Article Text
PDF
Abstract
Objective To study internal globus pallidus (GPi) activities and the interactions among the bilateral GPi and motor cortical areas during voluntary movements.
Methods Five patients with cervical dystonia who underwent bilateral GPi deep brain stimulation (DBS) were studied. Local field potentials from the GPi DBS electrodes and EEG were recorded while the patients performed externally triggered and self-initiated right wrist movements.
Results Movement related potentials were recorded at the GPi bilaterally before the onset of self-initiated but not externally triggered movements. In all movements studied, frequency analysis revealed a ∼10–24 Hz beta event related desynchronisation at bilateral GPi and with EEG recorded over the mid-frontal (Cz-Fz) and the bilateral sensorimotor cortical regions (C3/C4-Cz). A ∼64–68 Hz, gamma event related synchronisation was found with EEG recorded over the mid-frontal (Cz-Fz), the sensorimotor cortices (C3-Cz) and the GPi contralateral to movements. Both beta event related desynchronisation and gamma event related synchronisation occurred before the onset of self-initiated movements and at the onset of externally triggered movements. There was a resting ∼5–18 Hz coherence between the bilateral GPi, which attenuated for ∼1 s during movements. Gamma coherences were observed between EEG recorded over the mid-frontal (Cz-Fz), contralateral sensorimotor cortices (C3-Cz) and the GPi from 0 to 0.5 s after movement onset for externally triggered movements and from 0.5 s before to 0.5 s after movement onset for self-initiated movements.
Conclusions The beta and gamma frequency bands in the GPi are modulated by the preparation of self-initiated movements and the execution of self-initiated and externally triggered movements. The 5–18 Hz coherence at the bilateral GPi may be related to dystonia and its attenuation may facilitate voluntary movements.
Statistics from Altmetric.com
Introduction
The human internal globus pallidus (GPi) is the main output nucleus of the basal ganglia (BG). Bilateral GPi deep brain stimulation (DBS) is an established treatment for patients with dystonia1 2 although its therapeutic mechanisms are not well understood. A thorough study of how GPi and cortical activities are modulated by voluntary movements will lead to greater understanding of the pathophysiology in dystonia and may improve neuromodulation strategies.
Slow, negative, cortical movement related potentials (MRP) before the onset of self-paced movements but not movements triggered by randomised cues have been consistently observed in cortical motor areas.3–5 In the BG, MRPs were observed in the subthalamic nucleus (STN).6 7 In the GPi, MRP was reported in two patients with epilepsy8 but only referential recording was used and there was no phase reversal between adjacent electrode contacts separated by 1.5 mm. The MRPs could potentially be explained by volume conduction from other BG or cortical areas.6
Similar to the findings in the cortex and the STN,9–11 voluntary movements were associated with bilateral desynchronisation of GPi local field potentials (LFP) at frequencies <30 Hz and synchronisation of contralateral GPi gamma (∼60–80 Hz) oscillations.12 However, coupling between GPi and cortical movement related activities in the gamma frequency range has not been demonstrated. Previous studies have reported abnormal oscillations in the BG in patients with Parkinson's disease (PD) and dystonia.13 14 Dystonic patients were found to have greater 4–10 Hz power in the GPi than PD patients.15 An excessive <18 Hz GPi activity in dystonic patients was found to be coherent with and drive activities of dystonic muscles.16–19 PD patients were found to have increased beta activities, and elevated beta coherence between the bilateral STN was reported in patients with PD.20 However, coherence between bilateral GPi in patients with dystonia has not been reported.
In this study, we examined GPi and cortical activities and their interactions during externally triggered and self-initiated movements using MRP, frequency, coherence and phase coherence analyses. Our first hypothesis is that GPi MRP would be present before self-initiated but not movements triggered by a randomised external cue. As previous studies suggested that dystonia is associated with excessive <18 Hz oscillatory activities in the BG, our second hypothesis is that there is coherence at <18 Hz between the bilateral GPi, and this coherence is reduced with voluntary movements.
Methods
Patients
We studied five patients with cervical dystonia who had DBS electrodes implanted bilaterally in the GPi (table 1 and figure 1). All patients gave written informed consent and the study was approved by the University Health Network Research Ethics Board.
