DIPOLE LOCALIZATION OF POTENTIALS ELICITED BY VISUAL PATTERN STIMULATION

This thesis was centered on two principal objectives. The first was an accurate description of the differences between single visual evoked potentials elicited by illuminated checkerboard stimulation of halves and quadrants of the visual field as determined by the feature selection technique. Emphasis was placed on the features that were consistent in the single potentials evoked by stimulation of one section of the visual field, but differed when another section of the visual field was stimulated. Forward sequential feature selection, employing a quadratic classifier for evaluating feature effectiveness, was used to identify the stimulus related features in four subjects. Each feature was restricted to be the voltage measurement recorded from one of six electrodes at a single latency. The resulting classifier recognitions of between 71% and 96% showed the extent to which these distinguishing features were present in the single evoked potentials. The second objective was to determine a relationship between the above features, the visual evoked potential at various electrode sites, and the possible neurological generators of these potentials. A single dipole source was chosen as a model of these neurological generators, and a homogeneous sphere was chosen as a model for the head. Dipole locations and moments were calculated for the average evoked potentials and average dipole locations and moments were calculated for the preprocessed single evoked potentials. The physiological appropriateness of these locations and moments were considered and a relationship was demonstrated between these parameters and differences (especially polarity inversions) observed in the average evoked potentials. A relationship was also demonstrated between the feature selection and classification results and the differences with respect to latency, electrode site, and the stimulus involved