LIN: Forschungsabteilungen > Akkustik, Lernen, Sprache > Unterpunkt Ebene 3 > Unterpunkt Ebene 4
Titel: LIN Layout
The auditory modality, and the phenomenon of hearing, is especially suited for studying the relationship between the neuronal processing of physical stimulus features on the one hand and the attribution of subjective meaning to such stimuli (Scheich and Ohl 2010; Scheich et al. 2011). We study both the physical and the semantic aspects of auditory perception on several scales of nervous system organization, from the microscopic level (single neuron activity), via the mesoscopic level (neuronal assemblies), to macroscopic level (EEG and fMRI) in both animals and humans.
On the level of single neurons we have developed new advanced methods of mathematical system analysis in collaboration with the laboratory of Jörn Anemüller (University of Oldenburg) that allow prediction of single-neuron responses to complex acoustic stimuli, such as human speech sounds. These methods improve previous techniques of the estimation of spectrotemporal receptive fields (STRFs) and are sensitive to two known important features of auditory neuron responses, viz. temporally extended response components (Campbell et al. 2010) and adaptivity to stimulus statistics on various time scales (Dahmen et al. 2010). The existence of neuronal plasticity simultaneously on various time scales is of general importance for the function of the auditory system (Brechmann et al. 2011).
On the mesoscopic level of cortical assemblies we have continued our previous work on the relevance of gamma oscillations for coordinating neuronal activity within and between cortical columns. Since gamma oscillations require the synchronization of many neurons across cortical space a potential role of long-range horizontal connections for their generation has been repeatedly suggested. So far, direct investigation of this hypothesis was hampered because no methodology existed to tease apart synaptic activity driven by horizontal connections from other synaptic activities without manipulating the connections themselves. To address this problem we have performed a CSD residual analysis (cf. topic Neuroprostheses) after surgically cutting horizontal connections orthogonal to the tonotopic gradient of frequency representations in auditory cortex (figure below). Analysis of the measures of horizontal activity (RelResCSD: relative residual CSD) and of total columnar activity (AveRectCSD: average rectified CSD).
The macroscopic level is the typical level of observation in human experiments. Here we have been able to demonstrate that characteristic features of electrophysiological responses as measured by human scalp EEG recording can be explained by individually different resonance frequencies of the neocortex (Zaehle et al. 2010). This relates macroscopic properties of neurophysiological oberservables in typical investigations on human subjects to mesoscopic mechanisms.
a) New experimental design to investigate the contribution of horizontal connections to cortical gamma oscillations. A single-channel Tungsten electrode and a multichannel-CSD electrode were implanted on either side of surgical cut of intracortical connections. b) CSD-derived measures of horiztontal activity (RelResCSD) and total columnar activity (AvgRecCSD). c) Time-frequency decomposition of these measures. d) Average total gamma activity was plotted separately for the 3 frequency groups Cut, BF and nonCut (red, blue and grey, respectively). The peak of the RelResCSD based gamma activity did not show a significant frequency dependency. In contrast, AvgRecCSD based gamma activity displayed a significant frequency relationship in that the BF elicited significantly higher peak values than nonBF frequencies.
SFB TRR 31