Item

Abstract

Under certain stimulus conditions we encounter pronounced perceptual suppression of suprathreshold visual stimuli. The brain mechanisms underlying these phenomena are poorly
understood. Binocular rivalry (BR) and Binocular Flash Suppression (BFS) provide us excellent
behavioural tools to study this phenomenon. During these paradigms visual stimuli are completely
extinguished from our awareness for a substantial amount of time despite being physically
present on our retinas. Therefore, we can study the dissociation between the neural responses
that underlie a mere sensory representation of the visual input and what is perceived. Primary
visual cortex (V1) has been implicated as an important candidate for the site of perceptual
suppression. However, interestingly electrophysiological studies in V1 have found only a very
small percentage of neurons to be correlated with the percept[1]. In contrast, human fMRI
studies[2,3] have shown that the BOLD signal during such perceptual alternations modulates
almost as much as when the stimuli are non-ambiguously presented separately. These
contradicting results led to the speculation that the local field potential (LFP) signals, which have
been shown to correlate with the BOLD signal, will also show correlations with perception in
agreement with the BOLD results and thus potentially solve the apparent controversy. To this
end, a recent study[4] claimed that low frequency (<30Hz) LFP signals in V1 correlate well with the
subjective experience of macaques during BR.
We have used BFS and recorded neural activity from large populations of well-isolated single
neurons (SUA) from V1 using chronically implanted and non-chronic tetrodes in awake behaving
macaques. In addition to the SUA we also simultaneously recorded multi-unit (MUA) and LFP
signals. In agreement with previous electrophysiology experiments we find a very small
percentage of single neurons (12, t-test: p<0.05) as well as MUA sites (15) to be correlated
with the animals¹ percept during the binocular presentation of two gratings of orthogonal
orientations. Interestingly, an even smaller percentage (7) of gamma-band LFP sites show a
significant modulation and no other LFP band (e.g. alfa or beta-bands) showed stronger
perceptually related modulation. In addition, the amplitude of the normalized population response
in all three signals shows a small fractional modulation in comparison with the monocular
presentation of the gratings (see figure). We therefore conclude that the activity in V1 is not a
good predictor of the perceptual alternations at least using the classical simple measures of firing
rate and power modulations of the signals.