Purchase this article with an account.
B.B. Lee, H. Sun; Impulse statistics and positional information in ganglion cell signals . Invest. Ophthalmol. Vis. Sci. 2004;45(13):4656.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: Vernier thresholds with different target configurations may become similar if visibility is equalized by normalizing contrast to detection threshold. We consider if this is simply a consequence of ganglion cell signal strength or whether more complex mechanisms are implicated. Methods: Cell responses to different moving targets were subjected to a neurometric analysis to assess the positional accuracy of their signals. Results were compared to psychophysical performance with similar targets. Results: Under certain circumstances, e.g., using either bars or edges are vernier targets, ganglion cell response amplitude and spatial precision change in a way which resembles psychophysical performance. Under other circumstances, the situation is more complex; if impulse statistics result in unreliable signals from individual members of the ganglion cell array, psychophysical data suggest that central mechanisms change their spatial properties to make better use of a noisy ganglion cell signal. Finally, using target pairs of opposite contrast results in a moderate increase in vernier threshold, but neurometric analysis shows that cells can give little spatial information about targets with contrast inappropriate to their centers, e.g., an on–center cell and a dark edge. Thus comparison of local signs delivered by different pathways is required. Conclusions: Mechanisms underlying determination of spatial location are dependent on the size of the ganglion cell signal, but also may adopt flexible strategies to optimize performance. Making targets equally visible by normalizing contrast to detection threshold may be a useful rule of thumb but underlying mechanisms are complex.
This PDF is available to Subscribers Only