“School of Cognitive Sciences”
Back to Papers HomeBack to Papers of School of Cognitive Sciences
| Paper IPM / Cognitive Sciences / 11543 |
|
||||||||
| Abstract: | |||||||||
|
It has previously been shown that stimulus contrast affects the degree to which local neuronal populations in visual cortex are synchronously active: as assessed by the amplitude of the spike-triggered local field potential (LFP), increasing contrast decreases correlations (Nauhaus et al., 2009). Is this effect due to stimulus contrast only or does it depend, in a more general way, on stimulus strength? We used a multi-electrode system and recorded single-unit activity and LFPs from area MT of three awake, behaving monkeys while presenting stimuli varying in strength. We compared neuronal activity during the absence of a visual stimulus inside the receptive field, and during the presentation of a null- or preferred-direction random dot pattern (RDP). These conditions provided no drive, weak drive, and strong stimulus drive, respectively. We assessed correlations of the local neuronal population by computing the spike-triggered LFP and found that its amplitude decreased with increasing stimulus strength. In all three stimulus conditions the spike-triggered LFP consisted of a negative-going deflection peaking at 5 ms after spike occurrence. This peak negativity was largest for the condition without sensory stimulus (0.90 � 0.04 s.e., z-score), intermediate for the condition with the null-direction (0.64 � 0.03), and smallest for the preferred-direction stimulus (0.39 � 0.02). The effect of stimulus drive on local neuronal correlations cannot simply be explained by differences in the number of spikes serving as triggers for averaging LFPs. Reducing the number of trigger spikes by random removal of half of the spikes in the spike train did not influence the amplitude of the spike-triggered LFPs. Therefore, independent of the absolute number of trigger spikes, local neuronal correlations are stronger in stimulus conditions with low firing rates and weaker in conditions with high firing rates. We also found that local correlations fall off over a distance of 300 m. Using spikes from one electrode as trigger points, we determined the amplitude of the averaged LFP from the same and up to four neighboring electrodes. The amplitude of the spike-triggered LFP was largest when the LFP was recorded from the same electrode as the trigger-spikes, and was already significantly reduced for LFPs recorded at the inter-electrode distance of 300 m. We conclude that local neuronal correlations decrease with increasing stimulus strength. Contrast seems to be only one stimulus feature that influences synchronous activity in a local neuronal neighborhood. Beyond contrast, other stimulus features modulating firing rates can also affect correlations of neuronal populations.
Download TeX format |
|||||||||
| back to top | |||||||||
