Dipole parameter estimation of M50 auditory evoked fields applied to the study of training-induced neuroplasticity in schizophrenia

Abstract

In a first step three different methods for dipole localization of M50 auditory sources evoked by the double click paradigm were compared in order to differentiate between 12 schizophrenic patients (1 female) and 10 controls (1 female). Dependent variables were sensory gating ratios (dipolemoment of themagnetic counterpart of click two divided by click one) and hemispheric asymmetry (left vs. right hemispheric dipole location). MEG was measured using a 148 channels magnetometer system. Dipoleswere fitted using (1) a spherical head model with 34 MEG channels over the left and right temporal lobe, respectively, (2) a spherical head model with 68MEGchannels over both temporal lobes, (3) a boundary element model (BEM) based on an averaged magnetic resonance imaging (MRI) dataset (using 68 MEG channels). Differentiation between patients and controls for both gating ratios and anterior-posterior asymmetry was most precise when dipole moments were calculated with the BEM model. Thus, in a second step, the BEM method was used to investigate changes across 4 weeks, during which patients completed a cognitive training program. In patients, gating ratios were decreased after training but no changes in the asymmetry were noticed. Stability of measures was confirmed in controls.