Reproducibility of digital image analysis for measuring corneal haze after myopic photorefractive keratectomy

Abstract

Purpose: To evaluate the usefulness of digital image analysis for quantifying corneal haze by determining the reproducibility of its measurements at the corneal plane.

Methods: In a prospective study, 20 randomly selected eyes that had undergone myopic photorefractive keratectomy were photographed focusing the slit beam on their anterior corneal surface. Each photograph was examined using computer image analysis techniques that detect the edge of the reticular pattern of the image. Quantification of the difference between two areas, treated and adjacent untreated cornea, each containing 3,750 pixels with a resolution of 256 gray levels, was performed. Intra-analyzer variation was determined by evaluating the photographs obtained by two analyzers under standard conditions on four separate visits. Interanalyzer variation was calculated using one measurement and the mean of the four measurements.

Results: The pooled standard deviation of the measurements for the analyzers was 0.63 and 0.62 gray levels (coefficient of variation, 4.1% and 3.3%). An association between less severe haze measurements and higher reproducibility scores was found (r = .42; P = .007). The mean interanalyzer variation was smaller for the average of four measurements, 0.55 +/- 0.37 gray levels, than for one measurement, 0.94 +/- 0.73 gray levels (P = .014).

Conclusions: Good reproducibility for haze measurements by digital image analysis of the differences between the treated and adjacent untreated corneal areas was obtained. When the average of four measurements was used instead of a single measurement, interanalyzer reproducibility increased significantly. This new technique may be used to quantify and analyze corneal haze after myopic photorefractive keratectomy.