Purchase this article with an account.
Ted Maddess, Andrew James, Thilini Perera, Maria Kolic, Corinne F Carle; Characteristics of gain control in the pupillary pathway - relevance to pupil perimetry. Invest. Ophthalmol. Vis. Sci. 2015;56(7 ):3181.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To understand aspects of the divisive gain control mechanism present in the pupillary system, which generates smaller responses/stimulus when stimuli are concurrently presented. This is an important factor in all forms of pupil based perimetry including multifocal pupillographic objective perimetry (mfPOP), which we have demonstrated has excellent sensitivity, specificity and structure/function relationships in glaucoma. Previous experiments had indicated that the gain control occurs at or after the Edinger-Westphal Nucleus (EWN). Two experiments examined: 1) the integration time of the gain control, and 2) the effect of the gain control on simulated visual field damage.
Experiment-A (ExA) had 16 participants (9 females, 23.4 + 4.0 SD years), and Experiment-B (ExB) had 13 participants (7 females, 24.1 + 3.6 SD years). Both experiments used dichoptic mfPOP stimuli and had 44-regions/eye within the central 60°. ExA explored temporal summation and used 7 stimulus types having mean presentation intervals (MI) that ranged from 0.27 to 16 s/region. ExB explored spatial summation using 9 stimulus types that each contained 4 categories of stimuli mimicking 9 degrees of visual field loss. Spatio-temporal gain control models attempted to predict the results from the 23 stimulus types. The models assumed an integrated version of the stimulus history divided the response to give the final response size for each protocol.
In ExpA response amplitude grew with increasing MI until MIs of 4 to 8 s/region with model gain governed by an integration time of 3 s. In ExB two models were compared: 1) a local gain control model in which each of the 3 region types was assumed to have its own control signal; and 2) a global gain control where the responses of the sets of 3 stimulus types were pooled before integration. The global model matched very well, predicting means and variances in the responses to within 15% of the observed values.
A global divisive gain control with a fixed integration time was indicated and global or near global spatial integration is highly likely.
This PDF is available to Subscribers Only