Cerebral Cortex Advance Access published online on October 3, 2008
Cerebral Cortex, doi:10.1093/cercor/bhn174
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
A Role of Beta Oscillatory Synchrony in Biasing Response Competition?
1 Research Institute MOVE, VU University Amsterdam, van der Boechorststraat 9, 1081BT Amsterdam, the Netherlands, 2 Behavioural Brain Sciences Centre, School of Psychology, University of Birmingham Birmingham, B15 2TT, UK
Address correspondence to Peter Praamstra, Behavioural Brain Sciences Centre, University of Birmingham, Birmingham B15 2TT, UK. Email: p.praamstra{at}bham.ac.uk.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionFundingReferences |
|---|
Beta-range oscillatory activity measured over the motor cortex and beta synchrony between cortex and spinal cord can be up- or downregulated in anticipation of a postural challenge or the initiation of movement. Based on these properties of beta activity in the preparation for future events, the present investigation addressed whether simultaneous up- and downregulation of beta activity might act as an online mechanism to suppress and select competing responses. Measures of local and long-range beta synchrony were obtained from electroencephalographic and electromyographic signals recorded during a cued choice reaction task. Analyses focused on task-related changes in beta synchrony during a 2-s delay period between cue and response signal. Analyzed separately, none of the beta measures (spectral power, corticospinal coherence, corticospinal phase synchronization) showed simultaneous up- and downregulation over opposite hemispheres controlling the competing responses. However, the combined pattern of beta measures showed beta power desynchronization associated with selection of a response and increased corticospinal coherence and phase synchronization associated with suppression of a response. These results indicate that concurrent up- and downregulation of different components of beta oscillatory activity is likely to have a functional role in response selection, resembling attentional modulation of alpha activity in visual selection.
Key Words: corticospinal coherence • electroencephalography • movement preparation • response selection • sensorimotor function
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingReferences |
|---|
Population measures of motor cortex neural activity show a characteristic modulation of oscillatory activity in the beta frequency range preceding and following movement. The modulation consists of an anticipatory decrease and postmovement increase in power. A similar pattern can be observed for the synchronization strength between oscillatory activity in motor cortex and spinal cord, exemplified by high beta-band corticospinal synchrony during steady grip force, which is abolished during movement. This modulation pattern suggested that beta activity is associated with a resting or idling state (Pfurtscheller et al. 1996
). Recently, Brown and coworkers provided evidence for an active role of cortical and corticospinal beta synchrony in postural stabilization (Gilbertson et al. 2005; Androulidakis et al. 2007), indicating that changes in beta activity are causally relevant and not merely epiphenomenal. Key for the proposed role is an anticipatory upregulation of beta activity prior to an expected postural challenge, contrasted with the traditionally reported downregulation of beta activity before a speeded motor response. Related evidence for an association between beta synchronization and response inhibition was reported in an inhibition-of-return task (Pastötter et al. 2008). Building on the existing evidence that beta activity can be prospectively increased or decreased in a task-relevant way, we investigated whether beta activity in neural populations implementing competing response alternatives is modulated to bias response competition in favor of the selected response. A role of beta activity in suppressing and facilitating competing response alternatives would resemble the regulation of alpha activity in visual attention. Attending selectively to left or right hemispace is suggested to involve antagonistic suppression and enhancement of alpha activity over the occipital cortex contralateral to the attended and nonattended side (Worden et al. 2000; Rihs et al. 2007).
Relevant physiological measures were extracted from electroencephalographic and electromyographic data (EEG and EMG, respectively) recorded in a cued choice reaction task, in which a cue provided information as to which hand to prepare for an upcoming response. This instruction cue was followed at a fixed interval by an imperative signal indicating the response hand. Within the interval between cue and response signal (referred to as delay period), we were interested in early and late effects, representative of selection and preparation processes, respectively. Regarding response selection, there is already evidence for opposing influences on neural excitability of left and right hemisphere motor cortex from movement-related potential studies (Vidal et al. 2003; Praamstra and Seiss 2005) and from transcranial magnetic stimulation (Leocani et al. 2000; van den Hurk et al. 2007), but is there also a role of oscillatory synchrony in the selective facilitation and inhibition of competing responses? As to response preparation, it is known already that the premovement decrease in beta activity shows hemispheric lateralization as a function of the cued hand in a cued choice reaction task (Doyle et al. 2005; Kilner et al. 2005), but is there a corresponding lateralization of corticospinal synchronization? Based on the work of Androulidakis et al. (2007) and on the modulation of corticospinal coherence by response readiness (Schoffelen et al. 2005), we expected differential beta power and corticospinal synchronization effects for the selected and nonselected response in appropriate time windows for response selection and response preparation.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingReferences |
|---|
Participants
Fifteen healthy volunteers participated in the experiment (9 male, age 29 ± 8 years). Three participants were left-handed and one ambidextrous according to the Edinburgh Handedness Inventory (Oldfield 1971). Data of 2 participants were excluded because of excessive noise in their EEG recording and substantial loss of trials due to incorrect behavioral responses. Two further participants were excluded because their data did not display significant corticospinal synchronization in any condition. All participants gave written informed consent and were paid for their participation. The research was approved by the local ethics committee.
