Abstract
Purpose: :
To detect keratoconus on corneal pachymetry maps.
Methods: :
A retrospective analysis was performed on visits from July 2007 to August 2008 at the Doheny Laser Vision Center. Fifteen eyes diagnosed with keratoconus were included in the study. The most recent 15 pre-LASIK normal subjects were used as control. All subjects were scanned using a high-speed Fourier-domain optical coherence tomography (FD-OCT) system (RTVue, Optovue Inc.) with a corneal adaptor module (CAM). The pachymetry mapping pattern was centered on the pupil and consisted of 6-mm line scans on 8 meridians. Corneal pachymetry maps were obtained by interpolating between meridians. The depth and width of focal thinning (cones) were quantified by fitting a rotationally symmetric two-dimensional Gaussian waveform to the inverse normalized pachymetry maps (INPM). The INPM is the average pachymetry map of normal eyes divided by the pachymetry map of the eye under examination - thus a localized thinning appears as a distinct peak. The Gaussian waveform was centered at the peak INPM value and the width was adjusted to maximize the cross-correlation coefficient. The cross-product of the best-fit Gaussian waveform and the INPM represents the cone amplitude in units of fractional focal thinning.
Results: :
The Gaussian function provided a good model of keratoconus, with an average cross-correlation of 0.86 with the INPM. The average cone amplitude was 0.30 ± 0.20 (unit: fractional thinning; range: 0.089 to 0.73) in keratoconic eyes and 0.036 ± 0.012 (range 0.019 to 0.057) in normal eyes. There was no overlap in cone amplitude between the 2 groups and the difference was highly statistical significant by t-test (P < 0.0001). The average cone diameter was 4.93 ± 1.64 mm in keratoconic eyes and 7.04 ± 2.85 mm in normal eyes (P = 0.019).
Keywords: keratoconus • imaging/image analysis: clinical • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)