A proposed regression analysis-based method for assessment of higher order aberration interrelationships in ocular...

To analyze the interrelations and difference in higher order aberration patterns in ocular wavefront sets with similar measures of central tendency and apparently similar origins.

This theoretical simulation and assessment study was performed at the cornea and refractive surgery services of a tertiary care, teaching hospital. A sample data set of myopic candidates was used to create distribution patterns of signed and absolute values. Histograms and quantile-to-quantile (Q-Q) plots were used to analyze the pattern. Regression models were used for interocular comparison between fellow eyes without enantiomeric considerations, to create a subtle difference in interrelation, but not in the overall mean of absolute values. The values were represented in the form of R-squared values and slope of the regression line.

The Q-Q plots suggested large deviations from normal values in analyzing the absolute values. Regression plots were used to find out the Zernike modes that predicted other modes, and the ones with the highest R-square values were compared between data sets. This revealed differences in the interrelationship of various Zernike modes from each other, which could have been missed in measures of central tendency.

The proposed regression analysis-based model could pick up subtle differences in the higher order aberration patterns and interrelation, which would have been ignored in measures of central tendency. Large multicentric pooled data could be used and stratified according to age and racial variations to establish the limits of normal ranges based on this model.