hide
Free keywords:
-
Abstract:
Recently, we have been able to read out a human/monkey face category-boundary from singleunit-activity (SUA) recorded from the inferior-temporal (IT) cortex of a macaque monkey
brain. This data was collected in an experiment where monkeys have to fixate at pictures of
human/monkey morphed faces at different levels of this ‘species-continuum’. Consistent with
our previous psychophysical experiments in which human subjects have to categorize morphed
faces as humans or monkeys, the perceptual boundary seems to be shifted towards the
‘own-species’ category (approximately 60 human/40 monkey in humans and the other way
around in the monkey data). Similar to the ‘other-race’ effect, this effect suggests a perceptual
bias that could be due to long-term learning.
The local field potential (LFP) refers to the low-frequency (< 300Hz) component of signals
recorded from the brain, and it has been associated with dendritic activity within a particular
recording area. In this work we investigate to what extent these LFP signals are stimulus
selective and weather they correlate with our previous results obtained from the simultaneously
recorded spiking activity (SUA).To achieve that, we first extract different features from the LFP
signals such peak amplitude, time-onset or the spectral power of different frequency bands. To
evaluate the information content of these features in relation to our stimulus and the spiking
data, we use statistical analyses, information theory and pattern classification.
Preliminary results show that features such as peak onset-time and peak-amplitude differ
significantly across stimulus-conditions. In contrast to the spiking data, when using these features,
the pattern classifiers set the face category-border without a consistent shift towards the
monkey category. Further analysis of these features using information theory will be needed to
test possible correlations with the spiking data and the stimulus properties.
