G.L. Romani

Analysis of interhemispheric asymmetries of somatosensory evoked magnetic fields to right and left median nerve stimulation.

This paper represents the first neuromagnetic systematic investigation of the asymmetries between the sources activated in the right and left hemispheres after electric median nerve stimulation. We focused our attention on the location and strength of the equivalent sources activated in the primary somatosensory cortex contralateral to the stimulated nerve in the 50 msec post-stimulus epoch. The spatial coordinates of the equivalent sources did not differ statistically significantly in the two hemispheres. Minor individual asymmetries are shown to be related to the interhemispheric differences in the position of the central sulcus as revealed by MRI investigation. The equivalent sources were significantly stronger in the left hemisphere. When comparing the location of the generators across individuals, we show that interhemispheric differences fluctuate less than absolute values. A quantitative evaluation of these findings is also given. Based on these results, a normative data set has been established, to be used as a baseline in following up changes of interhemispheric asymmetries due to hemispheric lesions and subsequent cortical reorganization.

Neuromagnetic somatosensory homunculus: A non-invasive approach in humans

The somatosensory homunculus has been identified during stimulation of median (at wrist and elbow), femoral, tibial and pudendal nerves of the left hemibody via the neuromagnetic imaging technique. The somatic representations of different body districts have been localized in the somatosensory cortex, by means of an equivalent dipole localization algorhythm. Dipole locations agree with the well-known somatotopic organization obtained with invasive techniques. The proposed method is, therefore, an important investigating tool for studies on normal and diseased subjects.

NEUROMAGNETIC LOCALIZATION OF THE LATE COMPONENT OF THE CONTINGENT NEGATIVE VARIATION

The contingent negative variation (CNV) in a warned choice reaction time task was studied in 24 healthy subjects by use of magnetoencephalography (MEG). Special interest was focused on the late component of the CNV, CNVL. Source localization of the magnetically recorded CNVL, mCNVL was performed on 13 subjects, selected on the basis of the strength and stationarity of the electrically recorded CNV, eCNVL. To achieve whole head mapping, up to 500 epochs from different scalp positions were recorded, including a pretrial learning period of 40 epochs. The neuromagnetic signals studied in this experimental protocol are thus related to neurological processes that are present after an initial learning period has occurred. In 11 subjects, a goodness of fit between 88% and 95% was achieved using a two-dipole model with one equivalent source localized close to the precentral cortex contralateral to the side of movement, at mean a depth of 30 mm. Estimates of ipsilateral equivalent sources were less consistent across subjects. In 9 subjects the estimated ipsilateral sources were located symmetrically to the contralateral source. The results of this study suggest that the dominant source of the mCNVL is located near the bottom of the sulcus precentralis at the anterior bank of the gyrus precentralis, close to the sulcus frontalis superior. This supports previous findings that the CNVL is closely related to the readiness potential, and that the major cortical activity is symmetrically located in the left and right premotor areas.

Neuromagnetic topography of photoconvulsive response in man

The neuromagnetic method was applied to the study of photoconvulsive responses. The identification of specific magnetic field distributions over the scalp was achieved by; (a) a stimulation paradigm consisting of series of trains of flicker stimuli randomly presented to the epileptic patient, after eye closure, to get epileptic responses while avoiding seizures; (b) a novel procedure for data analysis, to select consistent responses. These patterns, when sufficiently stable in time and dipolar in shape, were used for source localization in the usual biomagnetic framework of the equivalent current dipole source representation. The results of this approach suggest that different specific cortical areas are repeatedly and randomly activated, involving mainly the frontal, occipital and temporal areas, often with a hemispheric prevalence.

Neuromagnetic imaging studies discriminate proprioceptive and cutaneous cortical inputs during median nerve stimulation in man

The localizing capabilities of the neuromagnetic imaging technique have been used to non-invasively discriminate mainly proprioceptive from cutaneous afferences to the contralateral hemisphere in human volunteers during separate median nerve stimulation at wrist and digits. The high time-resolution achieved in the localization of the equivalent sources of the early latency (15-30 ms) evoked fields permitted to follow their apparent movement as represented by an early and deep dipole (15 ms, about 5 cm deep, probably a subcortical source), a shallower one (20 ms, about 1.5 cm), and an intermediate one (23 ms, about 3 cm). These sources are supposed to be the manifestation of time-lagged parallel proprioceptive and cutaneous afferences from the thalamic ventro basal complex to different primary sensory areas in the postcentral gyrus. Both deeper and shallower dipoles could not be identified during selective stimulation of cutaneous afferents from the digital branches of the median nerve.