The accuracy of 'power' maps to display curvature data in corneal topography systems.

PURPOSE To quantify the error introduced by videokeratographic corneal topography devices in using a paraxial formula to calculate power over the entire corneal surface, including areas removed from the central paraxial region where the formula is known to be invalid. METHODS Corneal refractive power and two paraxial power approximations were computed as a function of distance from the apex for three theoretical surfaces, a sphere and two ellipsoids with 0.3 and 0.5 eccentricities. Color dioptric maps were then theoretically created. RESULTS For the spherical surface, both curvature-based paraxial power approximations were uniform over the entire surface because curvature is constant. However, the corneal refractive power increased from center to periphery, demonstrating the known phenomenon of spherical aberration. For the ellipsoids, which have been shown to model the human cornea, curvature-based power approximations decreased from center to periphery because the curvature flattens peripherally. However, refractive power increased from center to periphery. The limits of the central paraxial region for these surfaces was shown to be approximately 2 mm in diameter for the paraxial power approximation used by videokeratographic devices to measure 8 mm in diameter. CONCLUSIONS The direct correlation between corneal curvature and power with which clinicians are familiar is not valid in the peripheral regions measured by videokeratographic devices. Topographic devices measure curvature, which should not be interpreted as corneal power except in the central region. A recommendation to device manufacturers is to display "color curvature maps" instead of color dioptric maps, and to label the color bar with curvature values instead of power.