Functional magnetic resonance imaging and transcranial magnetic stimulation: Effects of motor imagery, movement and coil orientation

Abstract

Objective

To compare fMRI activations during movement and motor imagery to corresponding motor evoked potential (MEP) maps obtained with the TMS coil in three different orientations.

Methods

fMRI activations during executed (EM) and imagined (IM) movements of the index finger were compared to MEP maps of the first dorsal interosseus (FDI) muscle obtained with the TMS coil in anterior, posterior and lateral handle positions. To ensure spatial registration of fMRI and MEP maps, a special grid was used in both experiments.

Results

No statistically significant difference was found between the TMS centers of gravity (TMS CoG) obtained with the three coil orientations. There was a significant difference between fMRI centers of gravity during IMs (IM CoG) and EMs (EM CoG), with IM CoGs localized on average 10.3 mm anterior to those of EMs in the precentral gyrus. Most importantly, the IM CoGs closely matched cortical projections of the TMS CoGs while the EM CoGs were on average 9.5 mm posterior to the projected TMS CoGs.

Conclusions

TMS motor maps are more congruent with fMRI activations during motor imagery than those during EMs. These findings are not significantly affected by changing orientation of the TMS coil.

Significance

Our results suggest that the discrepancy between fMRI and TMS motor maps may be largely due to involvement of the somatosensory component in the EM task.

Introduction

Transcranial magnetic stimulation (TMS) is a method that allows for noninvasive stimulation of neurons in localized regions of cortex (Lazzaro et al., 2004). It is widely used as a research tool in neurosciences and therapeutic management of patients with a variety of neuro-psychiatric disorders (Schlaepfer et al., 2003, Tassinari et al., 1990). However, with almost two decades of TMS use since it was introduced by Barker et al. (1985), the exact stimulation site on the cortex remains under debate despite multiple attempts to define it (Epstein et al., 1990, Terao et al., 1998, Thielscher and Kammer, 2002). This stimulation site, or TMS maximum, is the point of maximum electric field, running along the line perpendicular to the center of the figure-of-eight coil (Kobayashi and Pascual-Leone, 2003). Knowledge of TMS maximum is crucial to accurate positioning of the coil in studying normal and pathological cerebral functions.

Functional magnetic resonance imaging (fMRI) has been used to study the cortical effects of TMS noninvasively because of its high spatial and temporal resolution (Brett et al., 2002). fMRI measures the hemodynamic correlates of neural activity (Ogawa et al., 1992) and allows for mapping functional activity and connectivity in humans (Matthews and Jezzard, 2004). In fMRI experiments, fast sequences such as echoplanar imaging (EPI) techniques are used to measure changes in the blood oxygenation level dependent (BOLD) contrast as the result of brain activity (Ramsey et al., 2002) and activation maps are derived from the BOLD images.

Functional neuroimaging techniques such as fMRI and positron emission tomography (PET) have been used to examine cortical activity before, during and after the application of TMS. Some investigators used fMRI to localize areas of cortical activation during task performance, then utilized the cortically active areas as markers for positioning a TMS coil (Neggers et al., 2004). Others simultaneously recorded cortical activity by interleaving TMS and fMRI thus revealing the brain's direct responses to TMS as well as intracerebral functional connectivity of the stimulated areas (Bohning et al., 1999, Paus et al., 1997). The long-term effects of TMS on brain activation were investigated using functional imaging immediately post-stimulation which led to empirical insights into functional cortical plasticity (Siebner et al., 2001).

One consistent finding in the studies that combined fMRI and TMS was the discrepancy of the maps generated by TMS targeting the primary motor cortex and fMRI of brain activity during motor movement (Bastings et al., 1998, Boroojerdi et al., 1999, Bohning et al., 2001, Krings et al., 1997a, Krings et al., 1997b). These authors reported a mismatch of 4–22 mm between TMS centers of gravity (TMS CoG) and fMRI activation maximae. The most recent studies which showed the mismatch of 4.14 (Neggers et al., 2004), 10 (Herwig et al., 2002) and as high as 13.9 mm (Lotze et al., 2003) indicated that the cortical sites corresponding to the scalp TMS CoGs were consistently anterior to the fMRI activation maximae.

One plausible explanation for the TMS CoGs to be located anterior to the fMRI activations may be the presence of a somatosensory component in the BOLD activity during overt movements. In a recent review, Lafleur et al. (2002) compared the BOLD activity during imagined movements (IM activation) to that during the executed movements (EM activation), in studies conducted with fMRI, PET and single photon emission computed tomography (SPECT). Many of these studies demonstrated activation of several cortical regions including the primary motor cortex (M1), supplementary and cingulate motor areas and dorsal premotor cortex during both EMs and IMs (Binkofski et al., 2000, Gerardin et al., 2000, Kuhtz-Buschbeck et al., 2003, Naito et al., 2002, Porro et al., 1996). On the other hand, in addition to M1, EMs activated primary sensory (S1) and sensorimotor (SM) areas, whereas in most studies IMs activated only M1 (Lotze et al., 1999, Porro et al., 2000, Roth et al., 1996).

Another potential explanation for the mismatch between the TMS and fMRI motor maps may be the influence of TMS coil orientation which determines direction of the induced current with subsequent differential affects on underlying neuronal elements (Lazzaro et al., 2004, Sakai et al., 1997). Changing the coil orientation affects MEP amplitudes and stimulation threshold and therefore can result in different TMS maps (Guggisberg et al., 2001).

In the present study, we compared BOLD activations during IMs and EMs with TMS-evoked MEPs of the first dorsal interosseus (FDI) muscle in order to explicitly examine the contribution of the somatosensory component of EMs. In this study, three coil orientations were used, corresponding to anterior–posterior, posterior–anterior and lateral–medial directions of the induced tissue current, and differences between resultant MEP maps were examined.