- Email: [email protected]
Localization of sensorimotor cortical rhythms induced by tactile stimulation using spatially filtered MEG
www.elsevier.com/locate/ynimg NeuroImage 30 (2006) 899 – 908
Localization of sensorimotor cortical rhythms induced by tactile stimulation using spatially filtered MEG William Gaetz* and Douglas Cheyne Department of Diagnostic Imaging, Neuromagnetic Imaging Laboratory, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada M5G 1X8 Received 5 April 2005; revised 19 September 2005; accepted 5 October 2005 Available online 2 December 2005 We applied the synthetic aperture magnetometry (SAM) spatial filtering method to localize sensorimotor mu (8 – 14 Hz) and beta (15 – 35 Hz) rhythms following tactile (brush) stimulation. Neuromagnetic activity was recorded from 10 adult subjects. Transient brush stimuli were applied separately to the right index finger, medial right toe and lower right lip. Differential images of mu and beta band source power were created for periods during (event-related desynchronization; ERD) or following (event-related synchronization; ERS) tactile stimulation, relative to prestimulus baseline activity. Mu ERD to finger brushing was localized to the contralateral somatosensory cortex and was organized somatotopically. Mu ERS, however, was not consistently observed for each subject. Beta ERD was consistently localized to sensory cortical areas and organized somatotopically in the postcentral gyrus (SI), and beta ERS was observed to be organized motorotopically in the precentral gyrus (MI). Longer duration (2 – 3 s) stimulation of the index finger also produced beta ERS in the primary motor cortex, and its time course demonstrated that these oscillatory changes are an off-response to the termination of the presented sensory stimulus. Interestingly, lip and toe stimulation also produced poststimulus increases in beta rhythms in the bilateral motor hand areas for all subjects, suggesting that common neural systems in the primary motor cortex are activated during tactile stimulation of different body regions. D 2005 Elsevier Inc. All rights reserved.
Introduction Early reports measuring cortical (Jasper and Penfield, 1949) and scalp (Gastaut et al., 1952) recorded brain activity described changes in EEG rhythms accompanying the preparation and performance of voluntary movement. More recently, a number of neuroelectric (EEG) and neuromagnetic (MEG) experiments have demonstrated that voluntary movement is preceded by a reduction in mu (8 – 15 Hz) and beta (15 – 35 Hz) power originating over
* Corresponding author. Fax: +1 416 813 7362. E-mail address: [email protected] (W. Gaetz). Available online on ScienceDirect (www.sciencedirect.com). 1053-8119/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2005.10.009
contralateral sensorimotor areas about 1.5 s prior to movement initiation (see Pfurtscheller and Lopes da Silva (1999) for review). Upon movement termination, mu power recovers slowly and with considerable variability between subjects (Salmelin et al., 1995; Salmelin and Hari, 1994b). However, beta power consistently returns to or exceeds baseline levels, a phenomenon termed beta rebound typically beginning within a half second of movement onset and persisting for several hundred milliseconds (Pfurtscheller et al., 1996). The neurophysiological mechanisms producing these oscillations are poorly understood (Lopes da Silva, 1991), and little is known about their functional significance (Fetz et al., 2000; MacKay, 1997). Pfurtscheller and Aranibar (1979) were first to refer to movement-related decreases in oscillatory EEG power as an event-related desynchronization (ERD) of neural populations. Later, experiments measuring synchrony in primate local field potentials (LFP) from sensorimotor cortex (Murthy and Fetz, 1996a,b) supported the inference that ongoing (resting-state) synchronous cortical oscillations are interrupted (i.e., desynchronized) by activating neurons involved in preparing and performing movement. These observations are further supported by experiments showing that motor imagery alone can produce beta ERD (Pfurtscheller et al., 2005; Schnitzler et al., 1997). Thus, it is generally assumed that movement-related desynchrony (ERD) represents a period of motor cortex activity involved in movement preparation and performance, whereas movement-related synchrony (beta ERS) represents a period of reduced cortical activity (idling) (Pfurtscheller et al., 1996) or inhibition of these same motor cortical areas (Chen et al., 1998; Salmelin et al., 1995) as a consequence of their prior activity. One problem facing experiments designed to study how voluntary movement can change sensorimotor rhythms relates to the difficulty in controlling for proprioceptive or haptic feedback which typically accompanies normal ballistic motor behavior (Kristeva et al., 1991). Indeed, recent experiments have also shown beta oscillations to be sensitive to somatosensory stimulation, such as median nerve stimulation (Salenius et al., 1997) and passive movement (Alegre et al., 2002; Cassim et al., 2001). For example, Cassim et al. (2001) showed that beta ERD/ERS could be
900
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
elicited by both active and passive finger movements and could be abolished by temporary deafferentation (i.e., digit ischemia). More recently, Cheyne et al. (2003) reported findings of sensorimotor ERD/ERS following tactile (brush) stimulation of the fingers using a novel spatial filtering method called synthetic aperture magnetometry (SAM) (Robinson and Vrba, 1999; Vrba and Robinson, 2001). These observations raise the interesting possibility that sensory stimulation alone may be sufficient to change the ongoing activity of beta oscillatory networks in motor cortex. In the present experiment, we applied the SAM spatial filtering method to localize cortical oscillations following tactile stimulation. By systematically varying the location and duration of stimulation, we sought to determine whether beta ERS is somatotopically organized and to observe the correspondence between stimulus duration and the time course of the observed changes in cortical oscillatory power. These experiments were performed in order to question whether previous reports of movement-related beta ERS may be explained by uncontrolled sensory events and thus test the assumption that motor cortical beta ERS is functionally linked to prior motor cortex activity (ERD).
All brush stimuli were presented with a soft bristled toothbrush and presented with light pressure making contact with an area of the skin approximately 1.5 cm2. A fiber optic light trigger was positioned in the path of the brush motion such that the interruption of the light path by the brush produced a TTL trigger when the brush made contact with the skin surface (‘‘brush-on’’) and when the brush departed from the skin surface (‘‘brush-off’’). An experimenter sat next to the subject and lightly brushed the subject’s skin at random intervals separated by several seconds. Subjects were required to fixate on a centrally positioned visual cue throughout the MEG recording session. A visual barrier was placed in the subject’s right visual field so that all motion of the experimenter, and the presented brush stimuli was not visible to the subject.
Methods
Sustained brush condition Sustained brushing was applied to the index finger only. These stimuli were applied in an identical manner to the transient condition except that stimulation consisted of bi-directional brushing for a period of approximately 3 s (10 s ISI). One sustained brush stimulus occurred for each of 100 trials (8-s duration; 2-s pretrigger period; ¨3-s brushing). In order to confirm the reliability of our source localizations, all brush stimuli were preceded by activation of the right and left median nerves. Median nerve stimuli were applied separately with a constant current, square wave pulses (3.1 Hz, 0.2-ms duration) delivered transcutaneously at the wrist above motor threshold. Somatosensory-evoked fields (SEFs) were recorded at a sample rate of 2500 Hz (bandpass DC to 800 Hz) with 600 epochs of 200ms duration.
Subjects Ten subjects (5 male, all right handed; mean age 31.5 years) participated in this experiment after receiving informed consent using protocols approved by the Hospital for Sick Children Research Ethics Board. Subjects were asked to sit upright in a comfortable chair with eyes open in a magnetically shielded room. Each subject was fitted with three fiducial localization coils placed at the nasion and preauricular points in order to localize the position of the subject’s head relative to the MEG sensors. Digital photographs of coil placement aided the coregistration of MEG data to each subject’s magnetic resonance image (MRI).
Transient brush condition Single transient brush stimuli consisted of a unidirectional brush motion of approximately 100-ms duration applied separately to the distal ventral phalanx of the right index finger and right medial toe, as well as the right half of the lower lip (¨6 s mean ISI). Each transient brush occurred once for each of 100 trials (5-s duration; 2-s prestimulus).
MEG recordings SAM analysis Neuromagnetic activity was recorded using a whole head 151 channel MEG system (Omega-151; VSM MedTech Ltd., Vancouver, Canada) with a sample rate of 625 samples/s, and a bandpass of 0 to 200 Hz. MEG data collection was synchronized to the interruption of a fiber optic light switch (Omron, 3EX-A41) by movement of the brush across the skin surface. Stimuli Prior to finger brushing, the subject’s hand was immobilized on an armrest in a comfortable position (palm up), and the proximal and middle phalanges of the right index finger were taped to the armrest to minimize inadvertent joint movements during brushing. Lip stimulation required that the fiber optic cables were adhered to the subject’s cheek, and positioned to detect brush stimuli presented to the lateral extent of the right lower lip. Finally, toe stimulation was accomplished by placing the bare right foot on an elevated support (approximately 30 cm) which included an angled sponge. The subject’s foot was positioned to rest against the sponge, with the toes extending out. The fiber optic cables were positioned at the distal glabrous surface of the medial right toe.
Synthetic aperture magnetometry (SAM) is a spatial filtering approach to source reconstruction, that has been successfully applied to the localization of cortical oscillations without the need for time-averaging of stimulus locked responses (Ishii et al., 1999; Singh et al., 2002; Taniguchi et al., 2000; Gaetz and Cheyne, 2003). This is achieved by integrating power changes over a defined time window relative to baseline activity using a differential power estimate (pseudo-t) (Robinson and Vrba, 1999). These differential images of source power can be calculated using single-trial data over selected frequency bands and time windows throughout the brain resulting in 3-dimensional source images co-registered with each subject’s head frame of reference. The application of this method to the localization of sensorimotor cortex oscillations has been described previously (Cheyne et al., 2003). T1-weighted structural MR images (3D SPGR) were obtained for each subject using a 1.5 T Signa Advantage system (GE Medical Systems, Milwaukee, USA). SAM source images were co-registered with the MR by matching the location of the localization coils used to define the MEG coordinate system on the MRI. In addition, dipole fits were performed on the averaged
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
response to median nerve stimulation at the latency of the M20. The location of the M20 dipoles could be overlaid on the same MR image and were inspected for each subject to confirm the location in the postcentral gyrus of the contralateral hemisphere. Pseudo-t images were computed for mu (8 – 14 Hz) and beta (15 – 35 Hz) frequency bands over the entire brain at 2-mm resolution. For each trial, stimulus event markers were placed (offline) which identified the fiber-optic stimulus events for Fbrush-on_ and Fbrush-off_. The control window for each brush event was 1.75 to 0.75 s before the onset of each brush (‘‘brush-on’’). The initial active window was 0.5 to 1.5 s following each brush (‘‘brush-off’’), however, the onset time was adjusted on the basis of virtual sensor and time – frequency plot information to optimally capture the observed changes in frequency power (see time – frequency methods below).
