Reappraisal of field dynamics of motor cortex during self-paced finger movements

Abstract

Background: The exact origin of neuronal responses in the human sensorimotor cortex subserving the generation of voluntary movements remains unclear, despite the presence of characteristic but robust waveforms in the records of electroencephalography or magnetoencephalography (MEG).

Aims: To clarify this fundamental and important problem, we analyzed MEG in more detail using a multidipole model during pulsatile extension of the index finger, and made some important new findings.

Results: Movement-related cerebral fields (MRCFs) were confirmed over the sensorimotor region contralateral to the movement, consisting of a temporal succession of the first premovement component termed motor field, followed by two or three postmovement components termed movement evoked fields. A source analysis was applied to separately model each of these field components. Equivalent current diploes of all components of MRCFs were estimated to be located in the same precentral motor region, and did not differ with respect to their locations and orientations. The somatosensory evoked fields following median nerve stimulation were used to validate these findings through comparisons of the location and orientation of composite sources with those specified in MRCFs. The sources for the earliest components were evoked in Brodmann's area 3b located lateral to the sources of MRCFs, and those for subsequent components in area 5 and the secondary somatosensory area were located posterior to and inferior to the sources of MRCFs, respectively. Another component peaking at a comparable latency with the area 3b source was identified in the precentral motor region where all sources of MRCFs were located.

Conclusion: These results suggest that the MRCF waveform reflects a series of responses originating in the precentral motor area.

Keywords: Diploe sources; magnetoencephalography; motor cortex; movement-related cerebral fields; somatosensory evoked fields.

Figure 1

Movement-related cerebral fields following pulsatile extension of the index finger. Data from a representative subject. (A) Superimposed waveforms of all the channels without (a) and with low cut filtering (b). For the latter, the corresponding global field power (GFP) curve is shown in c, in which prominent peaks for specifying motor field (MF) and MEFI–III components are indicated by blue, red, green, and magenta arrowheads, respectively. (B) Six snapshots (a–f) of isocontour maps of evoked magnetic fields in the left hemisphere contralateral to the movement at chosen peaks are indicated by arrowheads in the GFP curve in A-c. The negative values in panels a-b and positive values in panels c-f indicate latency before (−) and after (+) the movement, respectively. Arrows in panels b–e show the location and orientation of estimated equivalent current dipoles (ECDs).

Figure 2

Spatiotemporal characteristics of source response modeled for movement-related cerebral fields (MRCFs). (A) Superpositions of four dipole sources (smf, sm1–sm3) on an MR image in posterior/superior oblique view. (B) The same superpositions of four dipoles sources in the sagittal plane view through the motor cortex region in the left hemisphere. Note that three or four plots seem to locate in nearly the same position in the posterior crown (smf, sm1, and sm2) in A or the wall of the precentral gyrus (smf, sm1–sm3) in B. (C) Comparison of the time course of source strength among four different dipole sources, smf (blue), sm1 (red), sm2 (green), and sm3 (magenta).

Figure 3

Spatial locations and orientations of four sources in the movement-related cerebral fields (MRCFs). (A) Plots for the locations of four independent sources (smf, sm1–sm3) in MRCFs in all subjects, in horizontal (a), sagittal (b), and coronal (c) planes of magnetoencephalography (MEG) coordinates. In each source, an ellipse represents a 95% confidence limit (z-value: 1.96) for plots in each of three orthogonal planes. (B) Orientation of four sources, averaged separately, in three orthogonal planes, corresponding to the sources in the left panels.

Figure 4

Temporal relationship between movement-related cerebral fields (MRCFs) and EMGs. (A) MRCF waveform modeled based on the parameters of smf, averaged across subjects. Black and gray lines represent unfiltered and filtered responses, respectively. The vertical line a represents the onset time of readiness fields, and peak latency for motor field (MF), MEFI, MEFII, and MEFIII are indicated in alphabetical order from b to e, respectively. The onset latency of the readiness field and peak latencies of MEFs were derived from the records of unfiltered and filtered MEG responses, respectively. (B) Rectified EMG signals recorded from two agonist muscles (ago1 and ago2) and one antagonist muscle (ant), both of which are time locked to the trigger pulse, and averaged across subjects. (C) Temporal relationship between the onset latency of activation of three muscles and readiness fields (a), and peak latency for MF (b), MEFI (c), MEFII (d), and MEFIII (e).

Figure 5

Grand-averaged source strength waveforms for cortical sources in somatosensory evoked fields (SEFs). From top to bottom panels, solid line represents averaged waveforms for s3b, s1/4, s5, sSIIi, and sSIIc across subjects are shown, whereas the thin lines represents ±SD. In each panel, a closed arrowhead indicates the peak time of initial source activities. Note in panel of s1/4, two additional peaks are indicated by gray arrowheads.

Figure 6

Simultaneous representation for spatial locations and orientations of four independent sources in somatosensory evoked fields (SEFs) and smf in MRCFs. (A) Plots for locations of all sources in SEFs (i.e., s1/4, s5, sSIIi, and sSIIc) and of smf are presented relative to the positions of s3b (defined as the cross-point of vertical and horizontal lines in the figure) in three orthogonal planes; from top to bottom, horizontal (a), sagittal (b), and coronal (c) planes. In each panel, an ellipse represents a 95% confidence limit (z-value: 1.96) for corresponding plots of source location in each of three orthogonal planes. (B) Grand-averaged source locations and orientations of four independent sources in SEFs and smf. The same sources to those depicted in A are averaged and depicted. In each source, the orientation of the current is represented by the line segment embedded in the plot for the location of the corresponding source. In three anatomical planes, source orientations are represented as their projection on three orthogonal planes, that is, horizontal (xy), sagittal (xz), and coronal (yz) plane, in panels a to c, respectively.