Share Email Print

Proceedings Paper

Automated image classification applied to reconstituted human corneal epithelium for the early detection of toxic damage
Author(s): Giovanni Franco Crosta; Chiara Urani; Barbara De Servi; Marisa Meloni
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

For a long time acute eye irritation has been assessed by means of the DRAIZE rabbit test, the limitations of which are known. Alternative tests based on in vitro models have been proposed. This work focuses on the "reconstituted human corneal epithelium" (R-HCE), which resembles the corneal epithelium of the human eye by thickness, morphology and marker expression. Testing a substance on R-HCE involves a variety of methods. Herewith quantitative morphological analysis is applied to optical microscope images of R-HCE cross sections resulting from exposure to benzalkonium chloride (BAK). The short term objectives and the first results are the analysis and classification of said images. Automated analysis relies on feature extraction by the spectrum-enhancement algorithm, which is made sensitive to anisotropic morphology, and classification based on principal components analysis. The winning strategy has been the separate analysis of the apical and basal layers, which carry morphological information of different types. R-HCE specimens have been ranked by gross damage. The onset of early damage has been detected and an R-HCE specimen exposed to a low BAK dose has been singled out from the negative and positive control. These results provide a proof of principle for the automated classification of the specimens of interest on a purely morphological basis by means of the spectrum enhancement algorithm.

Paper Details

Date Published: 2 March 2010
PDF: 11 pages
Proc. SPIE 7550, Ophthalmic Technologies XX, 75501N (2 March 2010); doi: 10.1117/12.841095
Show Author Affiliations
Giovanni Franco Crosta, Univ. degli Studi di Milano-Bicocca (Italy)
Chiara Urani, Univ. degli Studi di Milano-Bicocca (Italy)
Barbara De Servi, VitroScreen Srl (Italy)
Marisa Meloni, VitroScreen Srl (Italy)

Published in SPIE Proceedings Vol. 7550:
Ophthalmic Technologies XX
Fabrice Manns; Per G. Söderberg; Arthur Ho, Editor(s)

© SPIE. Terms of Use
Back to Top