MEG Single-event Analysis: Networks for Normal Brain Function and Their Changes in Schizophrenia.

Modern instrumentation offers the capability to capture in an instance the magnetic field around the head. There is agreement that the measurements are responsive to instantaneous changes of the current density inside the head, corresponding, at least partially, to neuronal events elicited by the stimuli and/or the task. There is however disagreement about the information that can be usefully extracted from the data. On the one hand theoretical arguments can be given why unambiguous reconstructions of generators from the data are impossible. On the other hand evidence has been gathering that surprisingly accurate information can be extracted from the data under minimal assumptions. Here we outline magnetic field tomography (MFT), a method for obtaining tomographic descriptions of brain activity from biomagnetic data. We show that if the underlying physics laws and biological complexity are to be respected then the non-linear reconstruction algorithm of MFT must be selected over the simpler minimum norm or other algorithms. Drawing on the enormous dynamic range of the measurements and the richness of information in the single trial data, MFT provides unique insights into brain function, as demonstrated by a detailed study of the brain activity and interactions elicited by a facial recognition task in healthy and schizophrenic subjects. [ABSTRACT FROM AUTHOR]