Clinical details of the patients studied
Postoperative MRI and schematic representation of electrode contacts in the internal globus pallidus (GPi). (A) Postoperative MRI image from a cervical dystonia patient. A coronal slice showing the positions of bilateral deep brain stimulation (DBS) electrodes. (B) Locations of GPi contacts. Images were reflected to the same side for ease of comparison. Cpia, internal capsule anterior limb; Cpip, internal capsule posterior limb; Cpig, internal capsule genu; pm, medial pallidum—GPi; pl, lateral pallidum— external globus pallidus; Put, putamen. The non-shaded circles represent the left DBS electrode contacts while the shaded circles represent the right DBS electrode contacts. Numbers inside the circles correspond to the number of each patient, as shown in table 1. DBS contact number: a represents contact 0, b contact 1 and c contact 2. The calibration bars indicate 10 mm. The numbers above the calibration bars indicate the distance (in mm) anterior (+) or posterior (−) to the anterior commissure. Modified and reprinted with permission from Schaltenbrand G and Wahren W. Atlas of stereotaxy of the human brain. New York: Georg Thieme, 1977.
Paradigm and recordings
The study was performed 1–3 days after electrode implantations when the leads were externalised. While sitting in a comfortable armchair, patients performed two tasks of wrist extension movements. With the arm resting on an armrest and the wrist in a relaxed position, the patient performed brisk wrist extension followed immediately by relaxation back to the resting position. In the externally triggered task, the patient attended to a computer screen showing a dark background that turned green for 0.5 s at random intervals between 6 and 10 s. As soon as the green signal was presented, the patient made a brisk right wrist extension movement. In the self-initiated task, the patient made self-initiated brisk right wrist extension movements approximately once every 10 s. Each movement task lasted ∼10–15 min. All patients had no dystonia in the arm and were able to perform the movements without difficulties.
Data recording
The quadripolar DBS electrodes (Medtronic model 3387) contain four contacts numbered 0–3, with contact 0 being the most ventral. The electrodes were 1.27 mm in diameter. Contacts were 1.5 mm in length and spaced 1.5 mm apart. GPi LFP from bilateral DBS electrodes and EEG at Fp1, Fz, Cz, C3 and C4 (international 10–20 system) were recorded with linked ears reference. EMG was recorded from the extensor carpi radialis and flexor carpi radialis muscles to monitor wrist movements. SynAmp amplifiers (Neuroscan Laboratories, El Paso, Texas, USA) were used for all recordings. Impedance was less than 5 kΩ for all electrodes. The sampling rate was 2.5 kHz. Filters were set at 0.05–100 Hz for the scalp and DBS electrodes, and 30–500 Hz for EMG.
Data analysis
Offline analysis was carried out with Scan 4.3 (Neuroscan Laboratories, El Paso, TX) and Brain Electrical Source Analysis (BESA) 5.2 software (MEGIS Software GmbH, Munich, Germany). DBS recordings were transformed into bipolar montage with the adjacent contacts (0–1, 1–2, 2–3) to obtain focal activities for MRP or frequency analyses. For MRP analyses, epochs of 4 s before and 1 s after movement onsets were created and the scalp electrodes were analysed in monopolar montage with linked ears reference. For frequency analyses, bipolar montages that represent the mid-frontal (Cz-Fz), ipsilateral (right) and contralateral (left) sensorimotor cortices (C4/C3-Cz) were used and epochs of 4 s before and 3.5 s after movement onset were examined. Movement onset (0 s) was determined manually using EMG traces. Epochs contaminated with eye or head movement artefacts were rejected.
For MRP analysis, at least 30 artefact free epochs were averaged. Pre-movement MRP or Bereitschaftspotential (BP) onset latencies and amplitudes were obtained from significant waveforms. Paired t tests were used to examine the effects of locations (Cz vs GPi) and laterality (ipsilateral (right) GPi vs contralateral (left) GPi) on BP onset latencies and amplitudes.
For frequency analysis, we used DBS contact pairs that most likely located at the GPi identified by postoperative MRI (figure 1B). Event related desynchronisation (ERD) and event related synchronisation (ERS) were analysed with event related spectral perturbation analysis in BESA to examine movement related frequency changes from 4 to 100 Hz at −4 to 3.5 s. The baseline interval was 4 to 3 s before movement onset. The frequency resolution was 1 Hz and time resolution was 50 ms. ERD/ERS bandpowers were calculated for each patient in frequency bands of interest to check for consistency among patients during wrist movements using bins of 125 ms and a moving window to smooth out short term fluctuations.
