Hannu J. Aronen

　Department of Radiology and BioMag Laboratory, Helsinki

　University Central Hospital, P.O. Box 381, FIN-00029 HYKS,Finland,

　hannu.aronen@huch.fi

Multichannel MEG measurements have been traditionally combined with　conventional MRI in order to localize the neuronal events during a given　task. However, the MEG inverse problem without any a priori constraints has　no unique solution. When combining the high spatial resolution of fMRI with　the high temporal resolution of MEG it is possible to study the living　brain with high accuracy. In an effort to combine fMRI to help solve the　MEG inverse problem, we have compared MEG and fMRI localization when　studying the function of the somatosensory system (Korvenoja et al., Hum　Brain Mapp 1999;8:13-27) and the visual system (Ahlfors et al., J　Neurophysiol 1999;82:2545-2555). In these studies, we have first compared　independently MEG and fMRI findings when using similar kind of stimuli at　both modalities. We also have successfully used fMRI data to constrain the　possible MEG source areas in the brain.

By combining MEG with conventional MRI, we have studied recovery-induced　changes in the responsiveness of the primary somatosensory cortex (SI) in　stroke patients with sensory and/or motor symptoms. When comparing the　measurements at the acute phase with those three months after the onset of　the symptoms, clinical improvement was significantly correlated with the　increase of P1m response recorded by MEG. We conclude that the recovery of　somatosensory stroke can be monitored by this kind of MEG techniques　(Wikstrom et al., Ann Neurol. In press.).