The proposed study extends our previous work on the physiological basis of magnetoencephalography (MEG) carried out with support from NINDS. The usefulness of MEG as a non-invasive technique depends very much on whether it can be used to determine the distribution of neuronal currents in the brain that reflect abnormal as well as normal neuronal activity. Our previous results with an in vitro cerebellar preparation indicated that the magnetic evoked field (MEF) may be directly related to intracellular currents produced by active neurons with the strong contribution coming from dendritic currents. These results have motivated our proposal here to determine the MEF that can be produced by different types of currents in the dendrites. The concept of post- synaptic currents has undergone dramatic transformations in the past twenty years with the discoveries of active conductances in the dendrites that can produce spikes in dendrites. These discoveries call for a re-examination of the role of dendritic currents in the generation of magnetic field as well as evoked potential. For this purpose we have received an instrumentation grant from National Science Foundation and installed a 4-channel, high-resolution magnetometer (muSQUID) that has a sensitivity of as much as 30-300 times the sensitivity of the more conventional, superconducting sensor we have used in our earlier studies. Its sensitivity enables us to see the MEF produced by a tissue as small as 1 mm3 without averaging with a bandwidth of 1 kHz. Our plan is to measure the MEF produced by the pyramidal cells in the longitudinal slice of the CA3 of the guinea pig hippocampus in the presence of various selective channel blockers and to interpret the MEF with the aid of the mathematical model developed by R. Traub of IBM. Traub model incorporates six active conductances (gNa, gCa, gk(DR), gA, gK(C), and gK(AHP) and NMDA-and quisqualate-channels within each cell and connects such cell with excitatory synaptic connections. This model and its earlier versions have been successfully used to account for intracellular potential in CA3. Based on this analysis we expect to be able to determine the relative contributions of these conductances to the MEF and, by extrapolation, to the evoked potential as well. The second series of study will evaluate the ability of MEG to determine the underlying current distribution. The usefulness of MEG will be limited unless one addresses this issue of MEG as a current-imaging tool. Therefore, we shift out approach from that of the earliest studies in biomagnetism which addressed the issue of accuracy of localization of the current dipole generation of MEF, and consider whether one can correctly infer the distribution of cortical activity based solely on measured external MEFs. Our plan is to produce simple and complex patterns of activation in the somatosensory cortex of a juvenile swine with a transcutaneous stimulation of the somatic afferents, measure the MEF directly over the exposed intact dura of this large brain (6x4x4 cm), then solve the inverse problem with minimum norm estimation algorithms and estimate the current distribution over the epidural and cortical surfaces. The prediction will be verified directly with a set of epi-and intracortical potential recordings over each of predicted active areas and with pharmacological manipulations.
Showing the most recent 10 out of 22 publications
