March 2012
Volume 53, Issue 14
Free
ARVO Annual Meeting Abstract  |   March 2012
Functional Principal Component Analyses of Mouse RPE Sheet Morphology Give Discriminatory Categories
Author Affiliations & Notes
  • J M. Nickerson
    Ophthalmology, Emory Univ, Atlanta, Georgia
  • X Qi
    Mathematics, Georgia State University, Atlanta, Georgia
  • Micah A. Chrenek
    Ophthalmology, Emory Univ, Atlanta, Georgia
  • Christopher Gardner
    Ophthalmology, Emory Univ, Atlanta, Georgia
  • Qing Zhang
    Ophthalmology, Emory Univ, Atlanta, Georgia
  • Hans E. Grossniklaus
    Ophthalmology, Emory Univ, Atlanta, Georgia
  • Yi Jiang
    Mathematics, Georgia State University, Atlanta, Georgia
  • Footnotes
    Commercial Relationships  J. M. Nickerson, None; X. Qi, None; Micah A. Chrenek, None; Christopher Gardner, None; Qing Zhang, None; Hans E. Grossniklaus, None; Yi Jiang, None
  • Footnotes
    Support  Research to Prevent Blindness, Foundation Fighting Blindness, Fight for Sight, and the National Eye Institute (R01EY16470; R24EY017045; P30EY06360; T32EY007092).
Investigative Ophthalmology & Visual Science March 2012, Vol.53, 1602. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      J M. Nickerson, X Qi, Micah A. Chrenek, Christopher Gardner, Qing Zhang, Hans E. Grossniklaus, Yi Jiang; Functional Principal Component Analyses of Mouse RPE Sheet Morphology Give Discriminatory Categories. Invest. Ophthalmol. Vis. Sci. 2012;53(14):1602.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: : To test if RPE morphology changes during and after retinal degeneration as a bystander effect. We compared the morphology of RPE cells from C57BL/6J and congenic rd10 mouse eyes of various age groups. A novel functional principal component analysis (FPCA) technique offers analysis of discrete samples measured at irregular intervals, and allows classification of RPE sheets into major categories.

Methods: : Cell borders of RPE/choroid flatmounts were stained with anti-ZO-1. Photoshop CS2 was used to photomerge the images. Morphometric analysis of cell shape, area, number of neighbors, etc was performed with CellProfiler. Density curves of cell size and aspect ratio were estimated using the penalized likelihood method. FPCA generated principal component (PC) scores, used to construct the classification rule. Three classification methods, LDA (linear discriminant analysis), QDA (quadratic discriminant analysis) and SVM (support vector machine), were applied to the matrix of PC scores. Leave-one-out cross validation was used to assess predictive accuracy.

Results: : Using cell shape (as described by aspect ratio) and cell size (area), FPCA segregates all RPE cell morphology into 4 distinct classes: young C57BL/6J, old C57BL/6J, young rd10, and old rd10. P70 best segregates the age groups for both genotypes into young and old. From 88 eye images, FPCA results in a 88 x 4 matrix of the first four PC scores, used to construct classification rules. The predictive accuracies were 96.6% (85 are correctly classified among 88 mice), 95.5%, and 95.5% for LDA, QDA, and SVM, respectively.

Conclusions: : In young wild type mice the RPE morphology resembles a regular hexagonal array of cells of uniform size. Old wild-type eyes develop a subpopulation of large cells. A clear disruption of the regular cell size and shape appears in rd10 mutants. Aspect ratio and cell area together give rise to discriminatory principal components that classify age and genotype. Here we demonstrated the use of RPE sheet morphometrics as a clear indicator of retinal disease stage despite age as a confound. These same analyses may be applied to patients noninvasively with suitable imaging instruments.

Keywords: retinal pigment epithelium 
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×