Abstract
Purpose::
To determine the reliability of central corneal thickness measurements acquired by in-vivo contact laser confocal microscopy compared with conventional ultrasonic measurements.
Methods::
Central corneal thickness was measured by two different methods--laser confocal microscopy using the Rostock Corneal Module (Heidelberg Engineering, Vista CA) and conventional ultrasound (DGH-550 Pachette 2, Exeter, PA). Both eyes of five New Zealand white rabbits were measured. The 95% limits of agreement were computed and compared using methods described by Bland and Altman.
Results::
Central corneal thickness measured by confocal microscopy was (mean ± SD) 443.5 ± 27.30 µm. Ultrasonic corneal thickness measurements were 443.2 ± 34.85 µm. These were not significantly different (paired t-test, p = .94). The mean difference between these two measurement methods was 0.3 µm. The 95% limits of agreement ranged from -24.74 to +25.38 µm.
Conclusions::
Our results demonstrate excellent agreement between these two methods of measuring central corneal thickness. A total of 6 of the 10 of the confocal microscopy measurements were within 10 µm of the ultrasonic measurements, a value previously reported as the reliability of ultrasonic measurements. A total of 9 of 10 measurements were within 15 µm. The mean difference between these two measurement methods was not significantly different from zero (0.3 µm), indicating that there was no systematic bias by method. Our results show that the 95% limits of agreement by laser confocal microscopy are low and less than what have been previously reported for other confocal microscopy methods using similar techniques.
Keywords: cornea: clinical science • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)