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 Videos

The registration of MR images using multiscale robust methods

M E Alexander1, R L Somorjai

  • 1National Research Council Canada, Institute for Biodiagnostics, Winnipeg, Canada. alexander@ibd.nrc.ca

Magnetic Resonance Imaging
|January 1, 1996
PubMed
Summary

This study presents a robust global image registration method for medical imaging, effective even with noise and distortions. The novel approach achieves subpixel accuracy, outperforming existing algorithms in challenging scenarios.

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

Rapid learning in a native predator shifts diet preferences towards invasive prey.

Biology letters·2022
Same author

Out of the pot and into the fire: Explaining the vulnerability of an endangered small headwater stream fish to black-bass Micropterus spp. invasion.

Journal of fish biology·2018
Same author

Creating robust, reliable, clinically relevant classifiers from spectroscopic data.

Biophysical reviews·2017
Same author

Feasibility of infrared spectroscopy with pattern recognition techniques to identify a subpopulation of mares at risk of producing foals diagnosed with failure of transfer of passive immunity.

Australian veterinary journal·2012
Same author

Class proximity measures--dissimilarity-based classification and display of high-dimensional data.

Journal of biomedical informatics·2011
Same author

Disease pattern recognition in infrared spectra of human sera with diabetes mellitus as an example.

Applied optics·2008

Area of Science:

  • Medical Imaging
  • Image Processing
  • Computational Anatomy

Background:

  • Accurate registration of MR images is crucial for reliable interpretation and analysis.
  • Motion artifacts and misregistration can significantly degrade image quality and lead to flawed conclusions.
  • Existing registration methods may struggle with noise and local distortions.

Purpose of the Study:

  • To introduce a novel global image registration method robust to noise and local distortions.
  • To evaluate the performance and accuracy of the proposed method compared to existing algorithms.
  • To demonstrate the method's ability to handle various transformations and noise levels.

Main Methods:

  • A two-stage registration approach involving optional Fourier phase-matching for initial alignment.

Related Experiment Videos

  • An iterative procedure employing robust nonlinear regression at prescribed registration points.
  • Accommodation of general linear and nonlinear registration transformations.

    • Main Results:

    • Subpixel accuracy achieved in noise-free conditions for scaled, rotated, and translated brain images.
    • Graceful degradation of accuracy with increasing levels of Gaussian noise.
    • Recovery errors under 3 pixels with 40% noise for small shifts, outperforming the Woods algorithm.

    Conclusions:

    • The proposed global registration method offers high accuracy and robustness in the presence of noise and distortions.
    • The algorithm demonstrates superior performance and convergence compared to the Woods algorithm in challenging registration tasks.
    • This method provides a reliable tool for accurate medical image registration, enhancing diagnostic and analytical capabilities.