901
T1-weighted MR images and applied to the SAM ERD/ERS images. The normalized SAM images were then resampled to produce 2 2 2-mm resolution images of source power in standardized stereotaxic space and averaged across subjects. Due to large variations between the peak magnitude of ERD/ERS power, SAM images for each subject were normalized to the largest (peak value) in the image prior to spatial normalization. Statistical thresholds were calculated using a non-parametric statistical permutation test (SnPM) of the mean voxel value (Nichols and Holmes, 2002). Mean stereotaxic coordinates were calculated from the most significant activated voxels (statistical peak values) within each cluster. The MNI coordinates were converted to Talairach and Tournoux (1988) coordinates using a non-linear transformation (Lancaster et al., 2000) implemented in mri3dX software package (http://www.aston.ac.uk/lhs/staff/ singhkd/mri3dX/mri3dX.jsp).
Virtual sensors and time – frequency plots In addition to the ability to generate volumetric images of source power changes over defined time intervals, the time course of oscillatory activity may be observed for any location of interest by observing wideband (e.g., 1 – 50 Hz) output from the spatial filter for a target location. This Fvirtual sensor_ time series reflects millisecond by millisecond changes in oscillation power for the target location in the single trial data. A time – frequency analysis of virtual sensor output for arbitrary locations in the brain was applied to the unaveraged data using a wavelet based technique (Tallon-Baudry et al., 1997). This produced a time – frequency (TF) plot that represented changes in the total power of both phase locked and non-phase locked oscillations over the duration of the recording epoch. Virtual sensor based TF plots were observed for locations showing peak activity in the SAM images. This information also allowed us to optimize the placement of the active window (1-s duration) to be aligned to the onset of beta ERS. Differential SAM images of mu and beta band source power were then recomputed for the identified periods of ERD/ERS relative to premovement baseline activity. Virtual sensors and frequency (FFT) Previous work by Neuper and Pfurtscheller (2001) demonstrated that the specific frequency of beta rhythm ERS was lower for hand (¨20 Hz) than foot (¨30 Hz) movements. In order to address whether the specific frequency of beta ERS could also be predicted by presenting tactile stimulation to different body areas, an average of the Fast Fourier Transformed (FFT) virtual sensor single trial data was calculated for all images showing peak SAM activity. The beta ERS frequency corresponding to the largest FFT power was then noted for each peak location observed within a given SAM image, with mean values calculated for all subjects and brush locations. Group averaging and statistical thresholds Group averaging of the SAM ERD/ERS images was performed in order to determine group effects. Methods for group averaging of SAM functional images were first described by Singh et al. (2003). Spatial normalization to the MNI (T1) template was carried out using SPM99 (http://www.fil.ion.ac.uk/spm) implemented in MATLAB (http://www.mathworks.com/). Both linear and nonlinear warping parameters were obtained from individual subject’s
Results Transient brush stimuli Beta band ERS showed a consistent, robust stimulus-locked activity for each subject and brush condition. Inspection of singletrial MEG waveforms from sensors overlying the sensorimotor areas in individual subjects showed that the changes in beta ERD/ ERS could be observed within single trials, and that the beta ERS appeared time locked to the termination of the brush stimulus (see Fig. 1A). SAM analysis of beta band ERS following the transient brush of the index finger revealed consistent increases in beta power following brush stimuli which were localized to the motor cortical hand area bilaterally and was largest contralateral to side of stimulation in 7 out of 10 subjects. Time – frequency analysis of SAM virtual sensors from the peak location of the contralateral ERS showed a consistent decrease in beta band power coincident with the brush contacting the skin surface, followed by a rebound in beta band power starting at approximately 300 – 400 ms following the cessation of stimulation. Results from a representative subject are shown in Fig. 1. Responses to the transient brush of the right lower lip resulted in beta ERS motorotopic activation of contralateral lip area (MI) observed in all subjects as well as bilateral motor cortex hand area activity (larger pseudo-t values contralaterally in n = 5) (Fig. 1D). Transient brush stimuli of the right medial toe also produced a motorotopic activation of contralateral toe area (MI) observed in 8 out of 10 subjects (Fig. 1E) as well as bilateral motor cortex activity for each subject (larger pseudo-t value contralaterally in N = 8). Consistent with previous observations (Salmelin et al., 1995; Salmelin and Hari, 1994a), mu power ERS showed considerable variability and was not observed for all subjects. Consequently, post-stimulus changes in mu power (mu ERS) were not analyzed further. Grand averaged beta ERS Group averaged SAM images of beta ERS following the transient finger brush showed a significant activation of bilateral motor cortical hand area. Beta ERS following the transient lip brush produced a similar bilateral activation of motor cortical hand area as well as a significant contralateral activation of motor
902
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
cortical lip area lateral to the hand area. Grand averages of beta ERS following transient toe stimulation produced contralateral activation of the motor cortex hand area as well as significant activation of contralateral motor cortical toe area on the medial aspect of the precentral gyrus. All activations were significant in the non-parametric permutation test (SnPM; P < 0.05) shown in Fig. 2.
FFT analysis of ERS using virtual sensors Mean FFT values based on the virtual sensor single trials (FFT window 1-s duration; onset of beta ERS) were significantly higher in frequency for toe brush (26.01 Hz) than finger (20.99 Hz) (paired t test; P < 0.01, Bonferroni corrected) (Fig. 3A). FFT power was also compared for each of the peak frequencies
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
903
Fig. 2. Spatially normalized statistically thresholded grand averages for each brush condition (beta ERS) are shown coregistered to a template brain. Normalized SAM images were permuted using the mean. Co-active sources in motor cortex are observed for each brush condition (SnPM P < 0.05). Bilateral hand area activation was observed for each brush location, with the exception of the toe brush condition. Ipsilateral hand area activity was observed for individual subjects (toe brush) but was relatively weaker than contralateral activity and did not reach significance using SnPM.
observed in the active window. Beta ERS was observed to be relatively larger for the contralateral motor area MI-C, relative to the ipsilateral hand area MI-I for each brush location. The average hand area beta ERS power ((MI-C + MI-I)/2) was largest following index finger stimulation relative to lip and toe stimulation (paired t test; P < 0.05, Bonferroni corrected) (see Fig. 3B). Sustained brush stimuli The application of a sustained brush to the index finger produced a prolonged decrease in mu and beta power for each
subject which continued until the sustained brush stimulus was removed. TF plots based on virtual sensor locations in contralateral MI hand area showed a distinct mu and beta ERD throughout the ¨3 second brush period, followed by beta ERS approximately 300 ms following the lifting of the brush from the skin surface (see Fig. 4). In each subject, motor cortical beta ERS was not observed until the sensory stimulus was terminated. Grand averaged SAM images were calculated for mu and beta suppression starting 1.75 s prior to the termination of the sustained brush (1-s duration). These averaged SAM activation images showed a significant decrease in mu and beta power localized to distinct areas of contralateral
Fig. 1. (A) Two MEG channels (MLF45 and MLC42) overlying contralateral sensorimotor cortex showing beta (15 – 35 Hz) band activity for two single trials (trials 50 and 74) selected to show examples of different time intervals between onset (brush-on) and termination (brush-off) of stimulation. Brush onset (time = 0 s) coincides with beta ERD, followed by beta ERS which occurs later for the longer duration transient brush (trial 50), suggesting that the beta ERS is temporally locked to the offset of the brush stimulus. In both traces, beta ERS shows a topographical pattern consistent with source activity in contra as well as ipsilateral sensorimotor cortical areas (lower panel, topographic map). Time – frequency and SAM images were based on MEG data realigned to the brush off events as time = 0 s. (B) SAM localization of beta ERS for index finger brushing displayed on 3D rendered image of the subject’s MRI. A cutaway through the axial plane at the level of the motor hand areas shows bilateral motor cortex activation anterior to localized N20m dipole locations (not shown). The SAM pseudo-t image was calculated from all single trial MEG responses based on the active (0.2 – 1.2 s) and control ( 1.75 to 0.75 s) time windows. (C) A time – frequency (TF) plot of a virtual sensor based on the location of the largest SAM activity (contralateral MI) observed in panel (B). Red TF values indicate increased source power during the active state, relative to a 1-s baseline period ( 2 s to 1 s). (D and E) SAM localized beta ERS activity following right lower lip brush stimulation (left) and medial toe (right) is shown. Bilateral motor cortex activity (motor hand knob) as well as body part specific (motorotopic) cortical activations was observed.
904
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
Fig. 3. SAM localized peak activity for each subject and brush condition was evaluated to determine the frequency corresponding to the largest FFT power value. The FFT was performed on the virtual sensor single trials. Panel A compares the frequency (averaged over subjects) observed for each SAM location. Data from two subjects were not included for the brush toe condition due to low SNR. Panel A shows the frequency of the motorotopic toe activation produces a higher beta frequency than that observed coincidentally from the motorotopic hand area (paired t test; P < 0.01 Bonferroni corrected). Pooled FFT power values ((MI-I + MI-C)/2) were observed to decrease from brush index, lip and toe (paired t test; P < 0.05 Bonferroni corrected) conditions with no corresponding change in frequency.
somatosensory cortex (SnPM; P < 0.05). Group averaged SAM images to the post brush ERS showed a corresponding increase in beta power which, similar to results following the transient brush, were observed maximally from the contralateral motor cortex hand area (see Fig. 4, SnPM results). However, the SnPM map for beta ERS does not include significant ipsilateral motor cortex activation in the group thresholded beta ERS map. Power differences between MI-C and MI-I following transient stimulation were not significant, however, MI-I was significantly weaker than MI-I following the termination of the sustained brush (paired t test; P < 0.05 Bonferroni corrected).
