Linear discriminant functions to improve the glaucoma probability score analysis to detect glaucomatous optic nerve...

Area of Science:

• Ophthalmology
• Medical Imaging
• Diagnostic Accuracy

Background:

• Glaucoma diagnosis relies on optic nerve head parameters.
• Heidelberg Retinal Tomograph 3 provides Glaucoma Probability Score (GPS).
• Improving GPS diagnostic capacity is crucial for early glaucoma detection.

Purpose of the Study:

• To develop a linear discriminant function (LDF) formula.
• To enhance the Glaucoma Probability Score (GPS) diagnostic capacity.
• To discriminate between healthy and glaucomatous eyes using optic nerve head parameters.

Main Methods:

• Multicenter cross-sectional study.
• Utilized Heidelberg Retinal Tomograph 3 for optic nerve head parameters.
• Calculated LDF formulas and evaluated diagnostic accuracy using ROC curves, sensitivity, and specificity.

Main Results:

• Two LDF formulas improved GPS algorithm results.
• GPS-only LDF achieved 79.7% sensitivity at 85% specificity and 71% at 95% specificity.
• GPS-RA LDF (including sectorial rim area) showed higher sensitivity: 85.5% at 85% specificity and 79.7% at 95% specificity.

Conclusions:

• Linear discriminant function (LDF) formulas slightly improved GPS diagnostic capacity.
• GPS LDF incorporating sectorial rim area parameters demonstrated superior diagnostic performance.
• Enhanced GPS LDF offers improved accuracy in distinguishing between healthy and glaucomatous eyes.