Behavioral Task and Stimuli
The task was a cued choice response task, adapted to measuring corticospinal synchronization by the use of compliant rubber bulbs as response force measurement devices, one for each hand. Because the focus of interest was on the modulation of cortical and corticospinal synchronization during the selection and preparation of a manual response, participants maintained a constant, instructed force during the presentation of the cue stimulus and the following fore period between cue and response signal. Participants initiated a trial by increasing left and right hand force levels to the target force range (see Fig. 1). When both force levels had stayed within target range for 2 s, the cue stimulus (S1) was displayed, consisting of arrows pointing either to the left or to the right. Participants had to maintain the same force level for another 2 s with both hands, after which a response stimulus (S2) was displayed, again consisting of arrows pointing either left or right. Participants were asked to respond by increasing the force level of the response hand as quickly as possible while maintaining the same force level with the nonresponse hand. After the response, grip of both hands was released and participants could initiate the next trial in their own time.
|
The experiment consisted of 8 blocks of 60 trials each (30 left hand responses, 30 right hand responses). In 4 blocks constituting the predictive cue condition, the cue information was 100% valid regarding the side of the response, that is, the cue direction was always identical to the direction indicated by the response signal. In the remaining 4 blocks (the nonpredictive cue condition) cue validity was 50%, rendering the cue direction nonpredictive for the response side. Predictive cue and nonpredictive cue blocks were arranged so that 2 successive blocks were of the same type, with their order counterbalanced across subjects. Prior to each block, participants were informed whether S1 was predictive with respect to the upcoming response or not. Right and left hand responses were randomized within each block.
The experiment was run in a normally illuminated room with participants comfortably seated in a chair with their 90° pronated forearms leaning on the armrests. The compliant rubber bulb (Ø 4 cm) response device was held between thumb and index finger and supported by the middle finger. Force was exerted by squeezing the bulb. Visual stimuli were presented in white against a gray background on a computer screen at 100 cm distance. Participants were instructed to center their gaze on a 1.4° x 1.4° fixation area in the center of the screen surrounded by brackets (see Fig. 1). Cue and response signals were presented within that area and measured 0.7° x 0.7° of visual angle with a display duration of 200 ms. Online feedback about left and right hand force levels was provided by means of 2 colored bars (combined size 0.9° x 3.6°) directly below the fixation area. With increasing force level, the height of the bar increased and its color changed from green to red. Target force level (
1 N) and range ( ±0.2 N) were marked on the 2 bars. Participants were instructed to monitor the produced force level while they built up force, but maintain their center of gaze onto the fixation area. Participants could rest between blocks and relax their fingers. Before recordings participants performed a practice block to familiarize themselves with equipment and task.
Instrumentation and Physiological Recordings
Experimental control of stimulus presentation was performed via custom written software. Measurement of force levels was realized via air pressure variation in 2 closed pneumatic systems, each consisting of a compliant rubber bulb connected by 2 m length of 2-mm ID tube to a piezo resistive pressure sensor (Honeywell 40PC015G1A). The pressure sensor output signal was fed through the EEG AD-converter and stored with EEG and EMG signals. For online feedback, the sensor output was AD-converted by a National instruments NI-PCI-MIO-16E board at the screen refresh rate (100 Hz). Calibration of the system was performed periodically between experiments using a secondary precision load cell. EEG was recorded with Ag/AgCl electrodes from 128 scalp electrodes, including left and right mastoid electrodes. A nylon cap was used to place the electrodes according to the 10-5 electrode system (Oostenveld and Praamstra 2001). Concurrently, surface EMGs were recorded above the first dorsal interosseous muscle and flexor pollicis brevis muscle of both hands, with electrodes placed on the muscle belly and tendon. All signals were low-pass filtered (cut-off frequency at 256 Hz) and amplified before sampling at a rate of 1024 Hz (BioSemi Active-Two amplifiers).
Data Analysis
Behavioral Responses
Reaction times and response accuracy were determined using the continuous force recordings. The baseline force level was defined as the mean force in the time window from –1000 ms until the onset of the cue signal (S1). Reaction time was defined as the time point after the response signal (S2) at which the force of the correct response hand exceeded a threshold of 105% of the baseline force. Responses were classified incorrect if the force level of a hand exceeded a threshold of 110% of the baseline force during the fore period between S1 and S2, if the force level of the nonresponse hand exceeded 105% baseline force level during the response interval, or if the latter dropped below 95% baseline force level before 350 ms after S2. All incorrect trials were removed before analysis of response times and left out from the EEG analyses. The response times were statistically evaluated with a repeated measures ANOVA with factors Cue information (predictive vs. nonpredictive) and Hand (left vs. right).
Movement-Related Potentials
After removal of incorrect responses and exclusion of artifact-contaminated EEG segments, there remained 
96 trials per condition for each subject for EEG/EMG analyses. For extraction of movement-related potentials, EEG was analyzed using BrainVision Analyzer software (Brain Products GmbH, Germany). The continuously recorded data were referenced to off-line averaged mastoid electrodes and then segmented into epochs of 3.5 s, that is, from 500 ms before S1 to 1000 ms after S2. Epochs were averaged separately for each of the 4 conditions, with a 500-ms prestimulus baseline. To evaluate whether participants used cue information, lateralized movement-related potentials were derived by computing difference potentials between homologous electrode sites over left and right hemisphere. That is, EEG at channels ipsilateral to the cued response hand was subtracted from the signal recorded at contralateral sites. Subsequently, averaging over left and right hand conditions yielded lateralized movement-related potentials, which were computed separately for the predictive and nonpredictive cue condition. In the nonpredictive condition, lateralized potentials were also calculated relative to the actual instead of the cued response side. The lateralized potentials of interest were the anterior directing-attention negativity (ADAN) associated with the selection of a response (Verleger et al. 2000; Praamstra et al. 2005), and the lateralized readiness potential (LRP) associated with preparation and execution of a response (Leuthold et al. 1996). ADAN and LRP were quantified in a time window from 400 to 450 ms and 1800 to 2000 ms after S1, respectively, in a selection of channels corresponding to the scalp maximum (ADAN: FCC3h/4h, FCC5h/6h, FC3/4, FFC3h/4h and FFC5h/6h; LRP: C1/2, C3/4, CCP3h/4h, and FCC3h/4h). For the purpose of topographic mapping using spherical spline interpolation (Perrin et al. 1989), the lateralized potentials were displayed in antisymmetric fashion over both sides of the head (cf. Praamstra et al. 1996). For the statistical evaluation of movement-related potentials we used paired-samples t-tests.
