fMRI and TMS: Effects of Motor Imagery,
Movement and Coil Orientation
D.M. Niyazov, A.J. Butler, Y.M. Kadah, C.M. Epstein and X.P. Hu
ISMRM-2005
|
|
Introduction
Transcranial magnetic stimulation (TMS) allows for noninvasive
activation of cortical neurons and is used in neuroscience research
and treatment of patients with neuropsychiatric disorders [1]. The
exact activation site on the cortex remains under debate [2]. We
tested whether somatosensory component in the fMRI activity during
movements may contribute to the discrepancy between the TMS and fMRI
motor maps. We also examined the influence of TMS coil orientation
which determines direction of the induced current.
Methods
Six healthy volunteers participated in this study. During scanning
subjects wore a swim cap with 45 vitamin E capsules corresponding to
a 1X2 cm grid which was used for TMS of the primary motor cortex (M1).
fMRI experiments were performed on a 3T Siemens Trio scanner.
A single-shot gradient-echo EPI sequence was used to acquire T2*
images over 28 oblique axial slices with TR/TE of 2000/35 ms,
matrix of 64X64, FOV 22X22 cm2 and slice thickness of
3mm with no gap. A block-design paradigm consisting of periods of
rest, executed (EM) and imagined (IM) right index finger movements
each lasting 20 seconds, was used. fMRI centers of gravity (CoG)
were defined by the average of the image coordinates of the activated
pixels in M1. A paired t-test was used to determine whether there was
a statistically significant difference between fMRI CoGs during IMs
and EMs. TMS was performed using MAGSTIM 200 with the coil handle
pointing anterior, lateral and posterior. For each coil orientation,
the scalp TMS CoG was determined by the center of mass of the motor
evoked potential distribution on the grid. The TMS CoGs were
transformed into the fMRI coordinate system based on their location
relative to the fiducials on the T1 images [Fig. 1].
The scalp TMS CoGs were then projected towards the cortex, 2 cm along
the line perpendicular to the plane defined by the tangential lines
in the sagittal and coronal views [Fig. 2].
The distance of the projection was chosen based on previous reports
[3].
Fig.1 Vertex (V) has scalp coordinates (0,0) and all fiducials are
separated by 1 cm along the X-axis (X) and 2 cm along the Y-axis (Y).
Red cross (TMS CoG) with scalp (5.5,1) and image (174,139,64) coordinates
is half way between the markers 5 (5,1 scalp and 172,139,61 image) and 6
(6,1 scalp and 180,139,61 image). CS-C central sulcus.
|
Fig.2 Lines in 2 perpendicular planes tangential to the scalp at the
TMS CoG (red cross). The TMS CoG with scalp (5.5,1) and image (174,139,64)
coordinates is projected towards the cortex, 2 cm along the line
perpendicular to the tangential plane defining TMS CoG cortical
projection point (white cross) with coordinates 162,140,74.
Results and Discussion
IM CoGs were 1.17+/-1.16 mm lateral to the cortical projection sites
of TMS CoGs while the EM CoGs were 9.50+/-1.15 mm posterior to
projected TMS CoGs [Fig. 3]. Thus IM CoGs closely
agreed with TMS CoGs while EM CoGs had an approximately 1 cm mismatch.
These results suggest that discrepancy between fMRI and TMS may be due
to a somatosensory component of EM task. In addition, there were no
statistically significant differences between TMS CoGs for the three
coil orientations suggesting that the mismatch between EM and TMS CoGs
was not due to orientation of the field induced by the coil.
Acknowledgment
This work was supported in part by the National Institutes of Health
(RO1EB002009 and HD40984) and Georgia Research Alliance. We thank our
colleagues from Emory University for their invaluable help.
|
|
Fig.3 Same-level slices of EM and IM maps in 6 subjects (A-F). Blue and
yellow are EM and IM maps respectively. Projected TMS CoGs are red crosses.
Black dots are EM and IM CoGs. Left and Ant - left and anterior sides of
the head respectively, in all figures.
References
- Lazzaro VD, Oliviero A, Pilato F, et al. Clin Neurophysiol 2004;115(2):255-266
- Thielscher A, Kammer T. Neuroimage 2002;17(3):1117-30
- Epstein CM, Schwartzberg DG, Davey KR, Sudderth DB. Neurology 1990;40:666-670
|
|
