Abstract
Keywords
Introduction
Cortical reorganization is known to occur during the post-stroke recovery process by various proposed mechanisms including interactions between the primary motor cortex in the ipsilesional and the contralesional hemisphere (Calautti et al., 2001; Duque et al., 2008; Grefkes et al., 2009; Murase et al., 2004; Talelli et al., 2006). In the context of experimental rehabilitative therapies in stroke, the model positing interhemispheric excitability imbalance and an important role of transcallosal connections provides a framework for hypotheses based on concepts: up-regulation of the excitability of intact portions of the ipsilesional motor cortex and down-regulation of the excitability of the contralesional motor cortex. Several ongoing clinical trials are based on this framework. The contralesional motor cortex is presumed to be disinhibited due to the lack of an inhibitory influence from the lesioned motor cortex, while at the same time, it exerts an excessive inhibitory influence on the lesioned motor cortex. Such excessive inhibitory influence on the ipsilesional hemisphere would impair recovery of function (Manganotti et al., 2002; Mansur et al., 2005; Murase et al., 2004).
Repetitive TMS (rTMS) is a non-invasive method for producing potent changes in cortical excitability promoting functional recovery in stroke patients via increased neuroplasticity (Ameli et al., 2009; Bashir et al., 2010; Malcolm et al., 2007; Mansur et al., 2005; Yozbatiran et al., 2009). The down regulation of the contralesional, disinhibited motor regions through neurostimulation is thought to balance the abnormal inhibitory influence on ipsilesional regions (Butefisch et al., 2008; Cramer et al., 2008; Schlaug et al., 2008). Pilot studies, using either rTMS (Ameli et al., 2009; Malcolm et al., 2007; Mansur et al., 2005; Yozbatiran et al., 2009) or transcranial direct current stimulation (tDCS) (Celnik et al., 2009; Fregni et al., 2005 and Hummel et al., 2005) have shown that these approaches can improve motor impairment, at least transiently.
However, reports addressing effects of 1 Hz rTMS on contralesional motor cortex are limited and it is still unclear how rTMS over the contralesional hemisphere might modulate the excitability of the ipsilesional hemisphere. One study only showed that 1 Hz rTMS on contralesional hemisphere decreased short-interval intracortical inhibition (SICI) in primary motor cortex of both hemispheres (Manganotti et al., 2002). In the present study, we aimed to further explore theexcitation/inhibition balance of both hemispheres after a similar procedure.
Although the role of coil positioning techniques has been evaluated in several studies (Herwig et al., 2003; Moisa et al., 2009; Rusjan et al., 2010; Ruohonen et al., 2010; Sack et al., 2007, and Sparing et al., 2008), the effects of navigated rTMS on motor function and neurophysiology in stroke has not been fully explored. In navigated rTMS, the location of stimulation can be defined accurately relative to various anatomical structures, allowing stimulation of brain areas with greater spatial accuracy and reproducibility than conventional TMS method in healthy subjects (Bashir et al., 2011). Most of these studies have shown that neuronavigation increased spatial precision of the stimulation target. Navigated rTMS is thus believed to lead to greater effects and is increasingly used worldwide. Another aim of the study was to verify that navigated rTMS was indeed leading to meaningful improvements of motor function and physiology.
Finally, a third aim of the study was to compare the effects of navigated rTMS between stroke patients and control subjects. Indeed, many rTMS protocols are first tested in healthy volunteers before being translated towards rehabilitation protocols in patients. However, different baseline physiology, e.g., in terms of interhemispheric balance in the case of stroke patients, might modulate rTMS effects which are believed to be strongly state-dependent.
Materials and methods
Subjects
Eight patients (mean age: 67.50±13.77 years) and eight control subjects of comparable age (mean age: 57.38±9.61 years) participated in this study. The investigation was carried out according to the latest version of the Declaration of Helsinki and was approved by the Beth Israel Deaconess Medical Center’s Institutional Review Board. All participants gave their written informed consent prior to enrollment in the study.
The participants exhibited normal cognitive status, as indexed by their Mini Mental State Examination (MMSE; Folstein et al., 2001) scores (normal range: 28– 30). Furthermore, neurological examination of the subjects revealed no abnormal signs that were suggestive of any underlying neurological or psychological condition other than stroke. The participants were not taking any medication known to affect motor cortical excitability at the time of the study and had no contraindications to receive TMS (Rossi et al., 2009). The demographic and clinical characteristics of the sample are shown in Table 1. Further information about the stroke patients is included in Table 2. In each participant, navigated rTMS was performed in a single session lasting approximately 4-5 hours.
