Measurement, modeling and visualization
of the shape of the cornea

F.M. Vos *,+, H.J.W. Spoelder *, I.H.M. van Stokkum *,
G.L van der Heijde +, F.C.A. Groen #

*: Vrije Universiteit Amsterdam
Department of Physics and Astronomy,
De Boelelaan 1081, 1081 HV Amsterdam, The Netherlands

+: Faculty of Medicine of the Vrije Universiteit
P.O. Box 7057, 1007 MB Amsterdam, The Netherlands

#: Faculty of Mathematics, Computer Science, Physics and Astronomy
of the University of Amsterdam
Kruislaan 403, 1098 SJ Amsterdam, The Netherlands
Abstract

We have developed a prototype instrument to determine the shape of the cornea. The measurement technique is based on the reflection of a colour coded, Pseudo Random Binary Array (PRBA), stimulus by the eye. The special properties of the PRBA enable unique characterization of positions both in the stimulus and in the reflected image. This allows for a sophisticated surface reconstruction.

The shape of the cornea is described by a set of tensor splines. Starting out from the reflected image we are, assuming a shape for the cornea, capable to calculate the associated stimulus pattern (backward raytracing). The best approximation of the corneal shape is determined by minimizing the difference between calculated and actual stimulus.

To allow interpretation of the data at the various stages visualization is necessary. The processing of the data falls into two steps:

Visual representation of the results at all intermediate stages is necessary both in the development of the prototype instrument and for the users of the instrument. We have chosen IBM's Data Explorer as platform to develop the user interface and the visualization routines.

This paper is divided into the following sections:

  1. Introduction
  2. Image processing
    1. Crossing Detection
    2. Graph Reconstruction
  3. Surface Reconstruction
    1. Introduction
    2. General Description
    3. Spline Theory
    4. Results
  4. Conclusions and future work
  5. References