Identification and segmentation of cells in images of donated corneas using mathematical morphology

The most effective optical method for evaluating donated corneas is to count the number of living endothelial cells. The evaluation of the whole endothelium is usually done in a slit lamp, and the aspects of the cells are subjectively observed. In order to overcome the limitations of subjective observation and to provide an analysis of several parts of the cornea, we have developed a system, attached to the slit lamp, with a dedicated software package for image treatment and cell counting using mathematical morphology. The results are in good agreement with measurements done with the interactive counting method (the correlation coefficient is 0.95798).