Coherences between bilateral GPi, scalp EEG and surface EMG of arm muscles were analysed with complex demodulation implemented in BESA, using the same epochs as ERD/ERS analyses. Permutation statistics were used to test for significant ERD/ERS or coherences. Phase coherences between scalp EEG and the GPi were examined only where significant coherences were found, as indicated by the permutation tests. See supplementary material (available online only) for further details of data analyses.
Postoperative MRI and localisation of DBS electrode
The details of the technique used to localise DBS electrode contacts on postoperative MRI have been published elsewhere.21 Briefly, postoperative axial three-dimensional inversion recovery and T2 weighted images were transferred to a workstation and merged using the FrameLink 4.1 software (Mach 4.1, StealthStation, Medtronic, Surgical Navigation Technologies). Coronal and sagittal planes were reconstructed based on axial images. The anterior and posterior commissures (AC, PC) were targeted in the axial plane and three additional points were plotted in the midline. Thereafter, images were reformatted parallel to the AC–PC plane and orthogonal to the midline. DBS electrodes were visualised in all three planes and the location of each electrode was estimated. For the purpose of our study, we considered the centre of the sphere shaped artefacts as the centre of the contacts.22 Figure 1B shows electrode contacts from which GPi LFP were recorded.
Results
Movement related potentials
Significant pre-movement MRP preceding self-initiated but not externally triggered wrist movements were observed in scalp EEG and in bilateral GPi contact pairs in all patients (figure 2 and table 2). Phase reversals of MRPs between adjacent GPi contact pairs were observed in six of 10 sides studied. Externally triggered movements were associated with potentials only in the movement execution period in the cortex and bilateral GPi (figure 2B). We were not able to distinguish between different subcomponents of the MRP, probably due to the relatively small number of epochs recorded or the lower late MRP amplitude in dystonic patients.3 Table 2 presents the onset latencies and amplitudes of MRP at the GPi and at the Cz electrodes of all patients. The absolute values of MRP amplitudes were used for statistical analyses. The MRP onset latency was −2.5±0.2 s (mean±SD) with an amplitude of −33.2±9 μV at the Cz electrode. The MRP onset latency was −2.2±0.3 s with an amplitude of 2.7±1.3 μV for the ipsilateral GPi and −2.3±0.4 s with an amplitude of 2.5±1.6 μV for the contralateral GPi. Paired t tests showed no significant difference in MRP onset latencies between the ipsilateral, contralateral GPi contact pairs and the Cz electrode, nor differences in MRP amplitudes between the ipsilateral and contralateral GPi.
Examples of scalp and bilateral bipolar internal globus pallidus (GPi) movement related potentials (MRP) recordings. MRP recordings of patient No 4 during (A) self-initiated and (B) externally triggered wrist extension movements. Cz is the vertex scalp electrode, and 0–1, 1–2 and 2–3 represent bipolar montages of the quadripolar deep brain stimulation (DBS) electrodes at the ipsilateral right GPi (RGPi) and contralateral left GPi (LGPi). The lowest traces show the rectified and averaged EMG activity of the right extensor carpi radialis muscle. The horizontal lines represent the baselines. The scalp and GPi recordings show a slow MRP only before self-initiated movements. Potentials were observed during movement executions but not in the pre-movement period in externally triggered movements. The white arrows point to MRP onsets. The black arrows indicate phase reversals between adjacent DBS contact pairs. BP, Bereitschaftspotential.
Onset latencies and maximum amplitudes of scalp and GPi MRP during self-paced wrist movements
Movement related ERD/ERS at the cortex and GPi
A ∼6–28 Hz ERD was found at the mid-frontal (Cz-Fz) and the bilateral sensorimotor cortices (C3/C4-Cz) (figure 3A–F) during both externally triggered and self-initiated movements. A bilateral ∼10–24 Hz beta ERD was observed at the GPi (figure 3G–J). Permutation tests showed that movement related ERD at the mid-frontal, bilateral sensorimotor cortices and bilateral GPi started at movement onset (0 s) and ended at ∼1 s for externally triggered movement (figure 3A,C,E,G,I). For self-initiated movement, ERD at the mid-frontal, bilateral sensorimotor cortices and bilateral GPi started before movement onset at ∼−1 s and ended at ∼1.5 s (figure 3B,D,F,H,J). Movement related ERD was followed by post-movement ERS in the cortices and the GPi. Moreover, a ∼64–68 Hz gamma ERS was observed at the mid-frontal and the contralateral sensorimotor cortices (figure 3A,B,E,F) and the contralateral GPi (figure 3I,J) in both types of movements. For externally triggered movement, the permutation tests showed that this gamma ERS started at 0 s at the mid-frontal and contralateral sensorimotor cortices and the contralateral GPi and ended at ∼0.5 s at the mid-frontal and contralateral sensorimotor cortices and ∼1 s at the contralateral GPi (figure 3A,E,I). For self-initiated movement, the gamma ERS started at ∼−0.5 s and ended at ∼0.5 s at the mid-frontal and contralateral sensorimotor cortices and the contralateral GPi (figure 3B,F,J). To assess consistency among patients, ERD/ERS bandpowers were calculated for the 10–24 Hz and 64–68 Hz bands during self-initiated wrist movements for each patient (figure 3K,L). Both frequency bands displayed highly consistent movement related changes across patients.
