Hemodynamic response to oscillatory EEG rhythms in the human visual cortex

Abstract

Understanding the relationship between electrophysiological and hemodynamic signals is of superior importance to draw inferences from modern vascular based imaging techniques back to the underlying neuronal brain activity. Non-invasive studies fusing direct and indirect neuronal based methods such as electroencephalography (EEG) coupled to functional magnetic resonance imaging (fMRI) or near-infrared spectroscopy (NIRS) in human has become a significant approach to elucidate the neurovascular mechanisms and to validate invasive findings in animal. Currently, it is still unclear which electrophysiological components or which combinations of these components have the strongest influence on the hemodynamic signal. Therein, oscillatory brain activity in the alpha- and gamma à ¬range represents fruitful neuronal predictors because of their involvement in a widespread number of cognitive tasks and the assumption of oscillatory activity as a multifunctional coding mechanism in the visual system. The present dissertation aims to shed light on the coupling mechanisms between event-related oscillatory activity and the concomitant hemodynamic response in human. For this purpose, three studies with different experimental designs were performed in order to selectively induce oscillatory activity in the alpha- and gamma-range in the visual cortex and to test the influence of a particular oscillatory band on the hemodynamic response. To adequately address this neurovascular relationship EEG and NIRS techniques were applied simultaneously in all studies. The first study showed that the resonance phenomenon, a local maximum that appears when the stimulation frequency of flicker-light matches the endogenous alpha-frequency, is not accompanied by an increase in vascular parameters. Neither evoked potentials nor ongoing alpha-power or even a combination of both electrophysiological parameters predicted the magnitude of the hemodynamic response. It was therefore suggested that the resonance phenomenon is caused by a low energy demanding phase-resetting mechanism. Furthermore, it could be shown that the resonance boost observed in electrophysiological studies and the 8 Hz peak response observed with hemodynamic based techniques represent independent phenomena. Study II provided evidence for a predictive link between resting state alpha-parameters and evoked signals during flicker-light stimulation. It was shown that resting alpha-frequency is negatively related to the amplitude of evoked potential, to the induced alpha-power and to the magnitude of the hemodynamic response upon stimulation. The results provide further support for the assumption of a functional linkage between evoked potentials and alpha-rhythm (Study I&II). Study III showed a tight coupling between oscillatory activity in the gamma-range and the hemodynamic response during parametrical contrast variation of a visual moving grating. Also, it could be shown that during constant contrasts behavioural performance to the accompanied task was linked to the magnitude of gamma-activity and the hemodynamic response. Here, faster response times were preceded by a phasic enhancement of gamma-band activity. Thus, Study III provides further evidence for the superior role of gamma-oscillations in visual processing and behavioural performance and validates the close coupling between hemodynamic signal and gamma-activity noninvasively in human. In conclusion, by focussing on the same task with complementary methods, the studies of the present dissertation provide a further insight into the relationship between electrophysiological components and hemodynamic signals. Herein, simultaneous assessment of EEG and NIRS techniques provide a powerful tool to study the relationship between direct and indirect neuronal signals in humans.