Discussion We observed that transient tactile stimulation to different areas of the body produced both specific and general effects on ongoing rhythmic activity in the beta frequency band. Tactile input to the finger, lip and toe resulted in suppression and strong rebound of beta band power observed from precentral gyrus areas associated with movements from these same body parts. Thus, the beta rebound accompanying passive somatosensory input appears to be organized according to the homuncular distribution of motor output neurons of the primary motor cortex or motorotopically. This may reflect the close relationship between afferent projections of the somatosensory system to regions of the precentral gyrus by corticocortical connections with the somatosensory cortex (Jones, 1983) or by direct input from the thalamus (Horne and Tracey, 1979; Lemon, 1981) and the somatotopic organization of motor output corticospinal neurons to the same body regions. Using time – frequency analysis, we observed that sustained stimulation suppressed mu and beta rhythms maximally in contralateral sensorimotor areas over the duration of stimulus input, and that the termination of the stimulus produced a corresponding rebound in beta power arising from contralateral motor cortex hand area time-locked to the cessation of stimulation. This indicates that beta ERS to a single somatosensory stimulus most likely represents an Foff-response_ that is independent of the duration of stimulus. This observation is consistent with the hypothesis that movement-related beta ERS observed in the motor cortex may reflect the termination of sensory feedback resulting
from either passive or active movements (Cassim et al., 2001) rather than the return to resting state of motor cortex neurons following the initiation and execution of a voluntary movement as proposed in idling hypotheses of cortical oscillations. The term ‘‘idling’’ was first used by Adrian and Matthews (1934) to describe large EEG alpha rhythms observed from visual areas in the absence of visual stimulation. Kuhlman (1978) and later Pfurtscheller et al. (1996) have since suggested that the sensorimotor cortex also exhibits rhythmic activity corresponding to an idling state following the termination of a movement (ERS). Support for this suggestion has come from a transcranial magnetic stimulation (TMS) experiment that showed that the timing of beta ERS coincides with reduced excitability of motor cortex neurons (Chen et al., 1998). However, Cassim et al. (2001) argued against the idling hypothesis as they found that beta synchronization could be produced by passive finger movements, an event where the motor cortex is not first desynchronized. Like Cassim et al. (2001), we have observed sensorimotor beta ERD/ERS following somatosensory stimulation with no corresponding motor planning or production. Co-activation of hand regions of the motor cortex One of the more interesting observations in the current study is the fact that beta ERS from bilateral motor cortex hand areas were co-activated during lip and toe brushing, yet showed a similar peak frequency to that observed during stimulation of the digit alone (¨20 Hz). The magnitude of beta ERS in the hand region was observed to be significantly weaker during co-activation, in comparison to when the index finger was actually stimulated. The observation of hand area activation for lip and toe stimulation would suggest that beta rhythm changes observed in the motor cortex, particularly in the hand region, may be due to cortical inhibition through indirect pathways, presumably as the result of engaging a more distributed sensorimotor network and without the need for input from the periphery. We cannot rule out the possibility that this co-activation reflected activity in more rostral portions of area 4, even incorporating part of area 6 premotor neurons that are associated with more cognitive aspects of motor control. This would be consistent with functional magnetic resonance imaging (fMRI) studies that have shown that more
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
905
Fig. 4. Time – frequency (TF) plots on the left correspond to responses from a single subject based on the observed virtual sensor peak locations for mu ERD, beta ERD and beta ERS bands respectively. Group averaged and statistically thresholded activity for the active and control windows on the left are shown to the right (N = 10). Mu and beta band suppression is observed until the tactile stimulus is terminated. This suppression was localized to different areas of contralateral somatosensory cortex (SnPM; P < 0.05). Beta ERS appears immediately following the termination of the tactile brush and appears to be an ‘‘off response’’ arising from contralateral motor cortex.
rostral portions of the precentral gyrus may be active during motor tasks, particularly for the ipsilateral hemisphere (Hanakawa et al., 2005). This ‘‘non-specific’’ bilateral activation of the hand region of MI may therefore reflect a more general feature of motor cortex oscillations in sensorimotor tasks, including the observation of bilateral ERS in the precentral gyrus during actual finger movements (Cheyne and Gaetz, 2003; Neuper and Pfurtscheller, 2001). In a recent study, Jensen et al. (2005) showed that beta oscillation power can be increased by the administration of the GABAergic agonist benzodiazepine. Subjects administered benzodiazepines
showed an overall increase in resting, eyes-closed MEG beta band power, observed maximally over bilateral sensorimotor regions close to the hand area (Jensen et al., 2005). Further evidence of the specialized nature of hand area representations has come from recent primate research showing that unlike hindlimb regions of MI, the forelimb regions contain specialized inputs from areas of the prefrontal lobes as reflected by an abundance of indirect (polysynaptic) connections between these areas (Miyachi et al., 2005). Taken together, these results suggest that the hand region of the precentral gyrus may contain neuronal populations that
906
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
subserve a more general role in sensorimotor control that is independent of somatotopy—an observation that is important to take into account when interpreting the functional role of sensorimotor cortex oscillations in studies that focus exclusively on movements or stimulation of the digits. Pfurtscheller and Neuper (1994) observed that mu band EEG responses following simple hand movements produced both hand area ERD and foot area ERS (Pfurtscheller and Neuper, 1994). This phenomenon, later termed focal ERD/surround ERS (Suffczynski et al., 1999) was hypothesized to be the consequence of a directed focus of attention onto a specific motor act, thus producing ERD, whereas the surrounding cortex is outside the focus of attention, resulting in surround ERS. In the present experiment, close inspection of each individual’s response to tactile brush stimulation for both mu and beta ERD, as well as the group averaged ERD responses, failed to show any specific ERS in the surrounding cortical regions. It should be noted that our method involved the selection of time windows for each subject and stimulus location to optimally measure ERS power and location and avoid temporal overlap between post-stimulus ERD and ERS. The fact that we did not observe coincident ERD and ERS in the present experiment may have been the result of this methodological distinction or may indicate that the focal ERD/surround ERS observed by Pfurtscheller and Neuper (1994) is specific to mu rhythm changes following self-paced motor tasks, which are known to recover slowly and with considerable intersubject variability (Salmelin et al., 1995; Salmelin and Hari, 1994b). The somatosensory stimulation in the current study comprised a relatively salient stimulus that was attended to by the subjects, who were instructed to remain still and hence avoid any motor response to the tactile stimulus. Therefore, our task may have engaged cortical networks that involve the primary motor cortex and reflect higher order sensory control of movement, including the possible inhibition of a motor response to the tactile stimulus (withdrawal reflex). The involvement of MI in such cortical networks may explain why the modulation of these beta rhythms in motor cortex is observed for a wide range of somatosensory and motor tasks including motor imagery (Schnitzler et al., 1997; Pfurtscheller et al., 2005) or even observation of another individual’s digit being stimulated (Cheyne et al., 2003). The latter raises the possibility that mechanisms related to selective spatial attention, particularly with respect to the ability to anticipate a somatosensory input (Spence et al., 2000) could have engaged motor preparatory systems that would also explain the occurrence of beta ERS in the precentral hand area. However, in the current study, the stimulation was randomly presented and hidden from the subject’s view in order to minimize such anticipatory effects. In addition, the fact that beta ERS was highly time locked to the offset, rather than the onset of the stimulation would indicate that the subject’s anticipation of the stimulus did not play a role in the generation of the time-locked beta ERS.
higher peak frequency than that arising from cortical hand areas for stimulation of the corresponding body areas. Most notably, the observed changes in bilateral hand areas during toe and lip stimulation reflected the preferred frequency for finger stimulation rather than that for the body region that was being stimulated. This would suggest that each specific area of the primary motor cortex exhibits a stable and characteristic peak frequency, whereas the observed amplitude modulation of synchronized oscillations was only dependent on the presence or absence of sensory input. The reasons for the frequency specificity of beta ERS observed in the present study is unclear and raises some interesting questions about the underlying neural populations involved in the oscillatory response and what role various stimulus parameters may play in its generation, such as the number of afferent fibers activated (total area of stimulation), duration, or force. It is generally assumed that rapidly oscillating cell assemblies are comprised of fewer neurons than cell assemblies showing relatively slower oscillation frequencies (Singer, 1993). Using a neural network model, Lopes da Silva et al. (1976) observed the effect of varying the number of inhibitory interneurons on simulated alpha rhythmic activity. These simulations showed that an increased area (number) of synchronous inhibitory neurons reduces the peak frequency and increases the amplitude of the observed oscillation (Lopes da Silva et al., 1976). Although we did not vary the total area of stimulation or measure differences in receptive field size in the current study, the well-known differences in cortical representations of the body regions stimulated (e.g., finger tip versus toe) may support the interpretation that the different frequencies observed may have reflected size of the activated neural populations for different body regions. However, this does not explain why we observed similar frequencies of beta ERS in the hand region for both finger stimulation and stimulation of other body regions. Current studies are underway to explore the relationship between stimulus parameters (such as receptive field size) and the frequency and amplitude of movement and somatosensory induced ERD/ERS. In the present experiment, we were able to show consistent patterns of beta ERS across subjects by group averaging SAM images that were spatially normalized to stereotaxic space. However, we noted that with this approach weaker areas of activation (e.g., ipsilateral MI changes during transient toe brushing) that could be observed in individual subject images and were observed as peaks in the unthresholded group averages, failed to reach statistical significance when using an omnibus permutation test for whole brain images (SnPM). This was likely due to the known bias of the permutation distribution by presence of strong sources (Chau et al., 2004) that can result in an overly conservative test that will fail to detect weaker sources of activity that are co-active with the stronger sources as statistically significant. Further work is needed on SnPM or other significance tests involving omnibus derived permutation distributions which may be differentially biased by source strength.
Frequency specificity of beta ERS Conclusions We observed that input to different body regions appears to induce a beta ERS with a preferred frequency, consistent with previous EEG studies. For example, Neuper and Pfurtscheller (2001) observed that ERS hand movements produced a lower beta ERS frequency than foot movements (below 20 Hz for hand and above 20 Hz for foot). Similarly, the present results clearly show that beta ERS arising from toe motor cortex has a significantly
We have localized time-locked changes in rhythmic activity of primary somatosensory and motor cortex associated with tactile stimuli in the absence of motor planning and execution. Consistent with our previous results (Cheyne et al., 2003), transient stimulation produces beta ERD followed by beta ERS localized to bilateral primary motor cortical areas in the precentral gyrus. In
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
the current study, we also showed somatotopically organized cortical beta ERS for lip and toe brushing within the precentral gyrus, which suggests that the post-stimulus ERS is functionally related to the location of peripherally delivered somatosensory stimulation of the skin and reflects the known homuncular organization of movements of different body regions (motorotopy) within the same cortical area. Indeed, beta ERS may reflect the ongoing coordination of sensory input and motor output maintained by continuous input from the periphery via somatosensory afferents that in turn directly or indirectly inhibit associated areas of motor cortex.