Movement related event related desynchronisation (ERD) and event related synchronisation (ERS). Data from five cervical dystonia patients. ERD/ERS (left) and the corresponding permutation tests (right) are shown. The baseline period is from 4 to 3 s before movement onset (−4 to −3 s). ERD/ERS power changes are colour coded in percentage relative to the mean of the baseline period. Blue colour represents ERD while red colour represents ERS. From A to J, the abscissa denotes time in seconds, where the red marker at time 0 represents movement onset and the ordinate denotes frequency from 4 to 100 Hz. In the permutation tests, blue coloured areas indicate significant ERD while red coloured areas indicate significant ERS. (A) ERD/ERS of the mid-frontal cortex (Cz-Fz) during externally triggered and (B) self-initiated wrist extension movements. (C) ERD/ERS of the ipsilateral sensorimotor cortex (C4-Cz) during externally triggered and (D) self-initiated wrist extension movements. (E) ERD/ERS of the contralateral sensorimotor cortex (C3-Cz) during externally triggered and (F) self-initiated wrist extension movements. (G) ERD/ERS of the ipsilateral internal globus pallidus (GPi) during externally triggered and (H) self-initiated wrist extension movements. (I) ERD/ERS of the contralateral GPi during externally triggered and (J) self-initiated wrist extension movements. Event related band power at (K) 10–24 Hz and (L) 64–68 Hz of the contralateral GPi during self-initiated movements for all cervical dystonia patients. Each line represents the result from one patient. The abscissa denotes time in seconds, where time 0 represents movement onset and the ordinate denotes percentage change in power relative to the mean of the baseline period from −4 to −3 s. iGPi, GPi ipsilateral to the movement; cGPi, GPi contralateral to the movement.
Coherence and phase coherence between the cortex and the bilateral GPi
Coherences in the ∼5–18 Hz band between the right and left GPi were observed during the baseline period and after both externally triggered and self-initiated movements (−4 s to 3.5 s). Prominent attenuations of this 5–18 Hz coherence were observed during movements from 0 to ∼1 s (figure 4A,B). Phase analyses showed no consistent phase relationship between the bilateral GPi in the 5–18 Hz band during both externally triggered and self-initiated movements.
Movement related coherences between the cortices and bilateral internal globus pallidus (GPi). Coherences are shown on the left and the corresponding permutation tests on the right. The abscissa denotes time in seconds, where the red marker at time 0 represents movement onset and the ordinate denotes frequency from 4 to 100 Hz. In the permutation tests, red coloured areas indicate significant coherences. (A) Averaged coherences between the bilateral GPi during externally triggered wrist movements and (B) during self-initiated wrist movements. (C) Averaged coherences between the mid-frontal cortex (Cz-Fz) and the contralateral GPi during externally triggered wrist movements and (D) self-initiated wrist movements. (E) Averaged coherences between the mid-frontal cortex (Cz-Fz) and the ipsilateral GPi during externally triggered wrist movements and (F) self-initiated wrist movements. (G) Averaged coherences between the contralateral sensorimotor cortex (C3-Cz) and the contralateral GPi during externally triggered wrist movements and (H) self-initiated wrist movements. (I) Averaged coherences between the ipsilateral sensorimotor cortex (C4-Cz) and the contralateral GPi during externally triggered wrist movements and (J) self-initiated wrist movements. iGPi, GPi ipsilateral to the movement; cGPi, GPi contralateral to the movement.
