April 2014
Volume 55, Issue 13
ARVO Annual Meeting Abstract  |   April 2014
Factors influencing estimation of cone packing density in adaptive optics retinal images
Author Affiliations & Notes
  • Marco Lombardo
    Fondazione GB Bietti IRCCS, Rome, Italy
  • Sebastiano Serrao
    Fondazione GB Bietti IRCCS, Rome, Italy
  • Giuseppe Lombardo
    IPCF, CNR, Rende, Italy
  • Footnotes
    Commercial Relationships Marco Lombardo, None; Sebastiano Serrao, None; Giuseppe Lombardo, None
  • Footnotes
    Support None
Investigative Ophthalmology & Visual Science April 2014, Vol.55, 5198. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Marco Lombardo, Sebastiano Serrao, Giuseppe Lombardo; Factors influencing estimation of cone packing density in adaptive optics retinal images. Invest. Ophthalmol. Vis. Sci. 2014;55(13):5198.

      Download citation file:

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

  • Supplements

To evaluate the influence of demographic and technical factors on estimates of cone packing density in a healthy population of adults.


An adaptive optics (AO) retinal camera was used to obtain images of the parafoveal cone mosaic in fifteen healthy volunteers, aged 21-46 years old. Cone density and packing arrangement were estimated in windows of 320x320 µm, 160x160 µm and 64x64 µm at corresponding radial and linear distances from the fovea, i.e., 1.5 degree and 418 µm temporal and superior eccentricity. Moreover, densities were calculated either with or without compensating by the corrected retinal magnification factor (RMFcorr). The analysis of variance intraclass correlation coefficient (ICC) and Bland-Altman analysis were used to test significance and agreement between cone density measurements and preferred packing arrangements of cones taken within the different window comditions. A multiple linear regression analysis was used to analyze the effect of age and RMFcorr on cone density calculated within sampling windows of different size.


Cone density declined with aging (r≥-0.44; P<0.05). Moreover, density tended to decline with decreasing sampling area (320x320 µm and 64x64 µm areas had differences of 2100 cones/mm2; P=0.07). Although the differences in cone density between the windows of same size centered at radial and linear distance from the fovea were not statistically significant, ICC and Bland-Altman analysis showed that these estimates could not be directly compared. The use of individual RMFcorr did not influence density variation. The percentage of Voronoi tiles with hexagonal packing arrangement could be compared between sampling areas of different size. Sampling areas of 64x64 µm size or smaller are more vulnerable to limits of Voronoi diagrams, such as defects in the image and boundary conditions.


The strict use of a RMFcorr for each eye was shown not to be as important as other technical factors, such as window size and retinal eccentricity expressed in radial or linear distance, in order to minimize the proportional error when comparing cone densities between different studies. Age was the most important demographic factor to influence density variation. The graphical representation of preferred packing arrangements of cones by Voronoi tiles was slightly affected by window size.

Keywords: 550 imaging/image analysis: clinical • 648 photoreceptors  

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.