Abstract
Purpose: :
We present a new method to characterize the spatial statistics of the ocular aberrations and show how to apply it to a sample of healthy eyes. In particular, we assess the compatibility of simple eye aberration statistical models with the experimental centroid displacements provided by a Hartmann-Shack wavefront sensor (HS).
Methods: :
Initial information about the aberrations statistics is obtained from the centroid diagrams (defined as the plot of the positions of the centroids of the HS lenslet array for all the series measured for a given eye). We then compute the theoretical values of the second-order statistical functions of the sensor centroid displacements for two statistical models of aberration statistics: (1) defocus driven, and (2) power-law (Kolmogorov-like). Using a chi-square test, we check the compatibility of these predictions with the estimates of the same magnitudes from the experimental centroids: Two sets of uncorrected eyes under normal accommodation conditions were used: (a) 41 eyes from a middle-aged group (45 to 65 years) of 21 people and (b) 12 eyes from 6 people below 45 years. HS data were collected at 24 Hz using a laser diode at 780 nm.
Results: :
The eyes of the young-aged group that had a strong defocus fluctuation showed a characteristic inhomogeneous behavior, as their centroid diagrams and the plots of their centroid variances revealed. This is incompatible with the homogeneity of power-law models. The chi-square test performed for the centroid displacements structure function also dismissed power-laws (at a 5% significance level). For those eyes with weaker defocus fluctuation, the high variability of the data made this latter test not significant for both defocus and power-laws. The statistics observed for the middle-aged group showed a higher degree of homogeneity and isotropy; power-law models could not be ruled out as candidates for the description of these eyes.
Conclusions: :
The method presented skips the aberrations estimation stage and works with HS measurements instead. This procedure simplifies the comparison of results between labs (the specification of the reconstruction parameters is not needed) and avoids any loss of information due to the estimation step.The results obtained stress the importance of the defocus fluctuations in young eyes. Forthcoming models of the individual eye aberrations must take into account the high weight of this term in relation to that contained in power-law models.