Purchase this article with an account.
Denis Fize, Ghislaine Richard, Christophe Jouffrais, Michèle Fabre-Thorpe; EEG signals of rapid visual categorization in monkeys, in V4 area. Journal of Vision 2009;9(8):796. doi: 10.1167/9.8.796.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Recent experiments showed that macaque monkeys are able to perform accurately visual categorization of natural scenes with extremely fast saccadic reaction times, from 100ms (Girard, Jouffrais & Kirchner, 2008). Considering this time constraint and the neuronal response latencies in the inferotemporal cortex, it is hypothesized that mid-complexity visual cortical regions like V4 area could play a crucial role in the extraction of visual cues that could be sufficient for task performance. We recorded intracranial EEG on two head-free monkeys during natural scenes categorization, in which they had to respond by a button release and a screen touch to the presence of an animal in gray-scaled pictures flashed for 50ms. Corticograms were recorded using several electrodes placed in the neighborhood of V4 area; electrode locations were assessed using X-ray and MRI scans. Results showed that event-related potentials on correct trials (ERPs) recorded on V4 area exhibited significant higher amplitudes in response to target stimuli from 80ms after stimulus onset, during a regional recruitment that picked at 96ms. No effect of picture familiarity was observed at these latencies. At low stimulus contrasts, the differential activity between target and distractor ERPs remained of equal amplitude but was slightly delayed. Further experiments showed that this differential activity and its latency were robust to spectral amplitude equalization of the stimuli. These results likely indicate that feature shape could be the main visual cue supporting these fast visual processes. We conclude that relatively simple shape analyses performed by mid-complexity visual areas could be sufficient to perform rapidly object categorization in natural scenes.
This PDF is available to Subscribers Only