Time–Frequency Analysis
For time–frequency analyses, data were preprocessed in BESA (Brain Electrical Source Analysis, MEGIS Software GmbH, Germany) to create reference-free current source density derivations (Perrin et al. 1989) optimal for analysis of corticospinal synchronization (Mima and Hallett 1999). Both EEG and EMG signals were band-pass filtered (bidirectional fourth-order Butterworth filter, 5–150 Hz), and resampled to 512 Hz to optimize computation time and storage space.
Time–frequency analyses were performed in Matlab (ver 7.4, MathWorks, Natwick, MA). We used the open-source FieldTrip toolbox (www.ru.nl/fcdonders/fieldtrip) for the calculation of spectral power and coherence. All data were segmented into epochs of 4.5 s, starting 2000 ms before S1 and ending 500 ms after S2. After DC-removal, EMG data were rectified via the analytical signal constructed using the Hilbert transform (Myers et al. 2003). Time–frequency spectra for EEG and rectified EMG were calculated using a multitaper method with Slepian sequences based tapers. This involved the use of a sliding window with time steps of 50 ms and 2-Hz frequency bins from 5 to 69 Hz. Per frequency bin the window length was chosen as 5 periods of the central frequency, that is, 5/frequency. We further applied a spectral smoothing of ±0.4x frequency, which resulted in 3 tapers per window. All power spectra were log-transformed before averaging to improve normality (Halliday et al. 1995). The beta-band turned out to display the most pronounced task-related modulations throughout analyses (see Results). Hence, we selected the frequency range 15–30 Hz and averaged the power spectra over these frequencies. EEG power spectra were averaged across 4 channels overlying the left motor cortex (C1, C3, CCP3h, FCC3h) and across the homologous right hemisphere channels (C2, C4, CCP4h, FCC4h).
We adopted the approach of Baker and Baker (2003) for the calculation of time–frequency cross-spectra between each muscle and all EEG channels. Absolute coherence values were Z-transformed and corrected for a possible bias by subtracting the mean coherence over all time points within a frequency range in which significant coherence was absent during the whole epoch. Subsequently, Z-values were averaged over the 2 muscles per hand, the previously selected EEG channels (see above) and over the 15- to 30-Hz frequency band. All values were related to a significance level as defined in Rosenberg et al. (1989). As mentioned earlier, 2 participants who did not display significant coherence were excluded from further analysis.
For the analysis of corticospinal phase synchronization, defined as the phase uniformity of the relative phase between EMG and EEG signals (Mardia and Jupp 2000; see also Mormann et al. 2000 for application), we used consecutive 2-Hz frequency bins from 5 to 69 Hz (overlapping by 1 Hz). Per frequency bin the raw signals were filtered (bidirectional second-order Butterworth) and the analytical signal, calculated via the Hilbert transform, yielded the Hilbert-phase at every instant. Phase uniformity was calculated as mean resultant length of the difference between the phase of the EMG and EEG signals. Equivalent to coherence, the uniformity values were averaged over the 2 muscles per hand, over the earlier selected EEG channels and across the 15- to 30-Hz frequency band. Prior to averaging we applied Fisher's transform, Tan h–1, and used Tan h to invert the transform of the means (Amjad et al. 1997).
In order to correct for inter- and intraindividual differences in absolute values, the power and phase synchronization values were divided by the corresponding mean values estimated in the baseline interval –500 to 0 ms. Subsequently, power, coherence and phase synchronization measures were averaged across left and right hand conditions, to express them in terms of the hemisphere contralateral and ipsilateral to the cue direction. Mean absolute values averaged over left and right hand trials did not differ between conditions in the baseline interval as determined with paired-samples t-tests. Results were statistically evaluated with 2 x 2 repeated measures ANOVAs with factors Cue information (predictive vs. nonpredictive) and Hemisphere (ipsi- vs. contralateral to cued hand). Separate ANOVAs were performed on time windows 200–1000 ms and 1500–2000 ms, selected on the basis of visual inspection of the data. In addition, data points in the 200- to 1000-ms interval were compared with the baseline interval using serial 2-tailed paired t-tests, yielding time-evolving P values. In order to correct for type I errors due to the large number of comparisons (17), we considered effects to be significant only when P values of at least 2 consecutive data points were below 0.05 (see Lange et al. 1999).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingReferences |
|---|
Response Times
Participants responded significantly faster in trials with predictive cues (307 ± 53 ms) compared with trials with nonpredictive cues (421 ± 36 ms), expressed in a significant main effect of Cue information (F1,10 = 190.07, P < 0.001). There was no effect of Hand (F1,10 < 1) and no interaction effect (F1,10 < 1). In the nonpredictive cue condition, reaction times were not significantly different (F1,10 = 1.53, P = 0.156) between trials with congruent cues (426 ± 54 ms) and trials with incongruent cues (415 ± 54 ms). In view of the number of nonright handed participants, analyses were also performed in terms of dominant versus nondominant hand, instead of right versus left, yielding comparable results.
Movement-Related Potentials
The analysis of response times demonstrates that participants benefited from predictive cues relative to nonpredictive cues. We analyzed movement-related potentials to establish whether this response time benefit was associated with advance preparation during the fore period. Movement preparatory activity was found in the form of the ADAN and the LRP, represented in Figure 2a,b, respectively. The ADAN is generally elicited by directional information guiding the direction of spatial attention or the side of a manual response. It is typically distributed over the frontocentral scalp, peaking at about 400 ms. Accordingly, the present data displayed an ADAN peaking at 425 ms after the cue with an appropriate distribution. The amplitude was significantly different from zero, both in the predictive and in the nonpredictive cue condition (t(10) = 4.31, P = 0.002 and t(10) = 3.61, P = 0.005, respectively). Although higher in the former compared with the latter, this difference was not significant (t(10) < 1). When lateralized potentials in the nonpredictive cue condition were computed relative to the response side instead of the cue direction, the ADAN effects of left and right directed cues cancelled each other, eliminating the ADAN.
|
Results for the LRP are shown in Figure 2b. The scalp topography showed a more posterior distribution of lateralized activity compared with the ADAN. The slowly rising negativity before the response signal in the predictive cue condition is related to response preparation and was significantly different from the signal recorded for the nonpredictive cue condition. This was the case when the LRP of the latter condition was computed relative to the cue direction (t(10) = 3.62, P = 0.005) and when calculated relative to the response side (t(10) = 3.38, P = 0.007). Hence, faster response times in the predictive cue condition were accompanied by preparatory activity that was lateralized to the hemisphere contralateral to the cued response side before onset of the response signal. In summary, these results demonstrate that the task of maintaining a constant force level during the delay period did not prevent subjects to exploit the cue information for the advance selection and preparation of a response.
