The functional Role of Gamma-Band Synchronization in selective Routing and Network Configuration within the visual Cortex

Abstract

First psychophysical experiments performed more than 100 years ago by the German psychologist and physicist Hermann von Helmholtz, showed that visual attention is a central component of perception and, therefore, of substantial relevance for successful behavior. In the decades that followed, much research has been performed to investigate how attention modulates neuronal activity in order to explain the effects of attention on behavior and perception. A well-described finding is that visual neurons responding to the same attended object synchronize their activity in the gamma-frequency range (30 - 100 Hz). In chapter 2, I present the results of an experiment that was designed to find evidence for a causal role of gamma-band synchronization in selective information routing and processing. The underlying idea is that neurons, which synchronize their activity deliver their respective outputs (spikes) more precisely at times the receiving neuron is sensitive for it, i.e. the incoming spikes are more likely to evoke spikes of the receiving neuron. The selective synchronization between input and receiver neurons representing an attended and therefore relevant object could constitute a powerful selection mechanism. To test this gamma recorded neuronal activity in area V4 of two macaque monkeys while applying single electrical pulses to neurons located in area V2. Those V2 neurons delivered afferent input to the recorded V4 population, including the electrically evoked spikes. By relating the effects of these electrically evoked spikes to the gamma-oscillation in V4, I could show that the impact of stimulation on behavior and neuronal activity is causally dependent on the gamma-phase. In chapter 3, I investigated whether the effective processing of a given object requires a specific level of gamma-band synchronization within a local neuronal population. I hypothesized that different objects require different combinations of neurons of the same population to be functionally coupled with one another for effective processing. Furthermore, we hypothesized that this dynamic establishment of functional connections is implemented by gamma-band synchronization, resulting in a specific level of gamma-band synchronization for a specific stimulus. I tested these predictions by first recording neuronal activity in area V4 and quantifying the level of gamma-synchronization in response to two different single stimuli, which had to be attended. Second, I compared these levels to the level of gamma-synchronization when neurons received input of both stimuli simultaneously, and one of them was attended. The level of gamma-synchronization was almost 'as if' the attended stimulus was presented alone, strongly indicating that the processing of this stimulus requires this specific gamma-synchronization level. Chapter 4 describes and characterizes a method that I used for analyzing multi-unit activity in area V4. It does not rely on setting up an amplitude-threshold for separating spikes from background noise as standard procedures do. Thus, this measure takes the entire spike activity into account, which I, therefore, refer to as ESA. I used semi-chronically recorded data of five macaque monkeys in order to quantify the sensitivity of the ESA to detect neuronal responses. The ESA-signal was significantly more sensitive than the standard procedures, especially for data with low signal-to-noise ratio, but preserves information about receptive field sizes and orientation selectivity of the underlying neuronal population. The fifth chapter is describing a method for offline stimulation-artifact removal and restoration of the original broadband neuronal signal. I could show that in contrast to existing methods the here described procedure does not disturb the original signal and therefore allows for analysis of neuronal activity even shortly after electrical stimulation. In summary, the results presented here give further insight into the functional roles of gamma-band synchronization. I could show that (1) gamma-phase synchronization plays a causal role in selective information processing and routing, and (2) that a specific pattern of intra-areal gamma-synchronization is required for effective processing of a given stimulus.