This title appears in the Scientific Report : 1999 

Über lineare und nichtlineare Transformationen neuroelektromagnetischer Signale : Theorie und Anwendung am Beispiel einer Sprachstudie
Groß, J. (Corresponding author)
Institut für Medizin; IME
Jülich Forschungszentrum Jülich GmbH Zentralbibliothek, Verlag 2004
X, 184 p.
Düsseldorf, Univ., Diss., 1999
Dissertation / PhD Thesis
ohne FE
Berichte des Forschungszentrums Jülich 3673
Please use the identifier: in citations.
Magnetoencephalography (IvlEG) is an important technique for the investigation of cognitive processes like language processing since it offers a very high temporal resolution and the measured signals are rather independent of differences of conductivity within the human head, Modern whole-head systems allow the simultaneous measurement of the magnetic flux at more than hundred locations outside the head. The interpretation of the data is complicated due to the complex structure of the sensitivity profile (leadficld) of a single sensor. After an introduction to the basic concepts of magnetoencephalography the structure of data space is investigated and described based on the leadfields for a specific case. Figures of merit are introduced which allow a characterization of the relevant properties of the MEG system. The complex structure of data space can be modified with linear transformations in order to enhance desired properties of the data. Different methods are compared within the common framework of optimization of functionals. This framework necessitates the mathematical formulation of the desired properties. Another part describes the results from a language study. In this study the MEG correlates of syntactic processing were investigated by means of linear and nonlinear analysis methods. A characteristic component was evident in the data that has been linked to the fast processing of syntactic information. This finding is consistent with previous EEG studies. Generators of this component were localized in the vicinity of the auditory cortex and in inferior regions of the frontal lobe. The results from data measured with different ]VIEG systems were consistent