Time–Frequency Analyses
Time–frequency analyses started with determining the time–frequency windows and corresponding scalp distributions of task-related changes in synchronized activity, which were concentrated in the beta-range (see Fig. 3a). The beta cortical power modulations showed a frontocentral scalp distribution, as illustrated in Figure 3b. The figure represents the preparatory decrease in power (desynchronization) in the final 500 ms of the delay period. This decrease in power was lateralized to the hemisphere contralateral to the prepared response side in the predictive cue condition, but had a midline focus in the nonpredictive cue condition. The scalp channels overlying the lateralized maxima also appeared to be the 4 channels with the strongest corticospinal synchronization (Fig. 3b). These 2 groups of channels were therefore selected for further analyses of beta power modulation, corticospinal coherence and phase synchronization.
|
All measures of beta activity (15–30 Hz) were averaged across the 4 channels and their time courses were plotted separately for hemispheres contra- and ipsilateral to the response side (see Fig. 4). Power, coherence, and phase synchronization increased during the first part of the trial when participants increased grip force from zero into the target force range and reached a plateau at –500 to 0 ms serving as baseline interval. Differences between hemispheres and conditions were found in the interval between S1 and S2, which will be described below. Differences in the final 1000 ms were assumed to be related to response preparation, as already signaled by a steep reduction in beta activity during this time window. They were attributed to the processing of cue information and response selection in the first 1000 ms of the delay period.
|
Early Delay Period Effects
Effects in the first 1000 ms of the fore period were not straightforward to characterize due to multiple phases of beta power decreases and increases (see Fig. 4). Ignoring these multiple phases, that is, comparing beta power between conditions across the entire 200- to 1000-ms window, we found a significant main effect of Hemisphere due to lower beta power contra- than ipsilateral to the cue direction. In addition, there was lower beta power in the predictive than in the nonpredictive cue condition as expressed in a significant main effect of Cue information (see Table 1). The hemispheric difference in power being stronger in the predictive than in the nonpredictive cue condition yielded a significant interaction of Hemisphere by Cue information. Serial 2-tailed paired t-tests were conducted to compare power in the interval 200–1000 ms to baseline. As shown in Figure 4, power for the hemisphere contralateral to the cued hand in the predictive cue condition decreased significantly, whereas in the nonpredictive cue condition there were brief periods of significant increases.
|
Corticospinal coherence in the early 200- to 1000-ms interval followed the same time course as beta power (see Fig. 4) and demonstrated the same pattern of effects, except for a nonsignificant interaction of Cue information by Hemisphere (see Table 1). This similarity between beta power and beta-range corticospinal coherence was not entirely unexpected given that coherence estimates may be influenced by amplitude. Phase synchronization, by contrast, is by construction independent of amplitude. Figure 4 reveals that the most conspicuous difference between coherence and phase synchronization was the absence of the short-lasting amplitude drop around 500 ms, which also characterized the beta power time course. Instead, phase synchronization showed more sustained effects in the 200- to 1000-ms interval. Similar to coherence, phase synchronization stayed level for the hemisphere contralateral to the cued hand in the predictive cue condition, whereas it increased over the ipsilateral hemisphere, and increased more prominently when cue information was not predictive (see Fig. 4). This resulted in significant main effects of Cue information and Hemisphere, again without a significant interaction effect. The more sustained nature of phase synchronization effects was also evident from the sustained significant increase in the serial t-tests compared with baseline in the nonpredictive cue condition, where power and coherence demonstrated significant increases only around peak values. A similar sustained increase of phase synchronization was evident over the ipsilateral hemisphere in the predictive cue condition, although the window in which it reached significance relative to baseline was relatively short (see Fig. 4).
In summary, across the different measures of beta activity quantified in the 200- to 1000-ms window, there were instances of beta upregulation and beta downregulation. For none of the different beta measures, there was simultaneous up- and downregulation of activity over the hemispheres controlling the noncued and the cued response. Note, however, that in contrast to the later response preparation time window (see below), cue direction induced an interhemispheric lateralization also when it was nonpredictive. The interhemispheric lateralization signals that cue information, though irrelevant in the nonpredictive condition, was processed nonetheless and affected the motor system. The sustained increase of beta power in the nonpredictive relative to the predictive cue condition, across the 200- to 1000-ms interval, may reflect an upregulation of beta activity to prevent the cue information leading to differential preparation of left and right hand. Similarly, the increased beta phase synchronization, over both hemispheres in the nonpredictive cue condition and over the ipsilateral hemisphere in the predictive cue condition, "locks" the existing motor state to suppress a response from either hemisphere in the nonpredictive, and from the nonprimed hemisphere in the predictive cue condition.
Late Delay Period Effects
Response preparation effects in the late interval were straightforward, with a more pronounced beta desynchronization over the hemisphere contralateral to the cued response side in the predictive cue condition. This lateralization was not observed when subjects could not prepare their response in the nonpredictive cue condition. Together this yielded significant main effects of Cue information and Hemisphere, as well as a significant interaction (see Table 1). Like beta spectral power, the coherence and phase synchronization also decreased steeply in the late interval, with a lateralization only in the predictive cue condition. Again, this was reflected in significant main effects of Cue information and Hemisphere as well as a significant interaction for both synchronization measures (see Table 1).