Experimental set-up
Prior to TMS, all subjects underwent high-resolution T1-weighted structural MRI scanning. Imaging data were fed to the navigation software (eXimia 3.1, Nexstim Ltd, Helsinki Finland) for automatic 3D brain reconstruction, and the reconstructed images were used to guide navigation and deliver TMS. Nexstim 59 mm (mean winding diameter) focal figure-of-eight TMS coils were used. We used a TMS coil delivering monophasic pulses (type 201514P) for the paired-pulse stimulation paradigm and a TMS coil delivering biphasic pulses (type 201383P) otherwise. During stimulation, surface electromyography (EMG) was recorded and continuously monitored online (ME 6000, Mega Electronics Ltd, Kuopio, Finland) using pre-gelled, disposable Ag/AgCl electrodes (10 mm diameter). Active electrodes were attached to the skin overlying the first dorsal interosseus (FDI) muscle and reference electrodes were placed over the metacarpophalangeal joints. The EMG signals were filtered (8– 500 Hz), amplified, displayed and stored for off-line analysis. The TMS system delivered trigger pulses that synchronized the TMS and EMG systems. The measurement protocol has been previously described in detailed (Bashir et al., 2011; Säisänen et al., 2008).
The subjects sat in a comfortable recliner and held their hands in a supine position on their laps while measurements were recorded. They were instructed to remain silent during the study to avoid speech-induced modulation of cortical excitability. The subjects were also monitored for drowsiness and asked to keep their eyes open throughout the experiment, which was checked regularly by the experimenter. Muscle relaxation was controlled via continuous online visual EMG monitoring, and re-checked offline via visual inspection for absence of background activity.
Determination of the optimal site and motor threshold
In each session, the motor cortical output was mapped carefully to determine the optimal representation of the FDI muscle on both hemispheres in both groups. The site that evoked the MEPs of the highest amplitude (henceforth, the hot spot) was identified. Then, resting motor threshold (RMT) was determined as the minimum TMS intensity that produced at least five motor evoked potentials (MEPs) equal or greater than 50 μV peak-to-peak amplitude out of 10 consecutive stimuli delivered at an inter-stimulus interval of 5– 10 s.
TMS
Single-pulse
Following determination of RMT, 10 single stimuli 5– 10s apart were delivered to the hot spot at an intensity of 120% of RMT to determine baseline MEP amplitude and latency.
Paired pulse paradigm
The paired pulse protocol followed the design by Kujirai et al. (1993) with two conditioning-test pulse interstimulus intervals (ISI): 3 ms to assess SICI and 12 ms to assess ICF. RMT was determined as described above but this time the monophasic pulse was used. The conditioning stimulus was applied at an intensity of 90% of resting motor threshold, the intensity of the test stimulus 120% of resting motor threshold, and these intensities remained constant in all paired-pulse TMS trials. Conditioned MEP amplitude (i.e., recorded after a paired-pulse) was compared to unconditioned MEP amplitude (i.e., recorded after a test stimulus alone). The interval between pairs of stimuli was at least 4– 7s to avoid carry-over effects.
Cortical silent period
The contraction force was measured using a pinch grip simultaneously with both hands in a series of brief (4– 7s) maximal isometric efforts (maximum voluntary contraction, MVC). To assess the ipsilateral and the contralateral silent period (SP), force level was set to 20% of MVC. The muscle contraction was controlled by the subject with online visual force monitoring at the 20% MVC target. TMS-pulses were delivered during active tonic muscle contraction of approximately 4– 7 s duration, and the subjects were instructed to maintain muscle contraction for at least 2 s after the stimulus was delivered. The TMS intensity was set at 120% of the resting motor threshold.
rTMS intervention
All subjects received a train of 1200 stimuli of 1 Hz rTMS at 90% RMT to the right M1 in normal subjects, and the contralesional M1 in stroke patients, delivered over the optimal site for MEPs in the contralateral FDI muscle (hot spot). Repetitive TMS in the 1 Hz range has previously been shown, in most subjects, to suppress excitability of the motor cortex in the targeted hemisphere and increase excitability in the contralateral motor cortex (reviewed by Fregni and Pascual-Leone, 2007; Kobayashi and Pascual-Leone, 2003). In the navigated rTMS intervention procedure, the cortex and hot spot were visualized in real-time by the navigated brain stimulation software to aid the delivery of rTMS (Fig. 1B). The coil was hand-held, and no mechanical coil support was employed during the stimulation procedures, allowing the experimenter to continually steady the coil within the target boundaries across the procedure.
