Morphometric Analyses by a New Slit-Lamp Endothelial Biomicroscope

Criticality of the measurement of corneal thickness in specular reflection by digital biomicroscope

PURPOSE

The aim was to evaluate the effects on corneal thickness (t) measurement, of asymmetry of the CCD position for a digital biomicroscope in specular reflection.

METHODS

t was deduced from the distance between reflexes from anterior and posterior corneal surfaces using a biomicroscope (Takagi SM70 N), with the illuminator either, (a) farther from sensor side (150 eyes) or, (b) closer to it (134 eyes). The distance between reflexes was also measured on a glass slide and a reference lens, with nominal thicknesses of 1580 and 520 μm, respectively. Corneal thickness (t_PACH) was also measured by pachymeter (Canon TX-20 P).

RESULTS

When biomicroscope asymmetry was ignored, t for the glass slide was (a) 1760 and (b) 1404 μm. Correcting for the asymmetry provided corresponding values of 1588 and 1591 μm. For the lens, t was (a) 696 and (b) 543 μm, or 642 and 497 μm, when using the approximation of parallel surfaces (APS). Correcting for the asymmetry gave 565 and 552 μm (578 and 564 μm, with APS). Mean corneal t was (a) 560 and (b) 467 μm, (564 and 468 μm, with APS). Correcting for asymmetry gave (a) 506 and (b) 529 μm (508 and 530 μm, with APS). Mean t_PACH was 552 μm.

CONCLUSION

Biomicroscope asymmetry critically affects corneal thickness measurement in specular reflection. Induced errors can be accounted for and corrected, however. While the correction to the curvature radius is clinically relevant, it plays a minor role compared to asymmetry.

Variable-Size Bead Layer as Standard Reference for Endothelial Microscopes

PURPOSE

For morphometric analysis of the cell mosaic of corneal endothelium, checking accuracy and precision of instrumentation is a key step. In this study, a standard reference sample is proposed, developed to reproduce the cornea with its shape and the endothelium with its intrinsic variability in the cell size.

METHODS

A polystyrene bead layer (representing the endothelium) was deposited on a lens (representing the cornea). Bead diameters were 20, 25, and 30 μm (fractions in number 55%, 30%, and 15%, respectively). Bead density and hexagonality were simulated to obtain the expected true values and measured using a slit-lamp endothelial microscope applied to 1) a Takagi 700GL slit lamp at 40× magnification (recommended standard setup) and 2) a Takagi 2ZL slit lamp at 25× magnification.

RESULTS

The simulation provided the expected bead density 2001 mm and hexagonality 47%. At 40×, density and hexagonality were measured to be 2009 mm (SD 93 mm) and 45% (SD 3%). At 25× on a different slit lamp, the comparison between measured and expected densities provided the factor 1.526 to resize the image and to use the current algorithms of the slit-lamp endothelial microscope for cell recognition.

CONCLUSIONS

A variable-size polystyrene bead layer on a lens is proposed as a standard sample mimicking the real shape of the cornea and the variability of cell size and cell arrangement of corneal endothelium. The sample is suggested to evaluate accuracy and precision of cell density and hexagonality obtained by different endothelial microscopes, including a slit-lamp endothelial microscope applied to different slit lamps, also at different magnifications.