Logistic regression analysis for early glaucoma diagnosis using optical coherence tomography

Area of Science:

• Ophthalmology
• Medical Imaging
• Glaucoma Research

Background:

• Glaucoma is a leading cause of irreversible blindness.
• Early detection of glaucoma is crucial for preserving vision.
• Optical coherence tomography (OCT) measures retinal nerve fiber layer (RNFL) thickness, a key indicator in glaucoma diagnosis.

Purpose of the Study:

• To develop and validate a linear discriminant function (LDF) for diagnosing early glaucoma.
• To assess the diagnostic performance of the LDF using RNFL thickness at 12 clock-hour positions from OCT scans.
• To compare the LDF's accuracy against standard OCT parameters.

Main Methods:

• A prospective study involving healthy individuals and patients with open-angle glaucoma.
• Calculation of the LDF using RNFL thickness data from a training set.
• Validation of the LDF's diagnostic accuracy in an independent cohort using receiver operating characteristic (ROC) curve analysis.

Main Results:

• The developed LDF formula incorporates RNFL thickness from specific clock-hour segments.
• The LDF achieved high areas under the ROC curve (0.962 and 0.922) in both study populations.
• At 95% specificity, the LDF demonstrated a sensitivity of 74.5%, outperforming average RNFL thickness (67.8%).

Conclusions:

• The LDF significantly enhances the diagnostic capability beyond individual RNFL thickness measurements.
• The LDF exhibits superior sensitivity for discriminating between healthy and early glaucomatous eyes compared to standard OCT parameters.
• This novel LDF offers a promising tool for earlier and more accurate glaucoma detection.