Copy the page URI to the clipboard
Liu, LiChan
(1995).
DOI: https://doi.org/10.21954/ou.ro.0000fb74
Abstract
The biomagnetic inverse problem has no unique solution, nevertheless even a cursory look at the features shown in raw signal can often suffice to highlight strong superficial activity. To do a proper single epoch analysis is normally prohibitively expensive in terms of computing demands. Hence the original aim of this thesis was to use simple efficient signal transformations to characterize superficial generators and contrast the single epoch signature with that extracted from the average signal. The results have intrigued us sufficiently to go beyond the original goal and extract very preliminary estimates of activity across the cerebral hemisphere in single trials.
The original tool, and one that we have used for much of the work, is a simple vector signal transformation called V3. This signal transformation highlights nearby sources; it is a crude but quick estimator of generators directly from the raw MEG signals. Together with Magnetic Field Tomography (MFT), which relies on distributed source analysis of the MEG signals, we have tackled the following specific problems relating to aspects of normal brain function: efficient estimation of generators of magnetic fields; relationship between the average signal and single trials; and interhemispheric differences and relationship between the activity in the left and right hemispheres of the brain.
During the project, we have used as examples auditory evoked MEG measurements obtained from two multichannel systems and applied the V3 and MFT analysis to both the average and single trial signals. In particular, we chose the 40-Hz (or gamma band) auditory response as the study subject. We found that in single epochs similar patterns of high frequency activity are observed in the area around the auditory cortex well before, close to and well after stimulus onset; the sequence of events observed in the average can only represent the evolution of events in single trials in a statistical way; and deep and central areas of the brain may be the seeds for the main deflections observed in the auditory responses.
CORE (COnnecting REpositories)
CORE (COnnecting REpositories)
