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

Normalization procedures and factorial representations for classification of correlation-aligned images: a

M Unser1, B L Trus, A C Steven

  • 1Biomedical Engineering and Instrumentation Branch, National Institutes of Health, Bethesda, Maryland 20892.

Ultramicroscopy
|July 1, 1989
PubMed
Summary

Optimizing multivariate statistical analysis (MSA) for electron micrograph classification requires careful normalization and factorial representation. Constant mean and variance (CMV) normalization with correspondence analysis (CA) or principal components (PC) offers better discrimination of biological macromolecule images.

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

Activated STING in a vascular and pulmonary syndrome.

The New England journal of medicine·2014
Same author

Autonomous and self-sustained circadian oscillators displayed in human islet cells.

Diabetologia·2012
Same author

An Optimized Spline-Based Registration of a 3D CT to a Set of C-Arm Images.

International journal of biomedical imaging·2012
Same author

3-D PSF fitting for fluorescence microscopy: implementation and localization application.

Journal of microscopy·2012
Same author

Realistic analytical phantoms for parallel magnetic resonance imaging.

IEEE transactions on medical imaging·2011
Same author

Sum and difference histograms for texture classification.

IEEE transactions on pattern analysis and machine intelligence·2011

Area of Science:

  • Structural biology
  • Biophysics
  • Computational biology

Background:

  • Accurate classification of biological macromolecule images is crucial for structural determination.
  • Multivariate statistical analysis (MSA) is employed for image classification but requires optimization.
  • Normalization and factorial representation choices impact classification accuracy.

Purpose of the Study:

  • To optimize multivariate statistical analysis (MSA) procedures for classifying electron micrographs.
  • To compare the effects of different normalization methods and factorial representations on image discrimination.
  • To assess the resolution of distinct image classes in factorial space.

Main Methods:

  • Analysis of pre-aligned electron micrographs of two known protein classes.

Related Experiment Videos

  • Comparison of constant minimum and maximum (CMM) versus constant mean and variance (CMV) normalization.
  • Evaluation of correspondence analysis (CA) and principal components (PC) formalisms.
  • Quantitative assessment of inter-set discrimination.

    • Main Results:

    • Constant mean and variance (CMV) normalization proved superior to constant minimum and maximum (CMM) normalization.
    • CMV normalization with either CA or PC yielded qualitatively similar results.
    • Principal components (PC) showed slightly better quantitative discrimination than CA when using CMV normalization.
    • Spurious fluctuations were more pronounced with PC and CMM normalization.

    Conclusions:

    • CMV normalization is a more satisfactory method for analyzing electron micrograph data than CMM.
    • Both CA and PC are viable factorial representations when combined with CMV normalization.
    • Full resolution of image classes in factorial space remains a challenge, indicating potential ambiguities in practical image classification.