Original Article June 17, 2026 June 17, 2026

Epithelial thickness profiles in normal and keratoconus eyes identified by Artificial Intelligence

DOI: 10.37039/1982.8551.20260056

ABSTRACT

Objective:

To identify corneal epithelial thickness profiles for early keratoconus diagnosis with the aid of Artificial Intelligence.

Methods:

A total of 496 eyes (306 normal and 190 with keratoconus) were analyzed using optical coherence tomography images.

Results:

Epithelial segmentation revealed distinct patterns between groups, highlighting early changes in epithelial thickness in keratoconus eyes. The findings reinforce the potential of artificial intelligence in disease screening, enabling a more precise and objective diagnostic approach.

Conclusion:

Implementing algorithms such as k-means can significantly contribute to the early identification of keratoconus and the optimization of clinical management.