Nicole Rust, 2/18/03
<br>
<br>

Spike-triggered characterization of excitatory and suppressive stimulus dimensions in monkey V1
<br>
<br>

There are numerous means by which a phase-invariant, direction-selective detector can be constructed.  How does the visual cortex implement this computation?  To investigate, we applied a spike-triggered covariance analysis to binary noise data collected from monkey V1 neurons.  Starting with a high dimensional stimulus space, this technique allowed us to constrain a linear subspace of interest for a neuron, including dimensions of excitation and suppression.  We found evidence for more relevant stimulus dimensions than are predicted by standard models of visual processing:  as compared to the prediction of one excitatory dimension for a simple cell and two for a complex cell, we found no fewer than two excitatory dimensions for simple cells and as many as seven for complex cells.  We also found evidence for the existence of suppressive dimensions that were at least equal in number to excitatory dimensions.  These results suggest that extensions to standard models are required to fully describe the implementation of the direction selective computation in V1.