Can a Minimum-Norm Least-Squares Technique Help Map the Time Course of the Electrical Sources in the Brain Imaged by Functional MRI?

摘要

Functional Magnetic Resonance Imaging (fMRI), a new non-invasive modality for imaging brain function, provides millimeter spatial information about the changes in cortical metabolism and cerebral perfusion, which accompany the changes in neuronal activity. These neuronal changes occur on the order of milliseconds. However, the images produced by fMRI yield the activity occurring over several seconds (a result of biological constraints). At this time there is no way to identify when each source is active. The methodology of Evoked Response Potentials (ERP) provides millisecond information about the neuronal sources. A minimum-norm least squares technique is applied to ERP data to generate potential topographies inside the head at specific time points. These "cortical" topographies are compared to analogous fMRI data in order to map the time course of the neural sources depicted by fMRI.