Abstract
Abstract: :
Purpose:To test the level, repeatability, and visual significance of aberrations that are not adequately reproduced by Zernike polynomials Methods: Wavefronts of 20 eyes were measured three times each with a VISX WaveScan® system and analyzed using custom software that allows wavefronts to be reconstructed by two methods. One method allows an arbitrary number of Zernike coefficients to be used in the reconstruction to give the maximum amount of detail. The second method allows Fourier wavefront reconstruction at the full resolution of the Hartmann–Shack sensor. The results were analyzed to isolate any high order aberrations that were missed by the Zernike reconstructions. The two dimensional Pearson's product moment correlation was used to compare the very high order aberrations missed by the Zernike reconstruction to the other measurements of the same eyes. To measure the level of non–Zernike aberrations in a larger population, 108 wavefront exams were analyzed to determine residual RMS wavefront error. The visual significance of these unmeasured aberrations was estimated by calculating the Strehl ratio of each case. Results: When reconstructed with 6th order Zernikes, the residual (non–Zernike) aberrations were highly correlated from exam to exam (average r of 0.80). Of 108 exams, the average RMS wavefront error of the non–Zernike aberrations was: 4th order, 0.09 µm; 5th order, 0.07µm; and 6th order, 0.05µm. Strehl ratios of the residual error were: 4th order, 0.38; 5th order, 0.50; and 6th order, 0.63. Conclusions: In normal eyes, small but visually significant high order aberrations, although very consistent across successive measurements, are missed by wavefront reconstruction using Zernike polynomials.
Keywords: refractive surgery: optical quality • computational modeling • cornea: clinical science