Reconstruction of simulated functional brain images using generalised matrix inverses and algebraic reconstruction techniques

It is widely believed that Event-Related Potentials measured at the scalp include useful information about the underlying neural mechanisms related to stimuli processing, attention and memory. For this information to be extracted the inverse problem must be solved in order to compute the intracranial sources responsible for the externally measured potential quantities. Much work has been dedicated to localized dipole computation [1] whereas a new technique has been proposed for computing images of distributed cortex activity, producing interesting clinical results in the field of Psychiatry [2,3]. The purpose of the present work was to study comparatively the performance of iterative techniques known as Algebraic Reconstruction Techniques and of the Generalized Matrix Inversion Technique [4] in inverting simulated ERP data.