Behavioral tasks
To assess the functional significance of possible neurophysiological differences before and after rTMS, the participants completed a battery of tests to evaluate the motor function of their right and left hands. The behavioral tasks were administered immediately before the neurophysiological measurements, preceding the rTMS intervention, and the same tasks were completed immediately after the neurophysiological measurements following rTMS intervention. The tasks were presented in the order in which they are described below, with the right hand tested before the left hand (Fig. 1A).
Nine-hole peg test
The time required to insert nine pegs into a pegboard and then remove the pegs from the pegboard was recorded across 3 trials. A 1-minute break was given between each trial.
Pinch grip strength
Average pinch grip strength (kg force) was measured according to a previously described protocol that exhibits good validity and test-retest reliability (Mathiowetz et al., 1984). The subjects grasped a pinch key with the pad of their thumb opposed against the lateral aspect of the middle phalanx of the index finger. For each assessment, three pinch grip trials, lasting 10 seconds each, were performed and averaged. One minute of rest was allowed between trials.
Finger tapping task
The subjects were asked to tap a button with their index finger of each hand as quickly as possible for the duration of 10 seconds. The tapping assessment comprised 5 trials. A 15-second rest period was allowed between most trials, and there was a 1-minute break between the third and fourth trials.
Simple reaction time task
The subjects were instructed to press the space bar with their index finger as fast as possible in response to an on-screen stimulus, and the time (ms) between stimulus onset and response was recorded. For each assessment, two trials of 20 stimuli each were presented to the participant.
Data and statistical analyses
To determine the MEP in response to TMS, continuous EMG data were sampled in 350 ms epochs, 50 ms before and 300 ms after each TMS pulse. FDI muscle responses (latency and MEP) were analyzed using MegaWin software (Mega Electronics Ltd, Kuopio, Finland). Corticospinal excitability was assessed by measuring the peak-to-peak amplitude of MEPs in the contralateral FDI muscle in response to single TMS pulses. To minimize the variability of TMS-induced single-pulse responses, the largest and smallest amplitude MEP responses were excluded from the analysis. The baseline and post-rTMS MEP amplitude and latencies were determined as the mean response across eight trials for each subject. For SICI and ICF, the size of the conditioned MEPs was expressed as a percentage of the unconditioned MEPs.
All behavioral measures were assessed in each hand both before and after rTMS and averaged across all of the trials performed by each participant.
All statistical tests were two-tailed, with statistical significance defined at p < 0.05. All statistical analyses were performed using SPSS 16.0 software (SPSS Inc., Chicago IL).
The effects of navigated rTMS were evaluated through a repeated measures analysis of variance (ANOVA) with the HEMISPHERE (stimulated and non-stimulated) as a within-subjects factor and the GROUP (stroke patients and healthy controls) as a between-subjects factor.
Results
The subjects did not report any adverse effects during the course of the study.
Corticospinal excitability measures
Cortical excitability measures indicate the relative changes (Δ= [post-pre]/pre) of all the neurophysiological parameters recorded for each hemisphere and each group.
Resting motor threshold
Figure 2 shows the relative changes, induced by rTMS in corticospinal excitability measured using the RMT for each hemisphere and each group. The ANOVA demonstrated a significant main effect of HEMISPHERE (F = 5.020, p = 0.047). RMT in the stimulated hemisphere was increased following rTMS, whereas it decreased in the non-stimulated hemisphere. However, the ANOVA did not reveal any significant effect of the GROUP factor, and the interaction between the HEMISPHERE and GROUP was not statistically significant (all p > 0.05).