Acknowledgments This work was supported by the Canadian Institutes of Health Research (Grant 64279). The authors would like to thank Krish Singh for the valuable contributions to this work and for making available the mri3dX software package. References Adrian, E.D., Matthews, B.H., 1934. The Berger rhythm: potential changes from the occipital lobes in man. Brain 57, 355 – 385. Alegre, M., Labarga, A., Gurtubay, I.G., Iriarte, J., Malanda, A., Artieda, J., 2002. Beta electroencephalograph changes during passive movements: sensory afferences contribute to beta event-related desynchronization in humans. Neurosci. Lett. 331 (1), 29 – 32. Cassim, F., Monaca, C., Szurhaj, W., Bourriez, J.L., Defebvre, L., Derambure, P., Guieu, J.D., 2001. Does post-movement beta synchronization reflect an idling motor cortex? NeuroReport 12 (17), 3859 – 3863. Chau, W., McIntosh, A.R., Robinson, S.E., Schulz, M., Pantev, C., 2004. Improving permutation test power for group analysis of spatially filtered MEG data. NeuroImage 23 (3), 983 – 996. Chen, R., Yaseen, Z., Cohen, L.G., Hallett, M., 1998. Time course of corticospinal excitability in reaction time and self-paced movements. Ann. Neurol. 44 (3), 317 – 325. Cheyne, D., Gaetz, W., 2003. Neuromagnetic activation of oscillatory brain activity associated with voluntary finger and toe movements. New York. NeuroImage 19 (2), e2088R (Suppl.). Cheyne, D., Gaetz, W., Garnero, L., Lachaux, J.P., Ducorps, A., Schwartz, D., Varela, F.J., 2003. Neuromagnetic imaging of cortical oscillations accompanying tactile stimulation. Brain Res. Cogn. Brain Res. 17 (3), 599 – 611. Fetz, E.E., Chen, D., Murthy, V.N., Matsumura, M., 2000. Synaptic interactions mediating synchrony and oscillations in primate sensorimotor cortex. J. Physiol. Paris 94 (5 – 6), 323 – 331. Gaetz, W., Cheyne, D., 2003. Localization of human somatosensory cortex using spatially filtered magnetoencephalography. Neurosci. Lett. 340 (3), 161 – 164. Gastaut, H., Terzian, H., Gastaut, Y., 1952. E´tude d’une activite´ e´lectroence´phalographique me´connue: Fle rhythme rolandique en arceau_. Marseille Me´d. 89, 296 – 310. Hanakawa, T., Parikh, S., Bruno, M.K., Hallett, M., 2005. Finger and face representations in the ipsilateral precentral motor areas in humans. J. Neurophysiol. 93 (5), 2950 – 2958. Horne, M.K., Tracey, D.J., 1979. The afferents and projections of the ventroposterolateral thalamus in the monkey. Exp. Brain Res. 36 (1), 129 – 141. Ishii, R., Shinosaki, K., Ukai, S., Inouye, T., Ishihara, T., Yoshimine, T., Hirabuki, N., Asada, H., Kihara, T., Robinson, S.E., et al., 1999. Medial prefrontal cortex generates frontal midline theta rhythm. NeuroReport 10 (4), 675 – 679.
907
Jasper, H.H., Penfield, W., 1949. Electrocorticograms in man: effect of the voluntary movement upon the electrical activity of the precentral gyrus. Arch. Psychiatry Z. Neurol. 183, 163 – 174. Jensen, O., Goel, P., Kopell, N., Pohja, M., Hari, R., Ermentrout, B., 2005. On the human sensorimotor-cortex beta rhythm: sources and modeling. NeuroImage 26 (2), 347 – 355. Jones, E.G., 1983. The nature of the afferent pathways conveying shortlatency inputs to primate motor cortex. In: Desmedt, J.E. (Ed.), Motor Control Mechanisms in Health and Disease. Raven Press, New York, pp. 263 – 285. Kristeva, R., Cheyne, D., Deecke, L., 1991. Neuromagnetic fields accompanying unilateral and bilateral voluntary movements: topography and analysis of cortical sources. Electroencephalogr. Clin. Neurophysiol. 81 (4), 284 – 298. Kuhlman, W.N., 1978. Functional topography of the human mu rhythm. Electroencephalogr. Clin. Neurophysiol. 44, 83 – 93. Lancaster, J.L., Woldorff, M.G., Parsons, L.M., Liotti, M., Freitas, C.S., Rainey, L., Kochunov, P.V., Nickerson, D., Mikiten, S.A., Fox, P.T., 2000. Automated Talairach atlas labels for functional brain mapping. Hum. Brain Mapp. 10 (3), 120 – 131. Lemon, R.N., 1981. Functional properties of monkey motor cortex neurones receiving afferent input from the hand and fingers. J. Physiol. 311, 497 – 519. Lopes da Silva, F.H., 1991. Neural mechanisms underlying brain waves: from neural membranes to networks. Electroencephalogr. Clin. Neurophysiol. 79, 81 – 93. Lopes da Silva, F.H., van Rotterdam, A., Barts, P., van Heusden, E., Burr, W., 1976. Models of neuronal populations: the basic mechanisms of rhythmicity. In: Corner, M.A., Swaab, D.F. (Eds.), Perspectives in Brain Research, Progr. in Brain Res. Elsevier, Amsterdam, pp. 281 – 308. MacKay, W.A., 1997. Synchronized neuronal oscillations and their role in motor processes. Trends Cogn. Sci. 1 (5), 176 – 183. Miyachi, S., Lu, X., Inoue, S., Iwasaki, T., Koike, S., Nambu, A., Takada, M., 2005. Organization of multisynaptic inputs from prefrontal cortex to primary motor cortex as revealed by retrograde transneuronal transport of rabies virus. J. Neurosci. 25 (10), 2547 – 2556. Murthy, V.N., Fetz, E.E., 1996a. Oscillatory activity in sensorimotor cortex of awake monkeys: synchronization of local field potentials and relation to behavior. J. Neurophysiol. 76 (6), 3949 – 3967. Murthy, V.N., Fetz, E.E., 1996b. Synchronization of neurons during local field potential oscillations in sensorimotor cortex of awake monkeys. J. Neurophysiol. 76 (6), 3968 – 3982. Neuper, C., Pfurtscheller, G., 2001. Event-related dynamics of cortical rhythms: frequency-specific features and functional correlates. Int. J. Psychophysiol. 43 (1), 41 – 58. Nichols, T.E., Holmes, A.P., 2002. Nonparametric permutation tests for functional neuroimaging: a primer with examples. Hum. Brain Mapp. 15 (1), 1 – 25. Pfurtscheller, G., Aranibar, A., 1979. Evaluation of event-related desynchrozization (ERD) preceding and following voluntary self-paced movement. Electrocephalogr. Clin. Neurophysiol. 46 (2), 138 – 146. Pfurtscheller, G., Lopes da Silva, F.H.L., 1999. Event-related EEG/MEG synchronization and desynchronization: basic principles. Clin. Neurophysiol. 110 (11), 1842 – 1857. Pfurtscheller, G., Neuper, C., 1994. Event-related synchronization of mu rhythm in the EEG over the cortical hand area in man. Neurosci. Lett. 174 (1), 93 – 96. Pfurtscheller, G., Stancak Jr., A., Neuper, C., 1996. Post-movement beta synchronization. A correlate of an idling motor area? Electroencephalogr. Clin. Neurophysiol. 98 (4), 281 – 293. Pfurtscheller, G., Neuper, C., Brunner, C., da Silva, F.L., 2005. Beta rebound after different types of motor imagery in man. Neurosci. Lett. 378 (3), 156 – 159. Robinson, S.E., Vrba, J., 1999. Functional neuroimaging by synthetic aperture magnetometry (SAM). In: Nenonen, J., Ilmoniemi, R.J., Katila, T. (Eds.), Espoo. Helsinki University of Technology, pp. 302 – 305. Salenius, S., Schnitzler, A., Salmelin, R., Jousmaki, V., Hari, R., 1997.
908
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
Modulation of human cortical rolandic rhythms during natural sensorimotor tasks. NeuroImage 5 (3), 221 – 228. Salmelin, R., Hari, R., 1994a. Characterization of spontaneous MEG rhythms in healthy adults. Electroencephalogr. Clin. Neurophysiol. 91 (4), 237 – 248. Salmelin, R., Hari, R., 1994b. Spatiotemporal characteristics of sensorimotor neuromagnetic rhythms related to thumb movement. Neuroscience 60 (2), 537 – 550. Salmelin, R., Hamalainen, M., Kajola, M., Hari, R., 1995. Functional segregation of movement-related rhythmic activity in the human brain. NeuroImage 2 (4), 237 – 243. Schnitzler, A., Salenius, S., Salmelin, R., Jousmaki, V., Hari, R., 1997. Involvement of primary motor cortex in motor imagery: a neuromagnetic study. NeuroImage 6 (3), 201 – 208. Singer, W., 1993. Synchronization of cortical activity and its putative role in information processing and learning. Annu. Rev. Physiol. 55, 349 – 374. Singh, K.D., Barnes, G.R., Hillebrand, A., Forde, E.M., Williams, A.L., 2002. Task-related changes in cortical synchronization are spatially coincident with the hemodynamic response. NeuroImage 16 (1), 103 – 114. Singh, K.D., Barnes, G.R., Hillebrand, A., 2003. Group imaging of task-
related changes in cortical synchronisation using nonparametric permutation testing. NeuroImage 19 (4), 1589 – 1601. Spence, C., Pavani, F., Driver, J., 2000. Crossmodal links between vision and touch in covert endogenous spatial attention. J. Exp. Psychol. Hum. Percept. Perform. 26 (4), 1298 – 1319. Suffczynski, P., Pijn, J.P.M., Pfurtscheller, G., Lopes da Silva, F.H., 1999. Event-related dynamics of alpha band rhythms: a neuronal network model of focal ERD/surround ERS. In: Pfurtscheller, G., Lopes da Silva, F.H. (Eds.), Event-related Desynchronization. Rev. Ed. Elsevier, Amsterdam, pp. 67 – 85. Talairach, J., Tournoux, P., 1988. Co-Planar Stereotaxic Atlas of the Human Brain. Thieme, New York. Tallon-Baudry, C., Bertrand, O., Delpuech, C., Permier, J., 1997. Oscillatory gamma-band (30 – 70 Hz) activity induced by a visual search task in humans. J. Neurosci. 17 (2), 722 – 734. Taniguchi, M., Kato, A., Fujita, N., Hirata, M., Tanaka, H., Kihara, T., Ninomiya, H., Hirabuki, N., Nakamura, H., Robinson, S.E., et al., 2000. Movement-related desynchronization of the cerebral cortex studied with spatially filtered magnetoencephalography. NeuroImage 12 (3), 298 – 306. Vrba, J., Robinson, S.E., 2001. Signal processing in magnetoencephalography. Methods (Duluth) 25 (2), 249 – 271.
