Observations of cortical object representations in the human EEG - Problems and Solutions

It is a common finding that neurons synchronize their firing rates in a high frequency range above 30 Hz, the so‐called gamma band, if they are involved in the processing of the same stimuli. Furthermore, it was hypothesized that object representations are stored within the cortical networks that are formed of synchronously firing neurons. Consequently, induced gamma band responses in the electroencephalogram were utilized to investigate cortical object representations. A more recent finding by Yuval‐Greenberg et al. questioned this common interpretation. It was shown that miniature eye movements as they occur during fixation are accompanied by small artifacts that mimic activity in the gamma band. The question arose whether induced gamma band responses in the electroencephalogram are a marker of neuronal processing or a mere artifact caused by miniature eye movements. The present thesis aims at providing solutions for this dilemma by investigating whether evoked gamma band responses and steady state visual responses are suitable tools to examine cortical network activity in relation to object recognition. Both measurements are not affected by miniature eye movements. It was found that evoked gamma band responses are particularly qualified to study early processes of object recognition. Furthermore, steady state visual evoked potentials were sensitive to the semantic content of presented stimuli. Therefore they are feasible for studying object recognition, too. Additionally, the present thesis provides the COSTRAP algorithm, a new method to effectively identify and remove the eye movement artifacts related to miniature eye movements. The remaining induced gamma band responses are most certainly of cortical origin. They proved to be sensitive to object recognition and repetition priming. Finally, it was shown that miniature eye movements and cortical gamma band responses can be experimentally dissociated, further underpinning the different methodological importance of induced gamma band response when studying object recognition. In summary, it was shown that the investigation of cortical networks representing object knowledge can also be addressed by evoked gamma band responses and steady state visual responses. Furthermore, it was proven that induced gamma band responses remain a powerful tool in electrophysiological research when studying the neuronal dynamics behind object recognition and other cognitive processing