EEG and MEG Source Analysis of Somatosensory Evoked Responses to Mechanical Stimulation of the Fingers

Abstract

Previous MEG and EEG studies have successfully utilized electrical stimulation of the digits in order to study generators of early responses in the human primary somatosensory cortex [1–3]. Natural stimulation of the somatosensory system using vibratory pulses to the tips of the digits has been found to elicit qualitatively different responses in the EEG [4, 5] presumably due to the activation of different neuronal pathways. Similar responses have been observed in the MEG to transient mechanical stimuli producing large responses at latencies of 50 msec, corresponding to the electrical P50 response [6, 7]. A more recent study [8] also noted MEG responses at 70 msec latencies in some subjects resembling the N70 component described by Hamalainen and co-workers [4]. The earlier, P50 response is presumed to reflect activation of primary sensory cortex (SI). A study of mechanically evoked epidural and single unit responses in waking monkeys [5] found that this component was associated with a period of inhibitory input to neurons in areas 3b and 1, whereas the later, slow component (N70) was not associated with activity in SI and appeared to arise from other cortical areas, such as SII. The magnetically recorded P50 (P50m) appears to be the largest and most consistent MEG event recorded during transient tactile stimulation in humans and its potential application for somatotopic mapping studies [6, 7] warrants further investigation of its neural generation. Moreover, the marked orthogonality of the EEG and MEG topographies of the P50 response makes it an ideal candidate for the comparison and/or combination of EEG and MEG localization methods. The current study compared separate high-density, 32 channel EEG and 143 channel MEG recordings in two subjects using identical stimulation paradigms in order to compare dipole source locations in somatosensory cortex obtained separately for each method. In addition, 3-dimensional MRI was obtained for both subjects in order to constrain source model information as well as aid in the integration of coordinate systems for the MEG and EEG source models.