Localization of sensorimotor cortical rhythms induced by tactile stimulation using spatially filtered MEG William Gaetz* and Douglas Cheyne Department of Diagnostic Imaging, Neuromagnetic Imaging Laboratory, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada M5G 1X8 Received 5 April 2005; revised 19 September 2005; accepted 5 October 2005 Available online 2 December 2005 We applied the synthetic aperture magnetometry (SAM) spatial filtering method to localize sensorimotor mu (8 – 14 Hz) and beta (15 – 35 Hz) rhythms following tactile (brush) stimulation. Neuromagnetic activity was recorded from 10 adult subjects. Transient brush stimuli were applied separately to the right index finger, medial right toe and lower right lip. Differential images of mu and beta band source power were created for periods during (event-related desynchronization; ERD) or following (event-related synchronization; ERS) tactile stimulation, relative to prestimulus baseline activity. Mu ERD to finger brushing was localized to the contralateral somatosensory cortex and was organized somatotopically. Mu ERS, however, was not consistently observed for each subject. Beta ERD was consistently localized to sensory cortical areas and organized somatotopically in the postcentral gyrus (SI), and beta ERS was observed to be organized motorotopically in the precentral gyrus (MI). Longer duration (2 – 3 s) stimulation of the index finger also produced beta ERS in the primary motor cortex, and its time course demonstrated that these oscillatory changes are an off-response to the termination of the presented sensory stimulus. Interestingly, lip and toe stimulation also produced poststimulus increases in beta rhythms in the bilateral motor hand areas for all subjects, suggesting that common neural systems in the primary motor cortex are activated during tactile stimulation of different body regions. D 2005 Elsevier Inc. All rights reserved.
Introduction Early reports measuring cortical (Jasper and Penfield, 1949) and scalp (Gastaut et al., 1952) recorded brain activity described changes in EEG rhythms accompanying the preparation and performance of voluntary movement. More recently, a number of neuroelectric (EEG) and neuromagnetic (MEG) experiments have demonstrated that voluntary movement is preceded by a reduction in mu (8 – 15 Hz) and beta (15 – 35 Hz) power originating over
* Corresponding author. Fax: +1 416 813 7362. E-mail address: [email protected] (W. Gaetz). Available online on ScienceDirect (www.sciencedirect.com). 1053-8119/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2005.10.009
contralateral sensorimotor areas about 1.5 s prior to movement initiation (see Pfurtscheller and Lopes da Silva (1999) for review). Upon movement termination, mu power recovers slowly and with considerable variability between subjects (Salmelin et al., 1995; Salmelin and Hari, 1994b). However, beta power consistently returns to or exceeds baseline levels, a phenomenon termed beta rebound typically beginning within a half second of movement onset and persisting for several hundred milliseconds (Pfurtscheller et al., 1996). The neurophysiological mechanisms producing these oscillations are poorly understood (Lopes da Silva, 1991), and little is known about their functional significance (Fetz et al., 2000; MacKay, 1997). Pfurtscheller and Aranibar (1979) were first to refer to movement-related decreases in oscillatory EEG power as an event-related desynchronization (ERD) of neural populations. Later, experiments measuring synchrony in primate local field potentials (LFP) from sensorimotor cortex (Murthy and Fetz, 1996a,b) supported the inference that ongoing (resting-state) synchronous cortical oscillations are interrupted (i.e., desynchronized) by activating neurons involved in preparing and performing movement. These observations are further supported by experiments showing that motor imagery alone can produce beta ERD (Pfurtscheller et al., 2005; Schnitzler et al., 1997). Thus, it is generally assumed that movement-related desynchrony (ERD) represents a period of motor cortex activity involved in movement preparation and performance, whereas movement-related synchrony (beta ERS) represents a period of reduced cortical activity (idling) (Pfurtscheller et al., 1996) or inhibition of these same motor cortical areas (Chen et al., 1998; Salmelin et al., 1995) as a consequence of their prior activity. One problem facing experiments designed to study how voluntary movement can change sensorimotor rhythms relates to the difficulty in controlling for proprioceptive or haptic feedback which typically accompanies normal ballistic motor behavior (Kristeva et al., 1991). Indeed, recent experiments have also shown beta oscillations to be sensitive to somatosensory stimulation, such as median nerve stimulation (Salenius et al., 1997) and passive movement (Alegre et al., 2002; Cassim et al., 2001). For example, Cassim et al. (2001) showed that beta ERD/ERS could be
900
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
elicited by both active and passive finger movements and could be abolished by temporary deafferentation (i.e., digit ischemia). More recently, Cheyne et al. (2003) reported findings of sensorimotor ERD/ERS following tactile (brush) stimulation of the fingers using a novel spatial filtering method called synthetic aperture magnetometry (SAM) (Robinson and Vrba, 1999; Vrba and Robinson, 2001). These observations raise the interesting possibility that sensory stimulation alone may be sufficient to change the ongoing activity of beta oscillatory networks in motor cortex. In the present experiment, we applied the SAM spatial filtering method to localize cortical oscillations following tactile stimulation. By systematically varying the location and duration of stimulation, we sought to determine whether beta ERS is somatotopically organized and to observe the correspondence between stimulus duration and the time course of the observed changes in cortical oscillatory power. These experiments were performed in order to question whether previous reports of movement-related beta ERS may be explained by uncontrolled sensory events and thus test the assumption that motor cortical beta ERS is functionally linked to prior motor cortex activity (ERD).
All brush stimuli were presented with a soft bristled toothbrush and presented with light pressure making contact with an area of the skin approximately 1.5 cm2. A fiber optic light trigger was positioned in the path of the brush motion such that the interruption of the light path by the brush produced a TTL trigger when the brush made contact with the skin surface (‘‘brush-on’’) and when the brush departed from the skin surface (‘‘brush-off’’). An experimenter sat next to the subject and lightly brushed the subject’s skin at random intervals separated by several seconds. Subjects were required to fixate on a centrally positioned visual cue throughout the MEG recording session. A visual barrier was placed in the subject’s right visual field so that all motion of the experimenter, and the presented brush stimuli was not visible to the subject.
Methods
Sustained brush condition Sustained brushing was applied to the index finger only. These stimuli were applied in an identical manner to the transient condition except that stimulation consisted of bi-directional brushing for a period of approximately 3 s (10 s ISI). One sustained brush stimulus occurred for each of 100 trials (8-s duration; 2-s pretrigger period; ¨3-s brushing). In order to confirm the reliability of our source localizations, all brush stimuli were preceded by activation of the right and left median nerves. Median nerve stimuli were applied separately with a constant current, square wave pulses (3.1 Hz, 0.2-ms duration) delivered transcutaneously at the wrist above motor threshold. Somatosensory-evoked fields (SEFs) were recorded at a sample rate of 2500 Hz (bandpass DC to 800 Hz) with 600 epochs of 200ms duration.
Subjects Ten subjects (5 male, all right handed; mean age 31.5 years) participated in this experiment after receiving informed consent using protocols approved by the Hospital for Sick Children Research Ethics Board. Subjects were asked to sit upright in a comfortable chair with eyes open in a magnetically shielded room. Each subject was fitted with three fiducial localization coils placed at the nasion and preauricular points in order to localize the position of the subject’s head relative to the MEG sensors. Digital photographs of coil placement aided the coregistration of MEG data to each subject’s magnetic resonance image (MRI).
Transient brush condition Single transient brush stimuli consisted of a unidirectional brush motion of approximately 100-ms duration applied separately to the distal ventral phalanx of the right index finger and right medial toe, as well as the right half of the lower lip (¨6 s mean ISI). Each transient brush occurred once for each of 100 trials (5-s duration; 2-s prestimulus).
MEG recordings SAM analysis Neuromagnetic activity was recorded using a whole head 151 channel MEG system (Omega-151; VSM MedTech Ltd., Vancouver, Canada) with a sample rate of 625 samples/s, and a bandpass of 0 to 200 Hz. MEG data collection was synchronized to the interruption of a fiber optic light switch (Omron, 3EX-A41) by movement of the brush across the skin surface. Stimuli Prior to finger brushing, the subject’s hand was immobilized on an armrest in a comfortable position (palm up), and the proximal and middle phalanges of the right index finger were taped to the armrest to minimize inadvertent joint movements during brushing. Lip stimulation required that the fiber optic cables were adhered to the subject’s cheek, and positioned to detect brush stimuli presented to the lateral extent of the right lower lip. Finally, toe stimulation was accomplished by placing the bare right foot on an elevated support (approximately 30 cm) which included an angled sponge. The subject’s foot was positioned to rest against the sponge, with the toes extending out. The fiber optic cables were positioned at the distal glabrous surface of the medial right toe.
Synthetic aperture magnetometry (SAM) is a spatial filtering approach to source reconstruction, that has been successfully applied to the localization of cortical oscillations without the need for time-averaging of stimulus locked responses (Ishii et al., 1999; Singh et al., 2002; Taniguchi et al., 2000; Gaetz and Cheyne, 2003). This is achieved by integrating power changes over a defined time window relative to baseline activity using a differential power estimate (pseudo-t) (Robinson and Vrba, 1999). These differential images of source power can be calculated using single-trial data over selected frequency bands and time windows throughout the brain resulting in 3-dimensional source images co-registered with each subject’s head frame of reference. The application of this method to the localization of sensorimotor cortex oscillations has been described previously (Cheyne et al., 2003). T1-weighted structural MR images (3D SPGR) were obtained for each subject using a 1.5 T Signa Advantage system (GE Medical Systems, Milwaukee, USA). SAM source images were co-registered with the MR by matching the location of the localization coils used to define the MEG coordinate system on the MRI. In addition, dipole fits were performed on the averaged
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
response to median nerve stimulation at the latency of the M20. The location of the M20 dipoles could be overlaid on the same MR image and were inspected for each subject to confirm the location in the postcentral gyrus of the contralateral hemisphere. Pseudo-t images were computed for mu (8 – 14 Hz) and beta (15 – 35 Hz) frequency bands over the entire brain at 2-mm resolution. For each trial, stimulus event markers were placed (offline) which identified the fiber-optic stimulus events for Fbrush-on_ and Fbrush-off_. The control window for each brush event was 1.75 to 0.75 s before the onset of each brush (‘‘brush-on’’). The initial active window was 0.5 to 1.5 s following each brush (‘‘brush-off’’), however, the onset time was adjusted on the basis of virtual sensor and time – frequency plot information to optimally capture the observed changes in frequency power (see time – frequency methods below).
