Milan 2012 Programme Registration Exhibition Hotels Exhibitor Listing Satellite Meetings Visa Information
Search Abstracts by author or title
(results will display both Free Papers & Poster)

Comparison of endothelial cell count by manual and automated methods in normal cornea and in Fuchs’ endothelial dystrophy

Poster Details

First Author: S.Murthy INDIA

Co Author(s):    V. Rathi   D. Dutta   P. Kumar        

Abstract Details


To evaluate the difference in endothelial cell count between automated and manual methods in healthy cornea and Fuchs’ endothelial dystrophy.


L V Prasad Eye Institute, Hyderabad, India


- Archived images of endothlelium from the NIDEK Confoscan 4 were used to procure endothelial cell counts by automated and manual methods. Cell count difference between two methods of >250 cells/mm2 was considered significant.


Images from both healthy and Fuchs’ endothelial dystrophy were reviewed by one observer; 52 from healthy corneas and 65 from Fuchs’ dystrophy. In healthy endothelium, the average automated and manual cell count was 2471.24 ± 273.19 and 2444.50 ± 370.30 cells and the mean cell count difference between the two methods was 125.67 ± 86.9 cells (range 402 to 2 cells, p value ? 0.0463). In compromised endothelium, the mean cell count difference between the two methods was 633.04 ± 435.03 cells (range 54 to 1631 cells/mm2, p value <0.0001). On dividing into two groups: manual count >1000 cells/mm2 (49 images) and manual count < 1000cells/mm2 (51 images); the differences between the two methods were 312.96 ± 335.27 (p value <0.0001) & 953.12 ± 246.70 (p value <0.0001) cells/mm2.


Our study shows that while in healthy corneas, both methods showed similar counts; in the case of diseased cornea, the accuracy of the automated count decreased with decrease in total cells. Manual count is therefore more reliable in compromised endothelium as automated count overestimates the results, and can even give a value in normal range, thus misleading the clinician FINANCIAL DISCLOSURE?: No

Back to previous

loading Please wait while information is loading.