Analysis of the Accuracy of Reconstruction of a Human Cornea by Two Geometric Modelling Techniques: A Comparative Study

This study compares two techniques of reconstruction and representation of the anterior and posterior surfaces of the human cornea, both for healthy and for keratoconus eyes, in a CAD software environment using raw tomographic data, by two different methods: a mesh model (obtained from a grid of points) and a NURBS surface model. After reconstruction, a morpho-geometric analysis was made, and several parameters were defined and measured in a set of 100 healthy eyes and 61 keratoconus eyes, detecting the statistical analysis significant differences between methods for both groups. Mesh method proved to be more accurate, but less eyes could be modelled. Volumetric and surface measurement values showed a total equivalence between both methods, so these would be the ones to use when modelling eyes for keratoconus detection and characterization purposes.