Abstract
Purpose :
To describe the corneal wavefront error with principal components obtained fronm corneal Zernike coefficients and to compare the ability of the novel wave error representation with other wavefront error representations.
Methods :
From 792 normal eyes of 495 patients, corneal tomography scans were taken with a commercial Scheimpflug system (Pentacam HR, Oculus, Germany). Based on a ray tracing model, total corneal wavefront aberrations were calculated using a Zernike decomposition up to the 6th order over a pupil diameter of 6 mm. From 27 Zernike coefficients, a principal component analysis (PCA) based on the correlation matrix was performed (SPSS 11.0, Varimax rotation). Coefficient loads of less than |0.25| were ignored. For component selection, an eigenvalue of >1 was applied. Wavefront errors were built up in a stepwise fashion using the novel components, of which each contained a subset of Zernike coefficients. Similarly, wavefront errors were described with single Zernike coefficients by the root-mean square (RMS) of Zernike orders 3-6 and by the RMS of all coma, spherical and residual aberrations. For each wavefront error, the optical quality metric BCVSOTF (visual Strehl ratio based on the optical transfer function, simulated for best spectacle correction) was computed (VOL-Pro 7.14, Sarver and Ass.). The number of components to explain 95% of the variance of BCVSOTF was compared.
Results :
PCA produced 11 components with an eigenvalue >1, accounting for 72% of the total variance. The first 4 components accounted for 95% of the BCVSOTF variance. Using individual Zernike modes, 9 coefficients were necessary to describe 95% of the BCVSOTF variance. For wavefront description with RMS values three components were needed each to describe at least 95% of BCVSOTF variance.
Conclusions :
Novel wavefront components obtained by PCA were able to describe corneal optical properties as comprehensively as coarser representations such as RMS of Zernike orders.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.