One effect of a maximum possible current strength

摘要

One of the difficulties encountered when trying to interpret MEG/EEG experimental data is that the spatial extent of the neuronal (primary) current sources is often not known. For some classes of data it is thought that the sources are fairly well localized (focal); while for others they are expected to be distributed over a significant area or volume. In the former case, if the regions of activity are sufficiently small they can be well approximated by point current dipoles. If these are few in number there may be only one, or just a small set of acceptable solutions for the dipole parameters. At the other extreme one assumes that current is present everywhere in the source space. Even with anatomical constraints taken into account, however, this leads to a highly underdetermined problem. For example, at 1 cm resolution on the cortex it would take (approximately) 2,000 dipole moments to completely specify the current, and this is much larger than the number of measurements that can be made. Only by imposing physiological constraints can one hope to make this distributed source approach be of any use as a way to determine the actual current. This study examines the extent to which the ambiguity of a distributed source solution is reduced by imposing one specific constraint: a maximum possible value for the dipole moment per unit volume, or per unit area of cortex. There is experimental evidence in support of this limit.