Motor evoked potentials
Figure 3 presents the relative changes, induced by rTMS in corticospinal excitability (in this case MEP amplitude) for each hemisphere and each group. The ANOVA shows a main effect of HEMISPHERE. (F = 35.20, p < 0.001). The amplitude of the MEPs evoked by single-pulse over the stimulated hemisphere was decreased after rTMS, whereas the amplitude of the MEPs evoked by single-pulse over the non-stimulated hemisphere was increased. The ANOVA did not show any effect of the GROUP factor (p > 0.05). The interaction between the HEMISPHERE and GROUP was significant (F = 6.58, p = 0.026). Post-hoc testing revealed that the increase in MEP size over the non-stimulated hemisphere following rTMS was larger for patients than for healthy controls (p < 0.05). It should be noted, that on average the baseline (pre) MEP amplitude was slightly greater in the controls than in patients. This difference was significant for the non-stimulated hemisphere.
The MEP latencies elicited by TMS at 120% of the RMT in the relaxed muscle were also measured. The ANOVA showed a significant effect of HEMISPHERE (F = 6.769, p = 0.025). The latency of the MEPs evoked by single-pulse over the stimulated hemisphere was slightly increased after rTMS, whereas the latency of the MEPs evoked by single-pulse over the non-stimulated hemisphere was decreased. The ANOVA did not detect any significant effect of the GROUP factor, and the interaction between the HEMISPHERE and GROUP was not statistically significant (all the p > 0.05).
Cortical silent period
A stimulation intensity of 120% of the RMT was applied while sustaining 20% of MVC before and after TMS. The ANOVA showed that the HEMISPHERE and GROUP had no significant main effect on the relative changes observed in the cortical SP. There was no significant interaction between the HEMISPHERE and GROUP (p > 0.05).
Intracortical inhibition/facilitation
Figure 4 displays the relative changes, induced by rTMS in SICI and ICF for each hemisphere and each group.
With respect to SICI, the ANOVA did not reveal any significant effect of the HEMISPHERE and of GROUP and the interaction between the HEMISPHERE and GROUP was not statistically significant (p > 0.05). Repetitive TMS increased SICI (experimental values became more negative) regardless of the HEMISPHERE and GROUP.
For ICF, the ANOVA showed a significant main effect of the HEMISPHERE (F = 7.084, p = 0.022). ICF decreased in the stimulated hemisphere after rTMS was applied whereas it increased in the non-stimulated hemisphere. The ANOVA did not reveal any effect of the GROUP factor and the interaction between the HEMISPHERE and GROUP was not statistically significant (p > 0.05).
Behavioral outcome measures
Motor function
For the 9-hole peg task, the ANOVA showed a significant effect of the SIDE (F = 6.16 p = 0.03, Fig. 5, Table 3). Performance improved (i.e., the time taken to complete the task decreased) in the hand ipsilateral to the stimulation (the affected hand in patients) but not for the hand contralateral to the stimulation. The ANOVA did not show any significant effect of the GROUP factor, and the interaction between the SIDE and GROUP was not statistically significant (all p > 0.05, Table 3).
For the pinch grip strength task, the ANOVA showed a significant effect of the SIDE (F = 38.48, p < 0.000). Muscle strength increased for the hand ipsilateral to the stimulation (the affected hand in patients), but not for the hand contralateral to the stimulated side. The effect of the GROUP was not statistically significant (p > 0.05). The interaction between the SIDE and GROUP was significant (F = 17.26, p < 0.02). Post-hoc testing showed that the effect of stimulation was stronger for patients than for controls (p = 0.002, Fig. 6, Table 3).
For the finger-taping task, the ANOVA showed a significant effect of the SIDE (F = 56.72, p < 0.000). Finger tapping speed increased in the hand ipsilateral to the stimulation (the affected hand in patients), but not in the hand contralateral to the stimulation. The effect of the GROUP was not statistically significant (p > 0.05). The interaction between the SIDE and GROUP was significant (F = 39.88 p = 0.000)Post-hoc testing showed that these effects were greater for patients than for controls (p < 0.03, Fig. 7, Table 3).
For the simple reaction time task, the ANOVA showed a significant effect of the SIDE (F = 7.363 p = 0.02). Reaction time performance improved (i.e., reaction time decreased) for the hand ipsilateral to the stimulation (the affected hand in patients), whereas performance decreased (reaction time increased) for the hand contralateral to the stimulation. The ANOVA did not show any significant effect of the GROUP factor, and the interaction between the SIDE and GROUP was not statistically significant (all p > 0.05, Fig. 8, Table 3).
