A Novel Optical Coherence Tomography-based Keratoconus Diagnostic Index Incorporating Stromal and Epithelial Features

Area of Science:

• Ophthalmology
• Medical Imaging
• Biomedical Engineering

Background:

• Keratoconus diagnosis can be challenging, especially in early stages.
• Existing diagnostic tools have limitations in discriminating early keratoconus.
• Spectral-domain optical coherence tomography (SD-OCT) offers detailed corneal imaging.

Purpose of the Study:

• To develop a novel diagnostic index for keratoconus using SD-OCT.
• To evaluate the diagnostic performance of the new index.
• To compare the new index with established keratoconus diagnostic parameters.

Main Methods:

• SD-OCT and Scheimpflug tomography were performed on normal and keratoconic eyes.
• Machine-derived SD-OCT parameters were analyzed for cornea, stroma, and epithelium.
• A new index, Whole Information of Stroma and Epithelium (WISE), was developed using statistical modeling and validated against existing methods.

Main Results:

• The WISE index demonstrated high diagnostic accuracy: AUC of 0.76-0.83 for forme fruste keratoconus and 0.92-0.94 for early keratoconus.
• WISE index performance was comparable to established parameters like Belin-Ambrósio Deviation and Pentacam Random Forest Index.
• Individual SD-OCT parameters showed limited discriminatory power, highlighting the benefit of a combined index.

Conclusions:

• Combining stromal and epithelial SD-OCT data into the WISE index significantly improves keratoconus discrimination.
• The WISE index provides a powerful, OCT-based tool for diagnosing keratoconus, including early stages.
• WISE index accuracy is on par with current gold-standard Scheimpflug-based diagnostic indices.