Abstract
V2 but not V1 cells respond to illusory contours defined by inducers outside the cell's classical receptive field (von der Heydt 1984). Thus the “classical” rf (the region giving an audible response to a stimulus) must not comprise all the parts of the visual field that can contribute to the activation of a cell.
We studied 43 cells in V2 and 41 cells in V1. For each cell we first mapped the receptive field using an optimal moving bar stimulus and listening to the response; we will call this the “audible rf”. Second we re-mapped the receptive field using sparse noise, and defined the “noise-based rf” as those regions where the visually-driven spike rate was significantly above the spontaneous firing rate (p < 0.05). In V1 the two ways of measuring the rf gave comparable rf sizes, but in V2 the noise-based rf was invariably larger than the audible rf.
We then tested whether cells responded to illusory contours when:
The inducers were outside the audible rf but inside the noise-based rf
The inducers were outside the noise-based rf
Only in #1 did the cells respond to illusory contours; there was no significant response in # 2.
Thus regions outside the classical rf that can drive illusory-contour responseiveness are not qualitatively different from the classical rf, but are merely less effective in driving the cell, and therefore undetectable by ear. Non-linear response summation to collinear stimuli could then explain the responsiveness of V2 cells to illusory contours.
Supported by NIH (EY13135)