Evidently, the steep reduction in beta power, coherence, and phase synchronization in the second half of the fore period was modulated by Cue information, but not conditional on hand-specific response preparation. Were there, superimposed on this bihemispheric reduction, signs of antagonistic enhancement and suppression of beta activity over the hemispheres controlling the nonselected and the selected response in the predictive cue condition? If so, this would be manifested in higher (lower) values of beta activity contralateral to the nonselected (selected) response, relative to the nonpredictive cue condition. Post hoc t-tests indicated that for each of the beta measures there were significantly lower values over the hemisphere controlling the selected response, relative to the nonpredictive cue condition. However, there was no significant enhancement of beta activity contralateral to the nonselected response. Hence, in this late time window, our data did not support an inhibitory role of beta activity.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingReferences |
|---|
Although the initiation of a new movement is known to be preceded by a downregulation of beta activity, recent work has demonstrated that beta oscillatory synchrony can be prospectively upregulated when the task requires maintained posture (Gilbertson et al. 2005
; Androulidakis et al. 2007). This suggests a mechanistic role of beta synchrony in optimizing the motor state for current tasks. In the domain of visual spatial attention there is evidence for prospective up- and downregulation of activity in the alpha frequency band. Moreover, concurrent up- and downregulation of alpha oscillatory synchrony is suggested to suppress nonattended and facilitate attended input to opposite hemispheres (Worden et al. 2000; Thut et al. 2006). Both these examples concern anticipatory adjustments of oscillatory activity for future events. The present investigation addressed whether concurrent up- and downregulation of beta synchrony acts as an online mechanism implementing the selection between competing motor responses.
We examined both local synchrony and long-range synchrony in the form of corticospinal coherence and phase synchronization. Preparatory processes during the second half of the fore period between cue and imperative signal were associated with a steep reduction of beta synchrony. At the same time, there was a differential modulation of ipsi- and contralateral hemisphere when there was advance information regarding the response side. The contralaterally dominant reduction of beta power replicates earlier work (Leocani et al. 2001; Doyle et al. 2005). The current data, however, are the first to show that not only beta power but also corticospinal coherence and phase synchronization in the same time window differentiate between the hemisphere that is involved and the one that is not involved in the preparation of a response. Schoffelen et al. (2007) did not find such a modulation of corticospinal coherence, possibly due to task differences, in particular a variable S1-S2 timing, and because these authors did not apply a time-resolved frequency analysis. Although our data in the late time window show clear-cut hemispheric differences when participants could prepare a response, both in local and in long-range beta synchrony, the differential modulation does not amount to concurrent up- and downregulation. However, it is still conceivable that there is such a modulation in opposite directions, but that it is masked by simultaneous reduction of beta synchrony related to other processes, such as stimulus anticipation and temporal preparation (Alegre et al. 2006; Praamstra et al. 2006). This possibility is difficult to substantiate because there are also known interhemispheric cross-talk effects (Kilner et al. 2003).
Support for a role of beta synchrony in the selection and suppression of competing responses is suggested by the behavior of beta activity in the first half of the fore period. Based on the combined evaluation of power, coherence, and phase synchronization, we hypothesize that beta activity is subject to a relatively sustained condition-dependent modulation that is overlapped by a stimulus-induced beta desynchronization of shorter duration, thus accounting for the multiple phases of beta synchrony increases and decreases in power and coherence in this time window. The stimulus-induced desynchronization is by and large identical between conditions and hemispheres and peaks at about 500 ms, whereas the more sustained modulation is affected by the experimental manipulations and extends from shortly after stimulus onset to about 1000 ms. Because coherence estimates depend on power, it is unsurprising that the stimulus-induced beta desynchronization affected corticospinal coherence (see Fig. 4). In contrast, power does not influence corticospinal phase synchronization. Thus, the behavior of beta-range corticospinal phase synchronization captures an important part of the observed modulation of local beta synchrony, uncontaminated by the stimulus-induced influence on beta power. Crucially, it reveals increases of beta synchrony over both hemispheres in the nonpredictive cue condition and over the ipsilateral hemisphere in the predictive cue condition. These increases are plausibly construed as stabilizing the current motor state and inhibiting a motor response from the involved hemispheres. A beta increase related to response inhibition has been shown before in nogo response suppression (Alegre et al. 2004; Kühn et al. 2004) and, recently, in an inhibition-of-return paradigm (Pastötter et al. 2008) suggesting a possible role in response selection.
In the proposed interpretation, selection of the cued response and deselection of the noncued response are reflected in decreased spectral power and increased phase synchronization, respectively, in opposite hemispheres. A selective decrease of spectral power over the hemisphere controlling the response hand can obviously not be accounted for in terms of the condition-independent stimulus-induced beta desynchronization phase referred to above. That there is another factor that drives beta power down over the hemisphere that is primed for a forthcoming response is suggested by the finding that the ipsilateral–contralateral hemisphere difference in power was more pronounced in the predictive than in the nonpredictive cue condition (expressed in a significant Cue information by Hemisphere interaction), whereas the interhemispheric difference in phase synchronization was identical between predictive and nonpredictive cue conditions. This supports the antagonistic modulation of beta spectral power and phase synchronization in opposite hemispheres. The possibility of a dissociation between local synchronization and corticospinal synchronization has been established by pharmacological manipulations and by the behavior of beta rhythms in some disease states (Baker and Baker 2003; Riddle et al. 2004; Jensen et al. 2005). In the present context of response selection, the downregulation of cortical and upregulation of corticospinal beta-range synchronization might well represent a means of biasing the competition between alternative, simultaneously activated responses. In van Wijk et al. (2008), we have analyzed the present data from a different perspective, asking whether observed beta spectral power can be explained by a linear superposition of the sustained condition-dependent beta phase synchronization, as one component, and the overlapping, condition-independent stimulus-related beta desynchronization phase, as second component. Using principal component analysis, such a 2-component model explains the data remarkably well, except for the beta power reduction in the predictive cue condition over the primed hemisphere. This supports our view that phase synchronization and beta spectral power in opposite hemispheres can be modulated in an antagonistic way, acting together to suppress and select competing responses.
