September 2016
Volume 57, Issue 12
Open Access
ARVO Annual Meeting Abstract  |   September 2016
Accurate limbus demarcation using 3D anterior eye height data
Author Affiliations & Notes
  • Alejandra Consejo
    Biomedical Engineering, WROCLAW UNIVERSITY OF TECHNOLOGY, WROCLAW, Poland
  • D Robert Iskander
    Biomedical Engineering, WROCLAW UNIVERSITY OF TECHNOLOGY, WROCLAW, Poland
  • Footnotes
    Commercial Relationships   Alejandra Consejo, None; D Robert Iskander, Eaglet Eye (F)
  • Footnotes
    Support  Marie Curie ITN Grant, AGEYE, 608049
Investigative Ophthalmology & Visual Science September 2016, Vol.57, 5932. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Alejandra Consejo, D Robert Iskander; Accurate limbus demarcation using 3D anterior eye height data. Invest. Ophthalmol. Vis. Sci. 2016;57(12):5932.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : To develop a method for locating the transition from cornea to sclera using height data from a profilometry measurement of an anterior eye surface.

Methods : Data of a simulated example, and collected from an artificial bi-sphere test surface and four real eyes is used to validate the method. Data was acquired using Eye Surface Profiler (ESP, Eaglet Eye BV, Netherlands), a profilometer with the potential of measuring the corneo-scleral topography up to 20 mm diameter far beyond the limbus. 12 measurements were acquired from the right eye of each of the subjects. Raw 3D profilometry data was fitted with the 2nd radial order Zernike polynomial expansion. Fig. 1 (left) shows the difference between original and fitted data for a single semi-meridian. The residual error between the original surface and the fitted one is calculated. Then, the cumulative root mean square (RMS) of the residual surface for each semi-meridian is calculated. Further, the maximum change in the cumulative RMS in the region of interest (5<R<7 mm) is sought and assigned as limbus. Finally, a best-fit-circle is estimated using the points which demarcate the limbus for each semi-meridian. The radius of this circle is assigned as limbal radius.

Results : Determining the limbus position is possible using this technique. The proposed method is applied successfully on the simulated test surface (0.03% relative error) and the artificial bi-sphere (<2% relative error). For real eyes the standard deviation was smaller or equal than 10 µm for all four considered subjects. Fig 1 (right) shows an example of the limbal demarcation method applied to one of the subjects.

Conclusions : A method of limbus demarcation has been proposed. It is based on calculating the cumulative RMS of a residual error between the original data and low order Zernike polynomial expansion of the anterior eye topography, including cornea, sclera and the corneo-scleral region. The method achieves a small relative error compared to its predecessors based on, for example, fitting a bi-sphere surface to such data.

This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.

 

Fig 1. Original data (blue) and fitted data (red) in a semi-meridian of the artificial bi-sphere. Vertical dashed lines indicate the region of interest. The arrows show the biggest change (maximum error) between lines (left).Example of the estimated limbus position indicated by small overlapping circles and fitting circle as limbal radius estimator (right).

Fig 1. Original data (blue) and fitted data (red) in a semi-meridian of the artificial bi-sphere. Vertical dashed lines indicate the region of interest. The arrows show the biggest change (maximum error) between lines (left).Example of the estimated limbus position indicated by small overlapping circles and fitting circle as limbal radius estimator (right).

×
×

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.

×