Evaluation of a Method for Estimating Retinal Ganglion Cell Counts Using Visual Fields and Optical Coherence...

Area of Science:

• Ophthalmology
• Neuroscience
• Biomedical Engineering

Background:

• Glaucoma is a leading cause of irreversible blindness.
• Accurate estimation of retinal ganglion cell (RGC) counts is crucial for understanding glaucoma progression.
• Existing models aim to correlate structural and functional changes in glaucoma, but their generalizability requires validation.

Purpose of the Study:

• To assess the accuracy and generalizability of a published nonlinear model (H-NLM) for estimating RGC counts.
• To compare the H-NLM with a linear model (HK-LM) in relating structural and functional changes in glaucoma.
• To evaluate the underlying assumptions of the H-NLM using independent datasets and simulations.

Main Methods:

• Applied the H-NLM and HK-LM to an independent dataset including healthy controls, glaucoma patients, and patients with nonarteritic anterior ischemic optic neuropathy.
• Utilized frequency-domain optical coherence tomography and visual field data.
• Compared model predictions using topographic maps and performed simulations to test H-NLM assumptions.

Main Results:

• The HK-LM showed reasonable predictive performance (R2 = 0.31-0.64) when using the Garway-Heath et al. map.
• The H-NLM demonstrated poor predictive accuracy (R2 < 0) irrespective of the topographic map used.
• Simulations indicated that H-NLM estimates diverged significantly from histology-based RGC counts, and the added complexity of the H-NLM provided unclear value.

Conclusions:

• The assumptions underpinning the H-NLM require critical reassessment.
• Findings suggest that studies and models relying on H-NLM RGC estimates should be interpreted with caution.
• The HK-LM, particularly with specific topographic maps, offers a more reliable approach for glaucoma-related structural-functional correlations.