Estimation of neuromagnetic sources using a REwheighted, STandardized, Recursively Iterated Cerebral Tomography Algorithm (RESTRICTA)

One of the biggest challenges in neuroimaging is to localize as accurately and precisely as possible the brain activity of subjects in vivo during the execution of specific tasks. In fact, the better the localization of areas involved in the task, the better the chances for neurologists / psychologists to understand the correct physiology of the brain. In this paper we proposed a new algorithm of resolution of the inverse problem of MagnetoEncephaloGraphy (MEG): RESTRICTA. It is based on the application of a recursive minimum norm solver, properly standardized, regularized in ℓ_2,1 norm. It is shown how it is able to provide more accurate estimates in terms both of localization and of extension of the activation area compared to other algorithms of common use as eLORETA.