Previous studies of beta oscillatory activity in S1-S2 tasks have also described beta desynchronization and synchronization effects following the cue, both at scalp EEG level and for local field potentials in the subthalamic nucleus and thalamus (Doyle et al. 2005; Alegre et al. 2006; Klostermann et al. 2007). However, systematic early effects related to the meaning of the cue have, to our knowledge, only been reported in local field potential beta activity recorded from deep brain stimulation electrodes in the subthalamic nucleus in Parkinson's disease patients (Williams et al. 2003). Of relevance to our findings, the effects were also modulated by the predictive value of cues, though not by cue direction, and inferred to be related to the behavioral exploitation of cue information. Although reports of such early effects are rare, the effects are not entirely unexpected, not even in the situation where the cue information is not task-relevant, as in the nonpredictive cue information condition. For instance, arrow cues carrying no predictive or response-relevant directional information still produce an early lateralization of movement-related activity in the form of the ADAN or LRP components (Eimer 1995; Verleger et al. 2000), as they did here. As to oscillatory activity, beta activity is already known to be modulated by imagined and observed movements, commonly interpreted in terms of a covert motor command (Muthukumaraswamy and Johnson 2004; Caetano et al. 2007; Tkach et al. 2007). Our data indicate that a similar modulation of beta activity is elicited by overlearned symbolic information in the form of an arrow. Significantly, the prospective upregulation of corticospinal coherence in anticipation of a postural challenge, reported by Androulidakis et al. (2007), also occurred in a time window closer to the cue than to the perturbation, suggesting that it (partly) reflected a covert motor command elicited by the cue.
The role of beta synchrony in biasing response competition appears not only related to a biasing role of alpha activity in the selection of visual information (Worden et al. 2000; Kelly et al. 2006; Thut et al. 2006; Rihs et al. 2007), but also to the concept of "focal desynchronization/surround synchronization" (Pfurtscheller and Lopes da Silva 1999). The authors’ examples supporting this concept comprised antagonistic behavior of the rolandic mu rhythm and occipital alpha as well as antagonistic modulation of mu activity over hand and foot motor cortex. However, Pfurtscheller and Lopes da Silva (1999) also observed concurrent desynchronization and synchronization of beta activity over left and right motor cortex during motor imagery. Our data extend this observation and provide evidence that the antagonistic modulation can reflect a mechanism supporting the selection and suppression of competing responses. Related evidence for antagonistic modulation of competing response options comes from movement-related EEG potentials (Vidal et al. 2003; Praamstra and Seiss 2005), from transcranial magnetic stimulation (Leocani et al. 2000; van den Hurk et al. 2007), and from single-unit recordings in primates (Cisek and Kalaska 2005).
Within the context of research on the functional and anatomical loops connecting basal ganglia and cortex, it has been proposed that tuning of neural activity to different frequency bands may reflect a means of marking and segregating ongoing processes in different cortical–subcortical circuits (Fogelson et al. 2006). The operation of similar selection mechanisms carried by different frequency bands is in broad agreement with this proposal. With respect to selection operations carried by the sensorimotor beta rhythm, do the present results support the notion that differential modulation of beta-range activity implements a mechanism for the online selection between competing responses? If such a mechanism operates through antagonistic modulation of cortical beta power, then our data do not support the notion, because we found downregulation but not simultaneous upregulation of beta power. However, if the mechanism has a more complex mode of action, involving different components of beta-range oscillations, then our data do provide support for the investigated role, given the upregulation of corticospinal synchrony. The latter version probably represents a more plausible mechanism, recognizing that there are multiple components to the sensorimotor beta rhythm (cf. Jensen et al. 2005). Of relevance here, at the level of the striatum, task-related beta desynchronization occurs very focally amidst widespread beta oscillatory activity, giving rise to the proposal that beta oscillatory synchrony operates as a spatiotemporal filter to sharpen action-selection by cortico-basal ganglia networks (Courtemanche et al. 2003). The regulation of corticospinal synchrony during the selection between competing response options would fit the concept of such an action-selection network. This would mean that beta-range corticospinal synchronization is not exclusively a low-level motor control mechanism (Kilner et al. 2002).
|
Funding |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingReferences |
|---|
The Netherlands Organisation for Scientific Research (NWO) grant (no. 452-04-344) to A.D.
|
Acknowledgments |
|---|
Conflict of Interest: None declared.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingReferences |
|---|
Alegre M, Gurtubay IG, Labarga A, Iriarte J, Valencia M, Artieda J. Frontal and central oscillatory changes related to different aspects of the motor process: a study in go/no-go paradigms. Exp Brain Res. (2004) 159:14–22.[Web of Science][Medline]
Alegre M, Imirizaldu L, Valencia M, Iriarte J, Arcocha J, Artieda J. Alpha and beta changes in cortical oscillatory activity in a go/nogo randomly-delayed-response choice reaction time paradigm. Clin Neurophysiol. (2006) 117:16–25.[Web of Science][Medline]
Amjad AM, Halliday DM, Rosenberg JR, Conway BA. An extended difference of coherence test for comparing and combining several independent coherence estimates: theory and application to the study of motor units and physiological tremor. J Neurosci Methods. (1997) 73:69–79.[CrossRef][Web of Science][Medline]
Androulidakis AG, Doyle LM, Yarrow K, Litvak V, Gilbertson TP, Brown P. Anticipatory changes in beta synchrony in the human corticospinal system and associated improvements in task performance. Eur J Neurosci. (2007) 25:3758–3765.[CrossRef][Web of Science][Medline]
Baker MR, Baker SN. The effect of diazepam on motor cortical oscillations and corticomuscular coherence studied in man. J Physiol. (2003) 546:931–942.
