Identification of eyes at risk of developing idiopathic macular holes by support vector machines

Abstract

This work aims to discriminate between healthy eyes and eyes at risk of developing idiopathic macular hole (IMH). Fits of well known mathematical functions were used to model the topography of the retina with special emphasis on the foveal depression. Based on this set of fits, we are able to describe and, therefore, to analyze the shape of the retinal surface. The working hypothesis is that differences can be found within the parameters of the set of functions used to describe the retinal topography between the two groups of eyes. We have resorted to a pattern classification support vector machine algorithm to discriminate between groups through training using known cases.