial for clinical applications like surgical navigation and diagnosis.
  • Transformer models (e.g., TransMorph) show promise but require extensive data and parameters, limiting clinical use.
  • Conventional methods degrade high-frequency anatomical features during downsampling/upsampling.

Purpose of the Study:

  • To develop an efficient and accurate unsupervised medical image registration method.
  • To address the limitations of Transformer models regarding data requirements and spatial locality.
  • To preserve fine-grained anatomical structures often lost in traditional registration techniques.

Main Methods:

  • Proposed WaveMorph, a wavelet-guided multi-scale ConvNeXt method.
  • Introduced a novel multi-scale wavelet feature fusion downsampling module using Haar wavelet decomposition.
  • Implemented a lightweight dynamic upsampling module for reconstructing detailed anatomical structures.

Main Results:

  • WaveMorph achieved state-of-the-art performance on atlas-to-patient (IXI) and inter-patient (OASIS) datasets.
  • Achieved high Dice scores: 0.779 ± 0.015 (IXI) and 0.824 ± 0.021 (OASIS).
  • Demonstrated real-time inference capabilities (0.072 s/image).

Conclusions:

  • WaveMorph effectively integrates CNN inductive bias with Transformer advantages for medical image registration.
  • The model mitigates topological distortions and preserves high-frequency details.
  • WaveMorph offers a viable solution for clinical workflows demanding accuracy and speed.