Purchase this article with an account.
P. Iserovich, A. Rosensweig, J.M. Sanchez, J. Fischbarg; Computerized technology applied to measurements of corneal transendothelial water movements . Invest. Ophthalmol. Vis. Sci. 2004;45(13):424.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: The Dikstein and Maurice technique is a crucial tool to investigate the physiology and pharmacology of corneal endothelial fluid secretion. To eliminate subjectivity and improve the accuracy of the measurements, optical improvements and computerization of data collection and treatment appeared feasible. Methods: We used the standard deepithelialized rabbit corneal preparation as described earlier. Optical design was based on a Zeiss universal microscope modified so it operates as a specular microscope with white or He–Ne laser light. The corneal thickness was scanned via vertical micro movements of the microscope stage driven by a step motor under indexer and computer control. Light reflected form the different planes was collected by a photomultiplier and digitized at high rate via a A/D board and computer; data were smoothed by low–pass filtering. Software was written in LABVIEW 5.1. Results: The system allows clear recognition of two peaks in light intensity produced by interfaces at the corneal endothelium, and at the anterior part of the stoma. The algorithm rescans the local neighborhood of the peaks until the error decreases to a desired level. The use of an oil immersion 40X objective requires that the measurement procedure has to be standardized in each cycle. With this, corneal thickness can thus be monitored automatically, and detailed data are saved for later examination. The accuracy is greatly improved over visual methods given the high rate of data acquisition, high light sensitivity, and the multiple scanning for the peaks. Experiments which usually last several hours can thus be carried out automatically practically eliminating human factors. Conclusions:The accuracy is greatly improved over visual methods, and experiments which usually last several hours can thus be carried out automatically practically eliminating human factors.
This PDF is available to Subscribers Only