Discussion
In this study, we found that navigated rTMS at 1 Hz over the contralesional motor cortex (M1) reduced the corticospinal excitability of this region, while increasing in the ipsilesional hemisphere, corresponding with improvement of the motor function of the affected hand in stroke patients compared to healthy controls. Of note, significant ICF increment was observed in the ipsilesional motor cortex following rTMS in both stroke and control groups.
Post-stroke motor recovery relies on cortical reorganization of the ipsilesional and contralesional hemispheres through disinhibition of the intact motor cortex via transcallosal fibers (Tombari et al., 2004). The speculative notion is that the more severely a unilateral hemisphere is damaged, the more distinct the trans hemispheric disinhibition is. Therefore, significant contralesional activation indicates more severe damage and poorer recovery. Although the roles played by ipsilesional and contralesional hemispheres in post-stroke recovery remains controversial, previous studies have shown that increased recruitment of the ipsilesional hemisphere and reduced recruitment of the contralesional hemisphere relate to improved motor recovery (Calauttite al., 2001; Cao et al., 1999 and Carey et al., 2005; Cinzia, et al., 2007; Cramer et al., 1997, 2004; Coupar et al., 2012; Johansen-Berg et al., 2002; Lioumis et al., 2012; Stinear, 2010; Zemke et al., 2003;).
Several studies have suggested that the down regulation of the contralesional motor cortex results in an improvement in the motor function of the affected hand in chronic stroke (Avenanti et al., 2012;Hummel et al., 2005; Kobayashi et al., 2004; Maeda et al., 2000, Mansur et al., 2005; Pal et al., 2005; Ziemann, 2005). Behaviorally, we also found that navigated rTMS led to improvements in all of the tasks performed with the hand ipsilateral to the stimulated side. In the grip strength and finger tapping tasks, these effects were greater in stroke patients than in controls. Finger tapping is reported as an assay of human voluntary motor control in healthy adults and in stroke recovery (Calautti et al., 2010; Shimoyama et al., 1990 and Pineiro et al., 2011). Here we show an increase in this ability in the affected hand of stroke patients, and importantly, no reduction in performance of the unaffected hand; seldom assessed in protocols aiming to reduce corticospinal excitability in the unaffected hemisphere. These results are consistent with the neurophysiological finding of a greater interhemispheric impact in patients than in controls. Our results expand prior findings demonstrating that suppression of primary motor cortex cortical excitability using 1 Hz rTMS may result in faster serial button pressing with the ipsilateral hand in healthy humans (Bashir et al., 2011; Kobayashi et al., 2003). Furthermore, 1 Hz rTMS applied over the right primary motor cortex has been found to modify several kinematic parameters in a sequential finger opposition task performed with the ipsilateral hand (Avenanti et al., 2012).
Regarding changes in corticospinal excitability, previous studies showed that 1 Hz rTMS led to a reduction of MEP amplitude of stimulated primary motor cortex with increased MEP amplitude of the non-stimulated motor cortex in stroke (Fregni et al., 2005; Takeuchi et al., 2008) as well as in healthy controls (Gugino et al., 2001; Heide et al., 2006), which was similar to our study. As described above, this finding also follows the mechanism of reduced disinhibition of the ipsilesional hemisphere, plausibly by the disruption of transcallosal inhibition. When considered in stroke, disinhibition in the ipsilesional hemisphere is known to be prominent in early period of stroke (Cicinelli et al., 2003; Liepert et al., 2007; Manganotti et al., 2002) whereas ICF has been reported to be normal both during the first 2 weeks (Liepert et al., 2007; Manganotti et al., 2002; Shimizu et al., 2002) and in the chronic stage (Butefisch et al., 2003, 2005; Swayne et al., 2008). Our results regarding measurements of excitatory and inhibitory pathways showing increased ICF in the ipsilesional hemisphere would support this finding, which might have related to facilitation due to the activation of glutamatergic interneurons (Cicinelli et al., 2003). Although the results of studies investigating the clinical implications of the intracortical excitability of the contralesional hemisphere are controversial, increased intracortical excitatory activity has been demonstrated in the contralesional hemisphere of recovering stroke patients (Bütefisch et al., 2003; Lioumis et al., 2012; Talelli et al., 2006). Due to small sample size, it would be difficult to draw inference from this result, however it could be possible that increased ICF might be related to the patient characteristics of this study, including good motor recovery. The integrity of SICI in contralesional hemispheres after rTMS was also found to predict good motor recovery, because it reflects the maintenance of transcallosal inhibition by ipsilesional hemispheres (Manganotti et al., 2002; Liepert et al., 2007), however, we couldn’t find some difference when compared to controls, although it was relatively decreased in the patients group. The reason for this might be because of the difference of the time duration from onset of stroke, which was measured in chronic stage in this study.
