Zeitliche Muster neuronaler Aktivität : Stimulusgekoppelte und intrinsisch generierte Komponenten

Abstract

Temporal Patterns of Neural Activity - Stimulus-Coupled and Intrisically Generated Components The research projects reported in this thesis are concerned with the temporal structure of neuronal signals. Stimulus processing in the brain is accompanied by simultaneous activation of a large number of neurons. In this connection, we investigated the hypothesis that neurons enable stimulus processing in dynamically composed assemblies, where the synchronization of responses to the same stimulus represents the binding signal of an assembly. To this end, multiunit activity was recorded from cortical area 17 of anaesthetized cats while moving sinusoidal gratings were presented on a computer screen. On the one hand, we studied the temporal patterns that are caused by intrinsic synchronization of responses. On the other hand, flickering stimuli were shown in order to evoke a synchronization phase-coupled to the stimulus and to compare the two types of temporal patterns. The results presented here provide the first direct evidence of larger assemblies showing either the same or different synchronization patterns, depending on wether the neurons responded to the same or to different stimuli. This behaviour was observed to the same degree in sites responding exclusively to only one surface as in sites whose response to one of the two surfaces was up to 48% of that to the other (ch. 3.1). Also, we were able to show that neurons can phase-lock their responses to the temporal stimulus course even if high-frequent aperiodic flicker stimuli are used, and that the resulting patterns of phase-locked synchronization can relate these responses to either the same or to different stimuli (ch. 3.2). However, the cortex is able to override the phase-coupled temporal structure of neuronal activity. As shown here, the network properties are able to restrain the neurons from phase locking if the stimulus evokes sufficient degrees of gamma activity (ch. 3.3). The objection that the results provided by investigations on correlated neuronal activity were distorted by correlated fluctuations of the firing rate was refuted by an off-line filtering algorithm developed in this thesis. The results obtained with this algorithm, which removes slow changes from neuronal signals, confirm earlier reports on significant neuronal synchronization in the millisecond range (ch. 3.4).