Perimetric defects and ganglion cell damage: interpreting linear relations using a two-stage neural model

Area of Science:

• Computational neuroscience
• Ophthalmology
• Visual psychophysics

Background:

• Glaucoma is characterized by progressive visual field defects.
• These defects are linked to the loss and dysfunction of retinal ganglion cells.
• The precise relationship between cell damage and visual field loss remains incompletely understood.

Purpose of the Study:

• To develop a neural model interpreting the relationship between glaucomatous visual field defects and ganglion cell damage.
• To investigate how linear relationships observed empirically between perimetric defects and ganglion cell loss arise.

Main Methods:

• A two-stage neural model was employed, simulating ganglion cell mosaic responses and subsequent spatial filtering.
• Ganglion cell death and dysfunction were simulated in localized areas.
• Perimetric sensitivity was calculated across various spatial filter frequencies and ganglion cell densities.

Main Results:

• The model demonstrated nonlinear relationships between decibel-measured perimetric defect depth and percentage of ganglion cell damage.
• These relationships became linear at lower spatial frequencies when perimetric defect was expressed as a percentage of normal.
• Ganglion cell dysfunction, in addition to cell loss, was necessary to accurately fit experimental glaucoma data.

Conclusions:

• The spatial filtering and pooling mechanisms within the visual system can explain the empirically observed linear relationships between visual field defects and ganglion cell loss.
• The tuning of spatial filters significantly influences the observed relationship between cell damage and visual field sensitivity.