We did not detect any effects of rTMS on the cortical silent period (CSP). However, our assessment of the CSP could be improved by more quantitative approaches, and in any case, the dissociation in the neurobiological substrates of the CSP and other measures of corticospinal excitability have been repeatedly supported.
Navigated rTMS allows stimulation of brain areas with greater spatial accuracy and reproducibility than conventional TMS method in healthy subjects (Bashir et al., 2010). Neuronavigation was shown to improve accuracy and reproducibility; however the superiority of this targeting method has not yet been tested in the rehabilitation setting for stroke.
The most likely interpretation of this finding is that rTMS over the lesioned M1 reduced the excessive inhibition over the contralateral motor cortex. This disinhibition of the lesioned motor cortex may partly contribute to the functional improvement in the affected hand by unmasking the latent networks (Kobayashi et al., 2003). However, we did not observe any change in the SICI of the ipsilesional hemisphere following rTMS. This hypothesis needs to be investigated using a larger number of stroke patients.
Several studies have suggested that the up regulation of the lesioned motor cortex results in an improvement in the motor function of the affected hand in chronic stroke (Avenanti et al., 2012;Hummel et al., 2005; Kobayashi et al., 2004; Maeda et al., 2000; Mansur et al., 2005; Pal et al., 2005;Ziemann, 2005). In the present study navigated rTMS at 1 Hz induced an increase in pinch force. Therefore, rTMS may be important in the rehabilitation of patients after stroke, and imparting additional motor training while the changes are being generated by rTMS at 1 Hz and conducting rTMS cumulatively would sustain the effect of rTMS and improve the function of the affected hand.
In conclusion, our results demonstrate that navigated rTMS over the contralesional hemisphere can lead to an improvement in the motor function in the affected hand of chronic stroke patients. These findings may be pertinent to the design and optimization of neurorehabilitation strategies for patients after stroke. However, the present study has several limitations that should be considered when interpreting the results. First, we compared the stroke condition with healthy subjects (and not with a sham condition) making it difficult to draw inference about rTMS effect directly.
Second, we performed this TMS study only once, although previous studies have shown that there is an evolution of brain activation over time. Therefore, a longitudinal study would have undoubtedly benefited our predictions of clinical outcomes. Third, the subject group was not homogeneous with respect to stroke type. Moreover, the time since the occurrence of stroke varied among subjects (Table 1). This type of subject heterogeneity may have obscured the interpretation of the results. Fourth, there is limited power in the current study because of its small sample size. Despite the above limitations, we suggest that the integrity of the corticospinal output and normal inhibitory and excitatory intracortical processes, as demonstrated by navigated TMS in healthy subjects, may be relevant to the reorganization process required for motor recovery following stroke. Future studies of NIBS in healthy subjects should take caution speculating on likely effects for patients, because patients may have heightened susceptibility for improvement as this study shows.
Conflict of interest statement
Dr. Pascual- Leone serves on the scientific advisory boards for Nexstim, Neuronix, Starlab, Neuroelectrics, Axilum Robotics, Magstim, and Neosync; and is listed as an inventor on several issued and pending patents on the real-time integration of transcranial magnetic stimulation (TMS) with electroencephalography (EEG) and magnetic resonance imaging (MRI).
Footnotes
Acknowledgments
This study was supported by grants from the National Center for Research Resources, including the Harvard-Thorndike General Clinical Research Center at BIDMC (NCRR MO1 RR01032) and the Harvard Clinical and Translational Science Center (UL1 RR025758), as well as NIH grants K24 RR018875, R01HD069776 (DE and APL), and grant support from Nexstim Inc. and the Berenson-Allen Foundation to A.P.-L. The content of this manuscript is solely the responsibility of the authors and does not necessarily represent the official views of Nexstim Inc., the Berenson-Allen Foundation, the National Center for Research Resources or the National Institutes of Health. MV was supported by the Fyssen Foundation, France. SB was supported by grants from international research group (IRG14-26) from King Saud University, Saudi Arabia. The authors thank Andrea Vatulas,Cara Burzynski and Ann Connor for their administrative help.