A ∼64–68 Hz gamma band coherence was found between the mid-frontal, contralateral sensorimotor cortices and the contralateral GPi in both externally triggered (figure 4C,G) and self-initiated movements (figure 4D,H). This gamma coherence started before movement onset at −0.5 s for self-initiated movements and at movement onset (0 s) for externally triggered movements and ended at 0.5 s after onsets of both movements. The average strength of the 5–18 Hz and 64–68 Hz coherence was ∼0.3 in both types of movements. There was no consistent phase relationship in the 64–68 Hz band between the mid-frontal, contralateral sensorimotor cortices and contralateral GPi. Coherence was not observed between the mid-frontal cortex and ipsilateral GPi (figure 4E,F) or between the ipsilateral sensorimotor cortex and the contralateral GPi (figure 4I,J). No coherence was observed between EMG of the forearm muscles and the bilateral GPi or the cortices.
Discussion
We studied GPi MRP and oscillatory activities during both externally triggered and self-initiated movements. The novel findings include the presence of bilateral GPi MRP before self-initiated movements and a resting 5–18 Hz coherence between the bilateral GPi that attenuated with voluntary movements.
Movement related potentials at the human GPi
In both the Cz electrode and GPi, MRPs preceded self-initiated but not externally triggered movements (figure 2), consistent with previous findings that cortical MRPs were found before self-paced movements but not with movements triggered by randomised cues.4 5 The early BP recorded from the Cz electrode prior to self-initiated movements likely reflect activities of the supplementary motor area (SMA), as demonstrated by recordings from subdural electrodes.3 23 The bilateral GPi MRP onset latencies of ∼−2.2 s were similar to findings of MRPs recorded from the bilateral STN in PD patients6 and the ventral thalamus (∼−2 s) in tremor patients,7 24 consistent with the hypothesis that the BG–thalamocortical circuit is active during preparation of self-initiated movements and that the GPi is likely part of the subcortical circuit that contributes to the cortical MRP.6 7 24 25
Movement related ERD/ERS at the human GPi and cortices
In the GPi, bilateral 10–24 Hz beta ERD and contralateral 64–68 Hz gamma ERS during voluntary movements were consistent with our MRP findings and GPi single cell recording showing that the cortico-BG pathway is active during both externally triggered and self-initiated movements.7 26 Moreover, previous studies found that movement related ERD at the cortex, STN and GPi in PD and dystonia patients was <30 Hz and was bilateral, whereas movement related gamma ERS was mainly contralateral and was predominantly in the ∼60–80 Hz range.9 12 19 24 27–30 These consistent findings in different diseases suggest that our ERD/ERS results likely represent common movement related activities at the human cortico-BG pathway. However, the limitations of our study are that it is not possible to compare GPi activities of cervical dystonia patients with that of normal subjects and the recordings of LFP in postoperative patients may be affected by microlesion effects. Because of the small sample size, a larger study is required to confirm the findings.
Bilateral beta ERD in the cortico-BG circuit may represent general movement related functions, particularly the planning and execution of movements. Imagined wrist movements were associated with STN beta ERD and decreased beta band coherence between the cortex and STN, similar to actual wrist movements.30 Our findings of pre-movement beta ERD in the cortices and GPi before self-initiated movements (figure 3A–J) are consistent with the idea that pre-movement ERD in the BG–thalamocortical pathway reflects general motor planning.6 7 24 As beta ERD was also observed during movements, movement related ERD in the human BG is also related to execution of voluntary movements.29 30
Synchronisation of gamma activities between 60 and 80 Hz at both cortical and subcortical levels may facilitate motor processing. During voluntary movements, sharply tuned 60–80 Hz gamma peaks were consistently found in the contralateral primary motor cortex in normal subjects studied with EEG10 and magnetoencephalography11 and in epilepsy patients studied with electrocorticography.10 Gamma ERS was also observed in the contralateral STN and GPi of patients with PD and dystonia during voluntary movements.9 12 We found a similar sharply tuned gamma activity in the contralateral sensorimotor cortex and GPi in both externally triggered and self-initiated movements. Moreover, gamma ERS occurred in the pre-movement period in self-initiated movements (figure 3B,F,J), suggesting that movement related gamma ERS, similar to beta ERD, is also involved in movement preparation in the cortico-BG circuit.
