The magnetoencephalogram (MEG) is a recently developed method of noninvasively localizing and studying the summed intracellular currents of epileptic paroxysms in animal and man by mapping the extracranial magnetic fields that they produce. The long-term goal of this project is to use models of epilepsy in animal cortex to establish an empirical basis for the neurogenesis of epileptiform magnetic fields in man, and to combine MEG with detailed electrical recording to obtain information about the intra and extracellular currents produced by excitability changes in the in vivo epileptic neocortex. The present project will study a more complex cobalt focus in the more realistic gyrencephalic cortex of miniature swine. Our goal is to reduce the ambiguity and increase the realism of analytical solutions derived from physiological modeling of noninvasively recorded epilepsy data. We will approach this problem in three ways. First, we will combine information from both MEG and electroencephalogram (EEG) in all modeling procedures. Second, instead of simply comparing the results of modeling to the geometry of underlying anatomy, we will incorporate information about the location and shape of gyri and sulci into the modeling solutions directly. Finally, we will selectively filter data in both space and time to isolate contributions to the total system variance from sharp versus slow wave activity, and from focal versus regional activity respectively. We will evaluate and improve the accuracy of these modeling solutions by comparing them to detailed quantitative information about the intracranial distribution of cellular currents, information obtainable only invasively from an animal preparation. The results of this work will not only provide insights into membrane excitability changes that result in epileptic seizures, but will also be directly relevant to the interpretation of extracranial magnetic fields measured from normal and epileptic human neocortex.
Showing the most recent 10 out of 32 publications
