Detection of Keratoconus in Clinically and Algorithmically Topographically Normal Fellow Eyes Using Epithelial...

Area of Science:

• Ophthalmology
• Medical Imaging
• Biomedical Engineering

Background:

• Keratoconus is a progressive corneal ectatic disorder.
• It is often diagnosed unilaterally, but is increasingly recognized as a bilateral, albeit asymmetric, condition.
• Early detection in the fellow eye is crucial for monitoring and management.

Purpose of the Study:

• To evaluate the efficacy of a keratoconus detection algorithm.
• The algorithm utilizes epithelial thickness maps from Artemis very high-frequency (VHF) digital ultrasound.
• The study focused on clinically normal fellow eyes of patients with unilateral keratoconus.

Main Methods:

• The study included 10 patients with unilateral keratoconus.
• Epithelial thickness data were acquired using VHF digital ultrasound.
• A classification model was applied to identify keratoconus in the clinically normal fellow eyes.
• Corneal topography data (Pentacam BAD-D and Orbscan SCORE) were also analyzed for comparison.

Main Results:

• The VHF digital ultrasound epithelium model identified keratoconus in 5 out of 10 clinically normal fellow eyes.
• The Orbscan SCORE model classified 5 out of 9 fellow eyes as keratoconic.
• When combining VHF data with Pentacam BAD-D, one eye was identified by a combined model, while BAD-D alone did not detect keratoconus.

Conclusions:

• The findings suggest that epithelial thickness mapping can detect subclinical keratoconus in up to 50% of 'normal' fellow eyes.
• This supports the hypothesis of bilateral and asymmetric disease progression in keratoconus.
• The tested topographic parameters (Orbscan SCORE, Pentacam BAD-D) were less effective than epithelial thickness maps alone or in combination.