The 64–68 Hz coherence between the sensorimotor cortex and the GPi occurred only contralateral to (figure 4C,D,G,H) but not ipsilateral (figure 4E,F,I,J) to the movement side. Similarly, gamma band coherence was found between the contralateral STN and the cortex during voluntary movements in PD patients.27 Therefore, gamma activities may represent communications between the mid-frontal, contralateral sensorimotor cortex and contralateral BG and play a role in voluntary movements.31 Phase analysis indicated no consistent phase relationship in the gamma band between the cortices and the GPi, which suggests that there may be a third source driving this gamma activity or there is bidirectional communications between the two regions.32
The bilateral 5–18 Hz resting GPi coherence and dystonia
We observed a resting 5–18 Hz coherence between the bilateral GPi (figure 4A,B), which attenuated during movements, suggesting its suppression may be important for the execution of voluntary movements. Because there was no consistent phase relationship between the bilateral GPi, this 5–18 Hz activity was likely transmitted bidirectionally between the two GPi. Anatomically, there is no direct connection between the two GPi, suggesting that other cortical or subcortical structures may mediate the transmission of this coherence between the hemispheres. This 5–18 Hz coherence was not observed between the cortex and GPi (figure 4C–J). Therefore, this coherence is likely transmitted through subcortical structures such as the ventral thalamus.33 A previous study showed bilateral coherence of STN beta (13–35 Hz) activity in PD patients but EEG and coherence between the cortex and STN were not examined.20 The bilateral attenuation during movements suggests that a common process may modulate this coherence bilaterally, allowing information to flow through the bilateral BG during movement executions. Previous studies showed that low frequency (<30 Hz) oscillatory activities in the bilateral STN were driven by the SMA and motor cortical areas34 and stimulations of SMA and motor cortical areas suppressed STN beta frequencies in PD patients.35 Therefore, it is possible that the attenuation of the 5–18 Hz coherence between the bilateral GPi was due to cortical inputs from the SMA and motor cortical areas to the bilateral STN and subsequently to the GPi or to the putamen, external globus pallida and the GPi during movement executions. Further studies are required to investigate the physiological mechanisms responsible for the attenuation of the resting 5–18 Hz bilateral GPi coherence in cervical dystonia patients.
It has been hypothesised that excessive <18 Hz oscillatory activities in the human GPi may abolish reciprocal inhibitions of antagonistic muscles and cause co-contractions, leading to dystonic symptoms.17 19 36 In primary dystonia, excessive 3–12 Hz18 37 or 3–18 Hz19 GPi activities were found to drive activities of dystonic muscles. Similarly, in myoclonus dystonia, 3–15 Hz activities in the GPi were found to drive muscles activities.16 Similar to the excessive beta synchronisation in the BG in PD patients,13 20 the 5–18 Hz synchrony in the bilateral GPi may represent a feature rhythm in dystonia patients. However, we did not observe the coupling of <18 Hz frequencies between the GPi and EMG as previously reported in dystonic patients.16 18 19 38 This may be because we record EMG from forearm muscles which did not have dystonia.
Long term bilateral pallidal DBS may improve dystonia by reducing this <18 Hz activity.19 36 In dystonic hamsters, DBS of the endopeduncular nucleus reduced excessive low frequency (<30 Hz) synchronisation at the endopeduncular nucleus and the substantia nigra pars reticulate.39 These results support the idea that one of the therapeutic mechanisms of long term bilateral GPi DBS may be to reduce the pathological low frequency (<18 Hz) oscillations in the bilateral BG–thalamic network in dystonia patients.33
Conclusions
Activities of the human cortico-BG circuit in the beta and gamma frequency bands are modulated by the preparation of self-initiated movements and the executions of both externally triggered and self-initiated movements. The bilateral beta ERD in the GPi and sensorimotor cortices may play a general role in movement related processes whereas the contralateral gamma ERS may play a more direct role in movement preparation and execution. The 5–18 Hz coherence in the bilateral GPi network is associated with the resting state in dystonia and its reduction may be required for the execution of voluntary movements.
References
Footnotes
Funding This work was supported by the Canadian Institutes of Health Research (grant No CIHR, MOP15128 to RC). EWT was supported by a CIHR Canada Graduate Scholarship Doctoral Award. AML is supported by the Canada Research Chair in Neurosciences and RC is supported by a CIHR-Industry (Medtronic Inc.) Partnered Investigator Award and the Catherine Manson Chair in Movement Disorders.
Competing interests CH received honoraria from and is a consultant of St Jude Medical. EM received honoraria for serving on the educational advisory board for Medtronic, was a consultant for Medtronic Canada and received honoraria for lecturing from Medtronic. AML received honoraria and research grants from Medtronic Inc and St Jude Medical. MH has received research grant support from Medtronic Inc. RC received consulting fees from Medtronic Inc.
Ethics approval This study was conducted with the approval of the University Health Network Research Ethics Board.
Provenance and peer review Not commissioned; externally peer reviewed.