June 2015
Volume 56, Issue 7
ARVO Annual Meeting Abstract  |   June 2015
Customizing Macular Structure-Function Maps for Individuals
Author Affiliations & Notes
  • Andrew Turpin
    Computing and Information Systems, University of Melbourne, Parkville, VIC, Australia
  • Siyuan Chen
    Computing and Information Systems, University of Melbourne, Parkville, VIC, Australia
  • Juan Alejandro Sepulveda Ulloa
    Optometry and Vision Sciences, The University of Melbourne, Melbourne, VIC, Australia
  • Allison M McKendrick
    Optometry and Vision Sciences, The University of Melbourne, Melbourne, VIC, Australia
  • Footnotes
    Commercial Relationships Andrew Turpin, Centervue Inc (R), Haag-Streit AG (C), Heidelberg Engineering GmBH (F); Siyuan Chen, None; Juan Sepulveda Ulloa, None; Allison McKendrick, Carl Zeiss Meditech (R), Haag-Streit AG (C), Heidelberg Engineering GmBH (F)
  • Footnotes
    Support None
Investigative Ophthalmology & Visual Science June 2015, Vol.56, 632. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Andrew Turpin, Siyuan Chen, Juan Alejandro Sepulveda Ulloa, Allison M McKendrick; Customizing Macular Structure-Function Maps for Individuals. Invest. Ophthalmol. Vis. Sci. 2015;56(7 ):632.

      Download citation file:

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

  • Supplements

Purpose: To develop a model of Henle fibre length that is readily customizable to an individual based on their OCT images. By adjusting for an individual's Henle fibre length, it is anticipated that structural and functional measures of the macular will better correlate in glaucoma patients.

Methods: We begin with the model of Drasdo et al (Vision Research 47, 2007) as a basis, which uses population-based estimates of the density of retinal ganglion cells (RGCs) and the density of 'inputs' to those ganglion cells (RFs) in the macular region, combined with a simple counting argument, to predict Henle fiber length. In order to customize the computations to an individual, we adjust the density of both RGC and RF in proportion to the ratio of the appropriate layers in a spectral domain OCT image for a person to the population average. For RGCs we use the GCL+ layer (IPL less RNFL using the segmentation scheme of the Spectralis OCT, Heidelberg Engineering), and for the RF we use the Outer Nuclear Layer (ONL) (OPL less ELM).<br /> <br /> For an individual, the OCT data was collected with high resolution radial line scans at 45 degree intervals, and both the GCL+ and ONL thickness was fitted and interpolated to the whole retina with the model of Scheibe et al (Experimental Eye Research 119, 2014) using our own software. For examining the structure-function concordance, we used the RNFL layer of a cube scan of the macular.<br /> <br /> Individual Henle fiber-length maps for 30 control eyes were derived. To explore the utility of the model for structure-function computations in the macular, we fitted the model to the normal area (HFA 10-2 TD within normal limits) of 6 glaucomatous patients with known 10-2 hemifield-loss, extrapolated this to their damaged area, and examined the structure-function relationship both with and without Henle fiber correction.

Results: In the normal eyes, the location of some 10-2 test points when projected onto the retina shifted by more than one degree, and the range of shift between individuals was over one degree. For the glaucomatous eyes, the RNFL thickness of locations that were damaged (10-2 TD below normal limits) generally decreased when mapped to areas of the retina according to our model of Henle fibre length.

Conclusions: Using segmented OCT images to customize macular structure-function mapping for individuals is possible, and may lead to better quantification of macular damage in individuals with glaucoma.


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.