Neural network-based system for early keratoconus detection from corneal topography

Abstract

Some automatic methods have been proposed to identify keratoconus from corneal maps; among these methods, neural networks have proved to be useful. However, the identification of the early cases of this ocular disease remains a problem from both a diagnostic and a screening point of view. Another problem is whether a keratoconus screening must be performed taking into account both eyes of the same subject or each eye separately; hitherto, neural networks have only been used in the second alternative. In order to examine the differences of the two screening alternatives in terms of discriminative capability, several combinations of the number of input, hidden and output nodes and of learning rates have been examined in this study. The best results have been achieved by using as input the parameters of both eyes of the same subject and as output the three categories of clinical classification (normal, keratoconus, other alterations) for each subject, a low number of neurons in the hidden layer (lower than 10) and a learning rate of 0.1. In this case a global sensitivity of 94.1% (with a keratoconus sensitivity of 100%) in the test set as well as a global specificity of 97.6% (98.6% for keratoconus alone) have been reached.

Anterior Segment Characteristics of Keratoconus Eyes Using Scheimpflug-Placido Topography

“…28 Difficulty recognizing keratoconus arises with very early or preclinical stages of the ectatic disorder [24,[29][30][31].…”

Characterization of corneal structure in keratoconus

“…1 Several indices, algorithms, and even neural network approaches to the geometry and optical properties of the anterior corneal surface have been developed for keratoconus diagnosis and detection. [3][4][5][6][7][8][9][10][11][12][13] Therefore, moderate and advanced keratoconus detection is not a difficult issue using corneal topography in combination with biomicroscopic, retinoscopic, and pachymetric evaluation. 1 Detection problems arise with very early or preclinical stages of the ectatic disorder.…”

“…7,8,[59][60][61][62][63] The digital analysis of the image of the Placido disks projected on the cornea has also been shown to be a valid tool for keratoconus diagnosis, avoiding the use of estimated data or parameters obtained indirectly. 64 Specifically, the use of primary corneal indices characterizing the asymmetry of the mires seems to be especially useful for keratoconus detection.…”

Automated Keratoconus Detection Using Height Data of Anterior and Posterior Corneal Surfaces

“…11,12,16,20,29,31 The patients with WTR astigmatism were those with regular astigmatism of more than −1.5 D. The patients in the keratoconus group were separated into two groups: those with KC and those who were KCS. The nonkeratoconus group was made up of WTR astigmatic subjects, post-PRK subjects, and controls.…”

“…At present, several quantitative corneal indices have been developed for computed videokeratoscopic systems, and several automated keratoconusdetection systems using these indices are available. 6,[8][9][10][11][12][13][14][15][16][17][18][19] These systems are of great help because of they are easy to operate and make a rapid diagnosis possible during the preoperative screening for keratoconus.…”