Purpose
Forme fruste keratoconus (FFK) can be defined as a cornea that has no abnormal findings by both slit-lamp examinations and Placido-based corneal topography, with the fellow eye of clinical keratoconus. It is considered as the subjects with which one can create the better screening programs for detecting the early form of keratoconus. The purpose of this study is to conduct multiple logistic regression for differentiate FFK from normal eyes.
Methods
The subjects includes 38 eyes of forme fruste keratoconus (FFK) and 48 eyes of normal cornea. The corneal tomographic analysis was performed using a rotating Scheimpflug camera (Pentacam HR, OCULUS Optikgeräte GmbH, Germany). The ocular and corneal higher-order aberrations were obtained by an aberrometer (KR-1W, Topcon Corp., Japan). The parameters of corneal biomechanics were detected with s dynamic Scheimpflug analyzer (Corvis ST, Oculus Optikgeräte, Inc., Wetzler, Germany). Multiple logistic regression analyses and the receiver operating characteristic (ROC) curve analyses were performed on these data to develop the screening system of FFK from normal subjects.
Results
Multiple logistic regression analyses (forward stepwise method) showed the coma aberration of total cornea, coma aberration of anterior cornea, pachymetry, and Highest Concavity time were significant variables for differentiating FFK group from the normal group (P<0.001). The discriminant functions between FFK and normal cornea group reached an area under the ROC curve of 0.98. The sensitivity and the specificity were 100% and 98%, respectively.
Conclusions
Combining parameters from the corneal tomography, aberration, and biomechanics may be useful to improve the screening ability of FFK from the normal eyes.
Keywords: 479 cornea: clinical science •
574 keratoconus •
496 detection