The inverse problem in MEG studies: an instrumental and analytical perspective