Text

ETD_2751

http://hdl.rutgers.edu/1782.1/rucore10002600001.ETD.000056110

theses

Cortical information processing depends critically on an animal’s brain state. Previous research has revealed there is a great deal of variability in cortical responses to repeated stimuli. This thesis addresses the question of whether activity and response variability in rat auditory cortex depends on brain state. Specifically, we hypothesized that both spontaneous and evoked activity differ between states; furthermore that cortical responses in higher-frequency “desynchronized” EEG states would be less variable and follow sensory input up to higher temporal modulation frequencies. We first assessed the spontaneous activity of auditory cortex during silence. During synchronized “slow wave” EEG states the spike counts of individual neurons in sequential time bins were irregular, but this irregular firing was coordinated across the neural population. Spike counts were more regular following a tail pinch-induced shift to higher-frequency EEG, and the population-wide coordination disappeared. We also uncovered a set of high-firing neurons with independent, rhythmic activity during desynchronized states, peaking between 8 to 18 Hz. Next we characterized responses to loud single-click stimuli. Many neurons discharged short-latency spikes with similar latency across states. These preserved spike latencies manifested as brief, sub-50ms population sequences of activity with similar profiles in different brain states. In some experiments we observed late, long-lasting effect of clicks on firing rates in synchronized states. In our last study, we show that evoked local field potentials (LFPs) can follow high-frequency amplitude modulations of broadband noise during desynchronized regimes. Spikes also track input more reliably and can be better predicted from stimuli in desynchronized states than in slow-wave states. Finally, we address whether LFPs reliably predict neural activity, and show that in most cases LFPs explain more spiking variability than our amplitude-modulated white noise stimuli. Thus much ‘noise’ in neural responses is not cell-specific, but reflects a source shared across many cells; such variability is state-dependent, and can be accounted for by LFP dynamics. Our first studies demonstrate that despite clear changes in spontaneous activity, strong onset responses to discrete broadband stimuli are often preserved across states. The final study suggests the desynchronized state supports improved representation of temporally modulated stimuli in auditory cortex.

Ph.D.

Includes bibliographical references

Includes vita

by Stephan L. Marguet

doi:10.7282/T3XP74Q3

ETD doctoral

The author owns the copyright to this work.

ETD

application/pdf

application/x-tar

95088640

57b382a35c4d6f6e1f5d73bf9e5daf2c4592af8b