901
T1-weighted MR images and applied to the SAM ERD/ERS images. The normalized SAM images were then resampled to produce 2 2 2-mm resolution images of source power in standardized stereotaxic space and averaged across subjects. Due to large variations between the peak magnitude of ERD/ERS power, SAM images for each subject were normalized to the largest (peak value) in the image prior to spatial normalization. Statistical thresholds were calculated using a non-parametric statistical permutation test (SnPM) of the mean voxel value (Nichols and Holmes, 2002). Mean stereotaxic coordinates were calculated from the most significant activated voxels (statistical peak values) within each cluster. The MNI coordinates were converted to Talairach and Tournoux (1988) coordinates using a non-linear transformation (Lancaster et al., 2000) implemented in mri3dX software package (http://www.aston.ac.uk/lhs/staff/ singhkd/mri3dX/mri3dX.jsp).
Virtual sensors and time – frequency plots In addition to the ability to generate volumetric images of source power changes over defined time intervals, the time course of oscillatory activity may be observed for any location of interest by observing wideband (e.g., 1 – 50 Hz) output from the spatial filter for a target location. This Fvirtual sensor_ time series reflects millisecond by millisecond changes in oscillation power for the target location in the single trial data. A time – frequency analysis of virtual sensor output for arbitrary locations in the brain was applied to the unaveraged data using a wavelet based technique (Tallon-Baudry et al., 1997). This produced a time – frequency (TF) plot that represented changes in the total power of both phase locked and non-phase locked oscillations over the duration of the recording epoch. Virtual sensor based TF plots were observed for locations showing peak activity in the SAM images. This information also allowed us to optimize the placement of the active window (1-s duration) to be aligned to the onset of beta ERS. Differential SAM images of mu and beta band source power were then recomputed for the identified periods of ERD/ERS relative to premovement baseline activity. Virtual sensors and frequency (FFT) Previous work by Neuper and Pfurtscheller (2001) demonstrated that the specific frequency of beta rhythm ERS was lower for hand (¨20 Hz) than foot (¨30 Hz) movements. In order to address whether the specific frequency of beta ERS could also be predicted by presenting tactile stimulation to different body areas, an average of the Fast Fourier Transformed (FFT) virtual sensor single trial data was calculated for all images showing peak SAM activity. The beta ERS frequency corresponding to the largest FFT power was then noted for each peak location observed within a given SAM image, with mean values calculated for all subjects and brush locations. Group averaging and statistical thresholds Group averaging of the SAM ERD/ERS images was performed in order to determine group effects. Methods for group averaging of SAM functional images were first described by Singh et al. (2003). Spatial normalization to the MNI (T1) template was carried out using SPM99 (http://www.fil.ion.ac.uk/spm) implemented in MATLAB (http://www.mathworks.com/). Both linear and nonlinear warping parameters were obtained from individual subject’s
Results Transient brush stimuli Beta band ERS showed a consistent, robust stimulus-locked activity for each subject and brush condition. Inspection of singletrial MEG waveforms from sensors overlying the sensorimotor areas in individual subjects showed that the changes in beta ERD/ ERS could be observed within single trials, and that the beta ERS appeared time locked to the termination of the brush stimulus (see Fig. 1A). SAM analysis of beta band ERS following the transient brush of the index finger revealed consistent increases in beta power following brush stimuli which were localized to the motor cortical hand area bilaterally and was largest contralateral to side of stimulation in 7 out of 10 subjects. Time – frequency analysis of SAM virtual sensors from the peak location of the contralateral ERS showed a consistent decrease in beta band power coincident with the brush contacting the skin surface, followed by a rebound in beta band power starting at approximately 300 – 400 ms following the cessation of stimulation. Results from a representative subject are shown in Fig. 1. Responses to the transient brush of the right lower lip resulted in beta ERS motorotopic activation of contralateral lip area (MI) observed in all subjects as well as bilateral motor cortex hand area activity (larger pseudo-t values contralaterally in n = 5) (Fig. 1D). Transient brush stimuli of the right medial toe also produced a motorotopic activation of contralateral toe area (MI) observed in 8 out of 10 subjects (Fig. 1E) as well as bilateral motor cortex activity for each subject (larger pseudo-t value contralaterally in N = 8). Consistent with previous observations (Salmelin et al., 1995; Salmelin and Hari, 1994a), mu power ERS showed considerable variability and was not observed for all subjects. Consequently, post-stimulus changes in mu power (mu ERS) were not analyzed further. Grand averaged beta ERS Group averaged SAM images of beta ERS following the transient finger brush showed a significant activation of bilateral motor cortical hand area. Beta ERS following the transient lip brush produced a similar bilateral activation of motor cortical hand area as well as a significant contralateral activation of motor
902
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
cortical lip area lateral to the hand area. Grand averages of beta ERS following transient toe stimulation produced contralateral activation of the motor cortex hand area as well as significant activation of contralateral motor cortical toe area on the medial aspect of the precentral gyrus. All activations were significant in the non-parametric permutation test (SnPM; P < 0.05) shown in Fig. 2.
FFT analysis of ERS using virtual sensors Mean FFT values based on the virtual sensor single trials (FFT window 1-s duration; onset of beta ERS) were significantly higher in frequency for toe brush (26.01 Hz) than finger (20.99 Hz) (paired t test; P < 0.01, Bonferroni corrected) (Fig. 3A). FFT power was also compared for each of the peak frequencies
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
903
Fig. 2. Spatially normalized statistically thresholded grand averages for each brush condition (beta ERS) are shown coregistered to a template brain. Normalized SAM images were permuted using the mean. Co-active sources in motor cortex are observed for each brush condition (SnPM P < 0.05). Bilateral hand area activation was observed for each brush location, with the exception of the toe brush condition. Ipsilateral hand area activity was observed for individual subjects (toe brush) but was relatively weaker than contralateral activity and did not reach significance using SnPM.
observed in the active window. Beta ERS was observed to be relatively larger for the contralateral motor area MI-C, relative to the ipsilateral hand area MI-I for each brush location. The average hand area beta ERS power ((MI-C + MI-I)/2) was largest following index finger stimulation relative to lip and toe stimulation (paired t test; P < 0.05, Bonferroni corrected) (see Fig. 3B). Sustained brush stimuli The application of a sustained brush to the index finger produced a prolonged decrease in mu and beta power for each
subject which continued until the sustained brush stimulus was removed. TF plots based on virtual sensor locations in contralateral MI hand area showed a distinct mu and beta ERD throughout the ¨3 second brush period, followed by beta ERS approximately 300 ms following the lifting of the brush from the skin surface (see Fig. 4). In each subject, motor cortical beta ERS was not observed until the sensory stimulus was terminated. Grand averaged SAM images were calculated for mu and beta suppression starting 1.75 s prior to the termination of the sustained brush (1-s duration). These averaged SAM activation images showed a significant decrease in mu and beta power localized to distinct areas of contralateral
Fig. 1. (A) Two MEG channels (MLF45 and MLC42) overlying contralateral sensorimotor cortex showing beta (15 – 35 Hz) band activity for two single trials (trials 50 and 74) selected to show examples of different time intervals between onset (brush-on) and termination (brush-off) of stimulation. Brush onset (time = 0 s) coincides with beta ERD, followed by beta ERS which occurs later for the longer duration transient brush (trial 50), suggesting that the beta ERS is temporally locked to the offset of the brush stimulus. In both traces, beta ERS shows a topographical pattern consistent with source activity in contra as well as ipsilateral sensorimotor cortical areas (lower panel, topographic map). Time – frequency and SAM images were based on MEG data realigned to the brush off events as time = 0 s. (B) SAM localization of beta ERS for index finger brushing displayed on 3D rendered image of the subject’s MRI. A cutaway through the axial plane at the level of the motor hand areas shows bilateral motor cortex activation anterior to localized N20m dipole locations (not shown). The SAM pseudo-t image was calculated from all single trial MEG responses based on the active (0.2 – 1.2 s) and control ( 1.75 to 0.75 s) time windows. (C) A time – frequency (TF) plot of a virtual sensor based on the location of the largest SAM activity (contralateral MI) observed in panel (B). Red TF values indicate increased source power during the active state, relative to a 1-s baseline period ( 2 s to 1 s). (D and E) SAM localized beta ERS activity following right lower lip brush stimulation (left) and medial toe (right) is shown. Bilateral motor cortex activity (motor hand knob) as well as body part specific (motorotopic) cortical activations was observed.
904
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
Fig. 3. SAM localized peak activity for each subject and brush condition was evaluated to determine the frequency corresponding to the largest FFT power value. The FFT was performed on the virtual sensor single trials. Panel A compares the frequency (averaged over subjects) observed for each SAM location. Data from two subjects were not included for the brush toe condition due to low SNR. Panel A shows the frequency of the motorotopic toe activation produces a higher beta frequency than that observed coincidentally from the motorotopic hand area (paired t test; P < 0.01 Bonferroni corrected). Pooled FFT power values ((MI-I + MI-C)/2) were observed to decrease from brush index, lip and toe (paired t test; P < 0.05 Bonferroni corrected) conditions with no corresponding change in frequency.
somatosensory cortex (SnPM; P < 0.05). Group averaged SAM images to the post brush ERS showed a corresponding increase in beta power which, similar to results following the transient brush, were observed maximally from the contralateral motor cortex hand area (see Fig. 4, SnPM results). However, the SnPM map for beta ERS does not include significant ipsilateral motor cortex activation in the group thresholded beta ERS map. Power differences between MI-C and MI-I following transient stimulation were not significant, however, MI-I was significantly weaker than MI-I following the termination of the sustained brush (paired t test; P < 0.05 Bonferroni corrected).