Caetano G, Jousmaki V, Hari R. Actor's and observer's primary motor cortices stabilize similarly after seen or heard motor actions. Proc Natl Acad Sci USA. (2007) 104:9058–9062.
Cisek P, Kalaska JF. Neural correlates of reaching decisions in dorsal premotor cortex: specification of multiple direction choices and final selection of action. Neuron. (2005) 45:801–814.[CrossRef][Web of Science][Medline]
Courtemanche R, Fujii N, Graybiel AM. Synchronous, focally modulated beta-band oscillations characterize local field potential activity in the striatum of awake behaving monkeys. J Neurosci. (2003) 23:11741–11752.
Doyle LMF, Yarrow K, Brown P. Lateralization of event-related beta desynchronization in the EEG during pre-cued reaction time tasks. Clin Neurophysiol. (2005) 116:1879–1888.[CrossRef][Web of Science][Medline]
Eimer M. Stimulus-response compatibility and automatic response activation: evidence from psychophysiological studies. J Exp Psychol Hum Percept Perform. (1995) 21:837–854.[CrossRef][Web of Science][Medline]
Fogelson N, Williams D, Tijssen M, van Bruggen G, Speelman H, Brown P. Different functional loops between cerebral cortex and the subthalamic area in Parkinson's disease. Cereb Cortex. (2006) 16:64–75.
Gilbertson T, Lalo E, Doyle L, Di Lazzaro V, Cioni B, Brown P. Existing motor state is favored at the expense of new movement during 13-35 Hz oscillatory synchrony in the human corticospinal system. J Neurosci. (2005) 25:7771–7779.
Halliday DM, Rosenberg JR, Amjad AM, Breeze P, Conway BA, Farmer SF. A framework for the analysis of mixed time series/point process data—theory and application to the study of physiological tremor, single motor unit discharges and electromyograms. Prog Biophys Mol Biol. (1995) 64:237–278.[CrossRef][Web of Science][Medline]
Jensen O, Goel P, Kopell N, Pohja M, Hari R, Ermentrout B. On the human sensorimotor-cortex beta rhythm: sources and modeling. Neuroimage. (2005) 26:347–355.[CrossRef][Web of Science][Medline]
Kelly SP, Lalor EC, Reilly RB, Foxe JJ. Increases in alpha oscillatory power reflect an active retinotopic mechanism for distracter suppression during sustained visuospatial attention. J Neurophysiol. (2006) 95:3844–3851.
Kilner JM, Alonso-Alonso M, Fisher R, Lemon RN. Modulation of synchrony between single motor units during precision grip tasks in humans. J Physiol (Lond). (2002) 541:937–948.
Kilner JM, Bott L, Posada A. Modulations in the degree of synchronization during ongoing oscillatory activity in the human brain. Eur J Neurosci. (2005) 21:2547–2554.[CrossRef][Web of Science][Medline]
Kilner JM, Salenius S, Baker SN, Jackson A, Hari R, Lemon RN. Task-dependent modulations of cortical oscillatory activity in human subjects during a bimanual precision grip task. Neuroimage. (2003) 18:67–73.[CrossRef][Web of Science][Medline]
Klostermann F, Nikulin VV, Kühn AA, Marzinzik F, Wahl M, Pogosyan A, Kupsch A, Schneider GH, Brown P, Curio G. Task-related differential dynamics of EEG alpha- and beta-band synchronization in cortico-basal motor structures. Eur J Neurosci. (2007) 25:1604–1615.[CrossRef][Web of Science][Medline]
Kühn AA, Williams D, Kupsch A, Limousin P, Hariz M, Schneider GH, Yarrow K, Brown P. Event-related beta desynchronization in human subthalamic nucleus correlates with motor performance. Brain. (2004) 127:735–746.
Lange JJ, Wijers AA, Mulder LJ, Mulder G. ERP effects of spatial attention and display search with unilateral and bilateral stimulus displays. Biol Psychol. (1999) 50:203–233.[CrossRef][Web of Science][Medline]
Leocani L, Cohen LG, Wassermann EM, Ikoma K, Hallett M. Human corticospinal excitability evaluated with transcranial magnetic stimulation during different reaction time paradigms. Brain. (2000) 123:1161–1173.
Leocani L, Toro C, Zhuang P, Gerloff C, Hallett M. Event-related desynchronization in reaction time paradigms: a comparison with event-related potentials and corticospinal excitability. Clin Neurophysiol. (2001) 112:923–930.[CrossRef][Web of Science][Medline]
Leuthold H, Sommer W, Ulrich R. Partial advance information and response preparation: Inferences from the lateralized readiness potential. J Exp Psychol Gen. (1996) 125:307–323.[CrossRef][Web of Science][Medline]
Mardia KV, Jupp PE. Directional statistics (2000) West Sussex (UK): Wiley & Sons.
