Eszter Szalai1, Gábor Németh, András Berta, László Módis. 1. Department of Ophthalmology, Medical and Health Science Center, University of Debrecen, Debrecen, Hajdú-Bihar, Hungary. esztisz@gmail.com
Abstract
PURPOSE: To determine the difference between the endothelial cell density (ECD) and the morphologic measurements assessed with a recently developed noncontact specular microscope compared with a contact endothelial microscope. The agreement between the 2 independent operators was also investigated. METHODS: The central corneal endothelium was examined in the right eyes of 41 healthy individuals using automated image analysis programs without correction for a noncontact specular microscope and a contact specular microscope (EM-3000, EM-1000; Tomey, Tennenlohe, Germany). The evaluated parameters such as ECD,average cell area (AVG), and coefficient of variation of the cell area were determined by 2 operators. Conversion factors were also considered to adjust the ECD values. RESULTS: ECD measurements obtained with the noncontact instrument (2734 ± 287 cells per square millimeter, first operator; 2726 ± 282 cells per square millimeter, second operator) were significantly higher (P < 0.0001) in comparison with the contact microscope (2283 ± 206 cells per square millimeter, first operator; 2274 ± 242 cells per square millimeter, second operator) by both investigators. The contact specular microscope significantly overestimated (P < 0.0001) AVG and coefficient of variation of the cell area values. Statistically significant correlation (P < 0.0001) was detected in ECD (r = 0.77, first operator) and AVG measurements (r = 0.72, first operator) between the 2 endothelial microscopes. Higher degree of agreement between operators was given when using the noncontact specular microscope compared with the contact instrument. CONCLUSIONS: The significant differences and the poor agreement between the automated software options disclosed that the 2 instruments cannot be used interchangeably.
PURPOSE: To determine the difference between the endothelial cell density (ECD) and the morphologic measurements assessed with a recently developed noncontact specular microscope compared with a contact endothelial microscope. The agreement between the 2 independent operators was also investigated. METHODS: The central corneal endothelium was examined in the right eyes of 41 healthy individuals using automated image analysis programs without correction for a noncontact specular microscope and a contact specular microscope (EM-3000, EM-1000; Tomey, Tennenlohe, Germany). The evaluated parameters such as ECD,average cell area (AVG), and coefficient of variation of the cell area were determined by 2 operators. Conversion factors were also considered to adjust the ECD values. RESULTS: ECD measurements obtained with the noncontact instrument (2734 ± 287 cells per square millimeter, first operator; 2726 ± 282 cells per square millimeter, second operator) were significantly higher (P < 0.0001) in comparison with the contact microscope (2283 ± 206 cells per square millimeter, first operator; 2274 ± 242 cells per square millimeter, second operator) by both investigators. The contact specular microscope significantly overestimated (P < 0.0001) AVG and coefficient of variation of the cell area values. Statistically significant correlation (P < 0.0001) was detected in ECD (r = 0.77, first operator) and AVG measurements (r = 0.72, first operator) between the 2 endothelial microscopes. Higher degree of agreement between operators was given when using the noncontact specular microscope compared with the contact instrument. CONCLUSIONS: The significant differences and the poor agreement between the automated software options disclosed that the 2 instruments cannot be used interchangeably.