Discussion We observed that transient tactile stimulation to different areas of the body produced both specific and general effects on ongoing rhythmic activity in the beta frequency band. Tactile input to the finger, lip and toe resulted in suppression and strong rebound of beta band power observed from precentral gyrus areas associated with movements from these same body parts. Thus, the beta rebound accompanying passive somatosensory input appears to be organized according to the homuncular distribution of motor output neurons of the primary motor cortex or motorotopically. This may reflect the close relationship between afferent projections of the somatosensory system to regions of the precentral gyrus by corticocortical connections with the somatosensory cortex (Jones, 1983) or by direct input from the thalamus (Horne and Tracey, 1979; Lemon, 1981) and the somatotopic organization of motor output corticospinal neurons to the same body regions. Using time – frequency analysis, we observed that sustained stimulation suppressed mu and beta rhythms maximally in contralateral sensorimotor areas over the duration of stimulus input, and that the termination of the stimulus produced a corresponding rebound in beta power arising from contralateral motor cortex hand area time-locked to the cessation of stimulation. This indicates that beta ERS to a single somatosensory stimulus most likely represents an Foff-response_ that is independent of the duration of stimulus. This observation is consistent with the hypothesis that movement-related beta ERS observed in the motor cortex may reflect the termination of sensory feedback resulting
from either passive or active movements (Cassim et al., 2001) rather than the return to resting state of motor cortex neurons following the initiation and execution of a voluntary movement as proposed in idling hypotheses of cortical oscillations. The term ‘‘idling’’ was first used by Adrian and Matthews (1934) to describe large EEG alpha rhythms observed from visual areas in the absence of visual stimulation. Kuhlman (1978) and later Pfurtscheller et al. (1996) have since suggested that the sensorimotor cortex also exhibits rhythmic activity corresponding to an idling state following the termination of a movement (ERS). Support for this suggestion has come from a transcranial magnetic stimulation (TMS) experiment that showed that the timing of beta ERS coincides with reduced excitability of motor cortex neurons (Chen et al., 1998). However, Cassim et al. (2001) argued against the idling hypothesis as they found that beta synchronization could be produced by passive finger movements, an event where the motor cortex is not first desynchronized. Like Cassim et al. (2001), we have observed sensorimotor beta ERD/ERS following somatosensory stimulation with no corresponding motor planning or production. Co-activation of hand regions of the motor cortex One of the more interesting observations in the current study is the fact that beta ERS from bilateral motor cortex hand areas were co-activated during lip and toe brushing, yet showed a similar peak frequency to that observed during stimulation of the digit alone (¨20 Hz). The magnitude of beta ERS in the hand region was observed to be significantly weaker during co-activation, in comparison to when the index finger was actually stimulated. The observation of hand area activation for lip and toe stimulation would suggest that beta rhythm changes observed in the motor cortex, particularly in the hand region, may be due to cortical inhibition through indirect pathways, presumably as the result of engaging a more distributed sensorimotor network and without the need for input from the periphery. We cannot rule out the possibility that this co-activation reflected activity in more rostral portions of area 4, even incorporating part of area 6 premotor neurons that are associated with more cognitive aspects of motor control. This would be consistent with functional magnetic resonance imaging (fMRI) studies that have shown that more
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
905
Fig. 4. Time – frequency (TF) plots on the left correspond to responses from a single subject based on the observed virtual sensor peak locations for mu ERD, beta ERD and beta ERS bands respectively. Group averaged and statistically thresholded activity for the active and control windows on the left are shown to the right (N = 10). Mu and beta band suppression is observed until the tactile stimulus is terminated. This suppression was localized to different areas of contralateral somatosensory cortex (SnPM; P < 0.05). Beta ERS appears immediately following the termination of the tactile brush and appears to be an ‘‘off response’’ arising from contralateral motor cortex.
rostral portions of the precentral gyrus may be active during motor tasks, particularly for the ipsilateral hemisphere (Hanakawa et al., 2005). This ‘‘non-specific’’ bilateral activation of the hand region of MI may therefore reflect a more general feature of motor cortex oscillations in sensorimotor tasks, including the observation of bilateral ERS in the precentral gyrus during actual finger movements (Cheyne and Gaetz, 2003; Neuper and Pfurtscheller, 2001). In a recent study, Jensen et al. (2005) showed that beta oscillation power can be increased by the administration of the GABAergic agonist benzodiazepine. Subjects administered benzodiazepines
showed an overall increase in resting, eyes-closed MEG beta band power, observed maximally over bilateral sensorimotor regions close to the hand area (Jensen et al., 2005). Further evidence of the specialized nature of hand area representations has come from recent primate research showing that unlike hindlimb regions of MI, the forelimb regions contain specialized inputs from areas of the prefrontal lobes as reflected by an abundance of indirect (polysynaptic) connections between these areas (Miyachi et al., 2005). Taken together, these results suggest that the hand region of the precentral gyrus may contain neuronal populations that
906
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
subserve a more general role in sensorimotor control that is independent of somatotopy—an observation that is important to take into account when interpreting the functional role of sensorimotor cortex oscillations in studies that focus exclusively on movements or stimulation of the digits. Pfurtscheller and Neuper (1994) observed that mu band EEG responses following simple hand movements produced both hand area ERD and foot area ERS (Pfurtscheller and Neuper, 1994). This phenomenon, later termed focal ERD/surround ERS (Suffczynski et al., 1999) was hypothesized to be the consequence of a directed focus of attention onto a specific motor act, thus producing ERD, whereas the surrounding cortex is outside the focus of attention, resulting in surround ERS. In the present experiment, close inspection of each individual’s response to tactile brush stimulation for both mu and beta ERD, as well as the group averaged ERD responses, failed to show any specific ERS in the surrounding cortical regions. It should be noted that our method involved the selection of time windows for each subject and stimulus location to optimally measure ERS power and location and avoid temporal overlap between post-stimulus ERD and ERS. The fact that we did not observe coincident ERD and ERS in the present experiment may have been the result of this methodological distinction or may indicate that the focal ERD/surround ERS observed by Pfurtscheller and Neuper (1994) is specific to mu rhythm changes following self-paced motor tasks, which are known to recover slowly and with considerable intersubject variability (Salmelin et al., 1995; Salmelin and Hari, 1994b). The somatosensory stimulation in the current study comprised a relatively salient stimulus that was attended to by the subjects, who were instructed to remain still and hence avoid any motor response to the tactile stimulus. Therefore, our task may have engaged cortical networks that involve the primary motor cortex and reflect higher order sensory control of movement, including the possible inhibition of a motor response to the tactile stimulus (withdrawal reflex). The involvement of MI in such cortical networks may explain why the modulation of these beta rhythms in motor cortex is observed for a wide range of somatosensory and motor tasks including motor imagery (Schnitzler et al., 1997; Pfurtscheller et al., 2005) or even observation of another individual’s digit being stimulated (Cheyne et al., 2003). The latter raises the possibility that mechanisms related to selective spatial attention, particularly with respect to the ability to anticipate a somatosensory input (Spence et al., 2000) could have engaged motor preparatory systems that would also explain the occurrence of beta ERS in the precentral hand area. However, in the current study, the stimulation was randomly presented and hidden from the subject’s view in order to minimize such anticipatory effects. In addition, the fact that beta ERS was highly time locked to the offset, rather than the onset of the stimulation would indicate that the subject’s anticipation of the stimulus did not play a role in the generation of the time-locked beta ERS.
higher peak frequency than that arising from cortical hand areas for stimulation of the corresponding body areas. Most notably, the observed changes in bilateral hand areas during toe and lip stimulation reflected the preferred frequency for finger stimulation rather than that for the body region that was being stimulated. This would suggest that each specific area of the primary motor cortex exhibits a stable and characteristic peak frequency, whereas the observed amplitude modulation of synchronized oscillations was only dependent on the presence or absence of sensory input. The reasons for the frequency specificity of beta ERS observed in the present study is unclear and raises some interesting questions about the underlying neural populations involved in the oscillatory response and what role various stimulus parameters may play in its generation, such as the number of afferent fibers activated (total area of stimulation), duration, or force. It is generally assumed that rapidly oscillating cell assemblies are comprised of fewer neurons than cell assemblies showing relatively slower oscillation frequencies (Singer, 1993). Using a neural network model, Lopes da Silva et al. (1976) observed the effect of varying the number of inhibitory interneurons on simulated alpha rhythmic activity. These simulations showed that an increased area (number) of synchronous inhibitory neurons reduces the peak frequency and increases the amplitude of the observed oscillation (Lopes da Silva et al., 1976). Although we did not vary the total area of stimulation or measure differences in receptive field size in the current study, the well-known differences in cortical representations of the body regions stimulated (e.g., finger tip versus toe) may support the interpretation that the different frequencies observed may have reflected size of the activated neural populations for different body regions. However, this does not explain why we observed similar frequencies of beta ERS in the hand region for both finger stimulation and stimulation of other body regions. Current studies are underway to explore the relationship between stimulus parameters (such as receptive field size) and the frequency and amplitude of movement and somatosensory induced ERD/ERS. In the present experiment, we were able to show consistent patterns of beta ERS across subjects by group averaging SAM images that were spatially normalized to stereotaxic space. However, we noted that with this approach weaker areas of activation (e.g., ipsilateral MI changes during transient toe brushing) that could be observed in individual subject images and were observed as peaks in the unthresholded group averages, failed to reach statistical significance when using an omnibus permutation test for whole brain images (SnPM). This was likely due to the known bias of the permutation distribution by presence of strong sources (Chau et al., 2004) that can result in an overly conservative test that will fail to detect weaker sources of activity that are co-active with the stronger sources as statistically significant. Further work is needed on SnPM or other significance tests involving omnibus derived permutation distributions which may be differentially biased by source strength.
Frequency specificity of beta ERS Conclusions We observed that input to different body regions appears to induce a beta ERS with a preferred frequency, consistent with previous EEG studies. For example, Neuper and Pfurtscheller (2001) observed that ERS hand movements produced a lower beta ERS frequency than foot movements (below 20 Hz for hand and above 20 Hz for foot). Similarly, the present results clearly show that beta ERS arising from toe motor cortex has a significantly
We have localized time-locked changes in rhythmic activity of primary somatosensory and motor cortex associated with tactile stimuli in the absence of motor planning and execution. Consistent with our previous results (Cheyne et al., 2003), transient stimulation produces beta ERD followed by beta ERS localized to bilateral primary motor cortical areas in the precentral gyrus. In
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
the current study, we also showed somatotopically organized cortical beta ERS for lip and toe brushing within the precentral gyrus, which suggests that the post-stimulus ERS is functionally related to the location of peripherally delivered somatosensory stimulation of the skin and reflects the known homuncular organization of movements of different body regions (motorotopy) within the same cortical area. Indeed, beta ERS may reflect the ongoing coordination of sensory input and motor output maintained by continuous input from the periphery via somatosensory afferents that in turn directly or indirectly inhibit associated areas of motor cortex.