Mima T, Hallett M. Electroencephalographic analysis of cortico-muscular coherence: reference effect, volume conduction and generator mechanism. Clin Neurophysiol. (1999) 110:1892–1899.[CrossRef][Web of Science][Medline]
Mormann F, Lehnertz K, David P, Elger CE. Mean phase coherence as a measure for phase synchronization and its application to the EEG of epilepsy patients. Physica D. (2000) 144:358–369.[CrossRef]
Muthukumaraswamy SD, Johnson BW. Primary motor cortex activation during action observation revealed by wavelet analysis of the EEG. Clin Neurophysiol. (2004) 115:1760–1766.[CrossRef][Web of Science][Medline]
Myers LJ, Lowery M, O'Malley M, Vaughan CL, Heneghan C, St Clair GA, Harley YX, Sreenivasan R. Rectification and non-linear pre-processing of EMG signals for cortico-muscular analysis. J Neurosci Methods. (2003) 124:157–165.[CrossRef][Web of Science][Medline]
Oldfield RC. Assessment and analysis of handedness—Edinburgh Inventory. Neuropsychologia. (1971) 9:97–113.[CrossRef][Web of Science][Medline]
Oostenveld R, Praamstra P. The five percent electrode system for high-resolution EEG and ERP measurements. Clin Neurophysiol. (2001) 112:713–719.[CrossRef][Web of Science][Medline]
Pastötter B, Hanslmayr S, Bäuml KH. Inhibition of return arises from inhibition of response processes: an analysis of oscillatory beta activity. J Cogn Neurosci. (2008) 20:65–75.[CrossRef][Web of Science][Medline]
Perrin F, Pernier J, Bertrand O, Echallier JF. Spherical splines for scalp potential and current density mapping. Electroencephalogr Clin Neurophysiol. (1989) 72:184–187.[CrossRef][Web of Science][Medline]
Pfurtscheller G, Lopes da Silva FH. Event-related EEG/MEG synchronization and desynchronization: basic principles. Clin Neurophysiol. (1999) 110:1842–1857.[CrossRef][Web of Science][Medline]
Pfurtscheller G, Stancak A Jr, Neuper C. Post-movement beta synchronization. A correlate of an idling motor area? Electroencephalogr Clin Neurophysiol. (1996) 98:281–293.[CrossRef][Web of Science][Medline]
Praamstra P, Boutsen L, Humphreys GW. Frontoparietal control of spatial attention and motor intention in human EEG. J Neurophysiol. (2005) 94:764–774.
Praamstra P, Kourtis D, Kwok HF, Oostenveld R. Neurophysiology of implicit timing in serial choice reaction-time performance. J Neurosci. (2006) 26:5448–5455.
Praamstra P, Seiss E. The neurophysiology of response competition: motor cortex activation and inhibition following subliminal response priming. J Cogn Neurosci. (2005) 17:483–493.[CrossRef][Web of Science][Medline]
Praamstra P, Stegeman DF, Horstink MW, Cools AR. Dipole source analysis suggests selective modulation of the supplementary motor area contribution to the readiness potential. Electroencephalogr Clin Neurophysiol. (1996) 98:468–477.[CrossRef][Web of Science][Medline]
Riddle CN, Baker MR, Baker SN. The effect of carbamazepine on human corticomuscular coherence. Neuroimage. (2004) 22:333–340.[CrossRef][Web of Science][Medline]
Rihs TA, Michel CM, Thut G. Mechanisms of selective inhibition in visual spatial attention are indexed by alpha-band EEG synchronization. Eur J Neurosci. (2007) 25:603–610.[CrossRef][Web of Science][Medline]
Rosenberg JR, Amjad AM, Breeze P, Brillinger DR, Halliday DM. The Fourier approach to the identification of functional coupling between neuronal spike trains. Prog Biophys Mol Biol. (1989) 53:1–31.[CrossRef][Web of Science][Medline]
Schoffelen JM, Oostenveld R, Fries P. Neuronal coherence as a mechanism of effective corticospinal interaction. Science. (2005) 308:111–113.
Schoffelen JM, Poort J, Oostenveld R, Fries P. Local and long range beta-band synchronization in the human motor system during a visuomotor task. In: Neuronal communication through coherence in the human motor system—Schoffelen JM, ed. (2007) Nijmegen: F.C. Donders Centre for Cognitive Neuroimaging. 35–52. PhD thesis.
Thut G, Nietzel A, Brandt SA, Pascual-Leone A. Alpha-band electroencephalographic activity over occipital cortex indexes visuospatial attention bias and predicts visual target detection. J Neurosci. (2006) 26:9494–9502.
Tkach D, Reimer J, Hatsopoulos NG. Congruent activity during action and action observation in motor cortex. J Neurosci. (2007) 27:13241–13250.
van den Hurk P, Mars RB, van Elswijk G, Hegeman J, Pasman JW, Bloem BR, Toni I. Online maintenance of sensory and motor representations: effects on corticospinal excitability. J Neurophysiol. (2007) 97:1642–1648.
van Wijk BCM, Daffertshofer A, Praamstra P. Local and long-range beta synchrony in motor control. In: Kakigi R, Yokosawa K, Kuriki S, editors. Biomagnetism: interdisciplinary research and exploration. Sapparo, Japan: Hokkaido University Press (2008) 185–187.
Verleger R, Vollmer C, Wauschkuhn B, van der Lubbe RHJ, Wascher E. Dimensional overlap between arrows as cueing stimuli and responses? Evidence from contra-ipsilateral differences in EEG potentials. Cogn Brain Res. (2000) 10:99–109.[CrossRef][Medline]
Vidal F, Grapperon J, Bonnet M, Hasbroucq T. The nature of unilateral motor commands in between-hand choice tasks as revealed by surface Laplacian estimation. Psychophysiology. (2003) 40:796–805.[CrossRef][Web of Science][Medline]
Williams D, Kühn A, Kupsch A, Tijssen M, van Bruggen G, Speelman H, Hotton G, Yarrow K, Brown P. Behavioural cues are associated with modulations of synchronous oscillations in the human subthalamic nucleus. Brain. (2003) 126:1975–1985.
Worden MS, Foxe JJ, Wang N, Simpson GV. Anticipatory biasing of visuospatial attention indexed by retinotopically specific alpha-band electroencephalography increases over occipital cortex. J Neurosci. (2000) 20:RC63(1–6).
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  P. Praamstra, D. Kourtis, and K. Nazarpour Simultaneous Preparation of Multiple Potential Movements: Opposing Effects of Spatial Proximity Mediated by Premotor and Parietal Cortex J Neurophysiol, October 1, 2009; 102(4): 2084 - 2095. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||