Accuracy and precision of videokeratoscopes for test surfaces

Abstract

The present study evaluated the accuracy and precision performance
of three videokeratoscopes (the Keratron, TMS-1 and PAR
systems) in elevation topography with six test surfaces that were
designed to challenge each instrument’s ability to measure them.
The test surfaces comprised a sphere (r = 7.8 mm), an asphere
(r = 7.8 mm, Q = - 0.49), a multicurve (r1 = 7.84 mm, r2 = 7.54,
r3 = 7.84, r4 = 7.54, r5 = 7.84), a 5.0 bicurve (r1 = 5.0 mm,
r2 = 7.80), a 6.5 bicurve (r1= 6.5 mm, r2 = 7.80) and an 8.5 bicurve
(r1= 8.5 mm, r2= 7.80, sag = 50 μm).
For the spherical test surface, the root mean square error
(RMSE) data showed the instrument accuracies for the Keratron,
TMS-1 and PAR systems were 1.64, 2.61 and 20.89 μm, respectively.
The standard error for the Keratron system was small
(< ± 1 μm) for the same test surface, compared to the standard
errors of ± 3 and ± 44 μm for the TMS-1 and PAR systems.
Overall, each instrument performed well on certain test surfaces
but none of the instruments excelled on all the surfaces.
The study has confirmed the high accuracy of the Keratron system
in measuring test surfaces such as the sphere, the asphere
and the multicurve. The precision of the instrument was the
best among the three instruments. The TMS-1 system was slightly
less accurate and precise than the Keratron system in elevation
topography for the sphere and the asphere surfaces. Of the three
instruments, the PAR system has the poorest performance in
accuracy and precision.