Acknowledgments This work was supported by the Canadian Institutes of Health Research (Grant 64279). The authors would like to thank Krish Singh for the valuable contributions to this work and for making available the mri3dX software package. References Adrian, E.D., Matthews, B.H., 1934. The Berger rhythm: potential changes from the occipital lobes in man. Brain 57, 355 – 385. Alegre, M., Labarga, A., Gurtubay, I.G., Iriarte, J., Malanda, A., Artieda, J., 2002. Beta electroencephalograph changes during passive movements: sensory afferences contribute to beta event-related desynchronization in humans. Neurosci. Lett. 331 (1), 29 – 32. Cassim, F., Monaca, C., Szurhaj, W., Bourriez, J.L., Defebvre, L., Derambure, P., Guieu, J.D., 2001. Does post-movement beta synchronization reflect an idling motor cortex? NeuroReport 12 (17), 3859 – 3863. Chau, W., McIntosh, A.R., Robinson, S.E., Schulz, M., Pantev, C., 2004. Improving permutation test power for group analysis of spatially filtered MEG data. NeuroImage 23 (3), 983 – 996. Chen, R., Yaseen, Z., Cohen, L.G., Hallett, M., 1998. Time course of corticospinal excitability in reaction time and self-paced movements. Ann. Neurol. 44 (3), 317 – 325. Cheyne, D., Gaetz, W., 2003. Neuromagnetic activation of oscillatory brain activity associated with voluntary finger and toe movements. New York. NeuroImage 19 (2), e2088R (Suppl.). Cheyne, D., Gaetz, W., Garnero, L., Lachaux, J.P., Ducorps, A., Schwartz, D., Varela, F.J., 2003. Neuromagnetic imaging of cortical oscillations accompanying tactile stimulation. Brain Res. Cogn. Brain Res. 17 (3), 599 – 611. Fetz, E.E., Chen, D., Murthy, V.N., Matsumura, M., 2000. Synaptic interactions mediating synchrony and oscillations in primate sensorimotor cortex. J. Physiol. Paris 94 (5 – 6), 323 – 331. Gaetz, W., Cheyne, D., 2003. Localization of human somatosensory cortex using spatially filtered magnetoencephalography. Neurosci. Lett. 340 (3), 161 – 164. Gastaut, H., Terzian, H., Gastaut, Y., 1952. E´tude d’une activite´ e´lectroence´phalographique me´connue: Fle rhythme rolandique en arceau_. Marseille Me´d. 89, 296 – 310. Hanakawa, T., Parikh, S., Bruno, M.K., Hallett, M., 2005. Finger and face representations in the ipsilateral precentral motor areas in humans. J. Neurophysiol. 93 (5), 2950 – 2958. Horne, M.K., Tracey, D.J., 1979. The afferents and projections of the ventroposterolateral thalamus in the monkey. Exp. Brain Res. 36 (1), 129 – 141. Ishii, R., Shinosaki, K., Ukai, S., Inouye, T., Ishihara, T., Yoshimine, T., Hirabuki, N., Asada, H., Kihara, T., Robinson, S.E., et al., 1999. Medial prefrontal cortex generates frontal midline theta rhythm. NeuroReport 10 (4), 675 – 679.
907
Jasper, H.H., Penfield, W., 1949. Electrocorticograms in man: effect of the voluntary movement upon the electrical activity of the precentral gyrus. Arch. Psychiatry Z. Neurol. 183, 163 – 174. Jensen, O., Goel, P., Kopell, N., Pohja, M., Hari, R., Ermentrout, B., 2005. On the human sensorimotor-cortex beta rhythm: sources and modeling. NeuroImage 26 (2), 347 – 355. Jones, E.G., 1983. The nature of the afferent pathways conveying shortlatency inputs to primate motor cortex. In: Desmedt, J.E. (Ed.), Motor Control Mechanisms in Health and Disease. Raven Press, New York, pp. 263 – 285. Kristeva, R., Cheyne, D., Deecke, L., 1991. Neuromagnetic fields accompanying unilateral and bilateral voluntary movements: topography and analysis of cortical sources. Electroencephalogr. Clin. Neurophysiol. 81 (4), 284 – 298. Kuhlman, W.N., 1978. Functional topography of the human mu rhythm. Electroencephalogr. Clin. Neurophysiol. 44, 83 – 93. Lancaster, J.L., Woldorff, M.G., Parsons, L.M., Liotti, M., Freitas, C.S., Rainey, L., Kochunov, P.V., Nickerson, D., Mikiten, S.A., Fox, P.T., 2000. Automated Talairach atlas labels for functional brain mapping. Hum. Brain Mapp. 10 (3), 120 – 131. Lemon, R.N., 1981. Functional properties of monkey motor cortex neurones receiving afferent input from the hand and fingers. J. Physiol. 311, 497 – 519. Lopes da Silva, F.H., 1991. Neural mechanisms underlying brain waves: from neural membranes to networks. Electroencephalogr. Clin. Neurophysiol. 79, 81 – 93. Lopes da Silva, F.H., van Rotterdam, A., Barts, P., van Heusden, E., Burr, W., 1976. Models of neuronal populations: the basic mechanisms of rhythmicity. In: Corner, M.A., Swaab, D.F. (Eds.), Perspectives in Brain Research, Progr. in Brain Res. Elsevier, Amsterdam, pp. 281 – 308. MacKay, W.A., 1997. Synchronized neuronal oscillations and their role in motor processes. Trends Cogn. Sci. 1 (5), 176 – 183. Miyachi, S., Lu, X., Inoue, S., Iwasaki, T., Koike, S., Nambu, A., Takada, M., 2005. Organization of multisynaptic inputs from prefrontal cortex to primary motor cortex as revealed by retrograde transneuronal transport of rabies virus. J. Neurosci. 25 (10), 2547 – 2556. Murthy, V.N., Fetz, E.E., 1996a. Oscillatory activity in sensorimotor cortex of awake monkeys: synchronization of local field potentials and relation to behavior. J. Neurophysiol. 76 (6), 3949 – 3967. Murthy, V.N., Fetz, E.E., 1996b. Synchronization of neurons during local field potential oscillations in sensorimotor cortex of awake monkeys. J. Neurophysiol. 76 (6), 3968 – 3982. Neuper, C., Pfurtscheller, G., 2001. Event-related dynamics of cortical rhythms: frequency-specific features and functional correlates. Int. J. Psychophysiol. 43 (1), 41 – 58. Nichols, T.E., Holmes, A.P., 2002. Nonparametric permutation tests for functional neuroimaging: a primer with examples. Hum. Brain Mapp. 15 (1), 1 – 25. Pfurtscheller, G., Aranibar, A., 1979. Evaluation of event-related desynchrozization (ERD) preceding and following voluntary self-paced movement. Electrocephalogr. Clin. Neurophysiol. 46 (2), 138 – 146. Pfurtscheller, G., Lopes da Silva, F.H.L., 1999. Event-related EEG/MEG synchronization and desynchronization: basic principles. Clin. Neurophysiol. 110 (11), 1842 – 1857. Pfurtscheller, G., Neuper, C., 1994. Event-related synchronization of mu rhythm in the EEG over the cortical hand area in man. Neurosci. Lett. 174 (1), 93 – 96. Pfurtscheller, G., Stancak Jr., A., Neuper, C., 1996. Post-movement beta synchronization. A correlate of an idling motor area? Electroencephalogr. Clin. Neurophysiol. 98 (4), 281 – 293. Pfurtscheller, G., Neuper, C., Brunner, C., da Silva, F.L., 2005. Beta rebound after different types of motor imagery in man. Neurosci. Lett. 378 (3), 156 – 159. Robinson, S.E., Vrba, J., 1999. Functional neuroimaging by synthetic aperture magnetometry (SAM). In: Nenonen, J., Ilmoniemi, R.J., Katila, T. (Eds.), Espoo. Helsinki University of Technology, pp. 302 – 305. Salenius, S., Schnitzler, A., Salmelin, R., Jousmaki, V., Hari, R., 1997.
908
W. Gaetz, D. Cheyne / NeuroImage 30 (2006) 899 – 908
Modulation of human cortical rolandic rhythms during natural sensorimotor tasks. NeuroImage 5 (3), 221 – 228. Salmelin, R., Hari, R., 1994a. Characterization of spontaneous MEG rhythms in healthy adults. Electroencephalogr. Clin. Neurophysiol. 91 (4), 237 – 248. Salmelin, R., Hari, R., 1994b. Spatiotemporal characteristics of sensorimotor neuromagnetic rhythms related to thumb movement. Neuroscience 60 (2), 537 – 550. Salmelin, R., Hamalainen, M., Kajola, M., Hari, R., 1995. Functional segregation of movement-related rhythmic activity in the human brain. NeuroImage 2 (4), 237 – 243. Schnitzler, A., Salenius, S., Salmelin, R., Jousmaki, V., Hari, R., 1997. Involvement of primary motor cortex in motor imagery: a neuromagnetic study. NeuroImage 6 (3), 201 – 208. Singer, W., 1993. Synchronization of cortical activity and its putative role in information processing and learning. Annu. Rev. Physiol. 55, 349 – 374. Singh, K.D., Barnes, G.R., Hillebrand, A., Forde, E.M., Williams, A.L., 2002. Task-related changes in cortical synchronization are spatially coincident with the hemodynamic response. NeuroImage 16 (1), 103 – 114. Singh, K.D., Barnes, G.R., Hillebrand, A., 2003. Group imaging of task-
related changes in cortical synchronisation using nonparametric permutation testing. NeuroImage 19 (4), 1589 – 1601. Spence, C., Pavani, F., Driver, J., 2000. Crossmodal links between vision and touch in covert endogenous spatial attention. J. Exp. Psychol. Hum. Percept. Perform. 26 (4), 1298 – 1319. Suffczynski, P., Pijn, J.P.M., Pfurtscheller, G., Lopes da Silva, F.H., 1999. Event-related dynamics of alpha band rhythms: a neuronal network model of focal ERD/surround ERS. In: Pfurtscheller, G., Lopes da Silva, F.H. (Eds.), Event-related Desynchronization. Rev. Ed. Elsevier, Amsterdam, pp. 67 – 85. Talairach, J., Tournoux, P., 1988. Co-Planar Stereotaxic Atlas of the Human Brain. Thieme, New York. Tallon-Baudry, C., Bertrand, O., Delpuech, C., Permier, J., 1997. Oscillatory gamma-band (30 – 70 Hz) activity induced by a visual search task in humans. J. Neurosci. 17 (2), 722 – 734. Taniguchi, M., Kato, A., Fujita, N., Hirata, M., Tanaka, H., Kihara, T., Ninomiya, H., Hirabuki, N., Nakamura, H., Robinson, S.E., et al., 2000. Movement-related desynchronization of the cerebral cortex studied with spatially filtered magnetoencephalography. NeuroImage 12 (3), 298 – 306. Vrba, J., Robinson, S.E., 2001. Signal processing in magnetoencephalography. Methods (Duluth) 25 (2), 249 – 271.