|M.Sc Student||Maizlish Dov|
|Subject||Electrophysiological Correlates of Imagination Studied by|
Components Decomposition and Source Estimation
|Department||Department of Biomedical Engineering||Supervisor||PROFESSOR EMERITUS Hillel Pratt|
|Full Thesis text - in Hebrew|
Fourteen healthy subjects participated in the research. The subjects listened to a constant pace sequence of natural sounds (door knocks), and were asked to imagine hearing a sound of a door slam at the end of each sequence. In the main control paradigm, the subjects listened to the same sequence of door knocks, but were not asked to imagine anything after the sequence. The physical auditory stimuli were identical in the two paradigms, and no external response was required from the subjects.
The complexity of the imagination experience, the lack of knowledge on the unfamiliar waveforms (resulting from the almost complete absence of previous experiments on auditory imagery using ERP) and their high variability, necessitate the use of source separation tools and to develop a software tool for automatic and objective peak marking (first of its kind, to our knowledge) that will overcome problems of repeatability and objectivity in peak marking of the ERP waveforms. We found a long frontal positivity during the expected imagery period, which was noted only in the imagery paradigm.
ICA - Independent component analysis - was found effective in separating brain activity components common to all the subjects.
Source localization was performed by LORETA - LOw Resolution brain Electromagnetic Tomography Analysis. LORETA was applied once on the original waveforms and once on the waveforms that were accounted for by 2 ICA components assumed to reflect imagery processes.
Premotor and parietal areas were found to be more active in the imagery paradigm at the expected imagery period, and frontal areas were found to be less active in the imagery paradigm at that time. The premotor and the parietal activation can be ascribed to the difficulty to isolate the auditory aspect of imagination, as reported by the subjects. Decreased activation of frontal areas might be explained by inhibition of some cognitive processes when waiting with expectancy for an event or when attempting to produce imagery.
LORETA applied on the waveforms reconstructed by the selected 2 ICA components, discovered areas in the bilateral temporal lobes that were more active in the imagery paradigm at the expected imagery period. These findings support previous research in different imaging modalities (e.g. fMRI). The combination of ICA-LORETA procedures was found effective in revealing additional information hidden in the complete waveforms.