Localizing Broca’s area for transcranial magnetic
stimulation: Comparison of surface distance
measurements and stereotaxic positioning
Nora Weiduschat, MDa, Birgit Habedank, MDa, Birgit Lampea, Jo ¨rg Poggenborg, MDb,
Alexander Schusterc, Walter F. Haupt, MD, PhDa, Wolf Dieter Heiss, MD, PhDc,
Alexander Thiel, MD, PhDd
aDepartment of Neurology, University of Cologne, Cologne, Germany
bDepartment of Radiology, University of Cologne, Cologne, Germany
cMax-Planck Institute for Neurological Research, Cologne, Cologne, Germany
dDepartment of Neurology, McGill University, Montreal, Quebec, Canada
Precise placement of transcranial magnetic stimulation (TMS) coils over target regions is crucial for
correct interpretation of TMS effects. Modern frameless stereotaxic systems yield high accuracy, but
require extensive equipment and cannot be used in every setting, for example, during functional
The aim of this study was the development of a method for TMS-coil placement based on individual
imaging data without the need for external tracking devices.
We compared coil positioning over Broca’s area using an advanced stereotaxic navigation system with
placement according to the surface distance measurements (SDM) method. By using the SDM-method,
3-dimensional renderings adapted from individual T1-weighted magnetic resonance imaging (MRI)
data were created to identify Broca’s area and Broca’s homologue, respectively, and to define anatomic
landmarks on the skin’s surface. Distances between these landmarks were used to localize the real
target on the individual’s head.
The mean Euclidean distance between surface positions as determined with the two methods was 8.31
mm and the mean difference of estimated virtual electric field intensity at the target point was 7.37 V/m
corresponding to 4.01% of maximum field strength.
This study was supported by the Marga and Walter Boll-Stiftung.
Correspondence: Dr. Nora Weiduschat, Department of Neurology, University of Cologne, Kerpener Str. 62, 50924 Cologne, Germany.
E-mail address: firstname.lastname@example.org
Submitted June 26, 2008; revised September 8, 2008. Accepted for publication September 10, 2008.
1935-861X/09/$ -see front matter ? 2009 Elsevier Inc. All rights reserved.
Brain Stimulation (2009) 2, 93–102
Our findings suggest that, compared with a state-of-the-art frameless stereotaxy system, the SDM-
method yields a reasonable accuracy for positioning of a TMS-coil over Broca’s area in terms of spatial
? 2009 Elsevier Inc. All rights reserved.
rTMS; target localization method; coil placement; neuronavigation; TMS aiming
Mapping of cortical brain functions has been the domain
of neuroimaging techniques such as positron emission
tomography (PET) and functional magnetic resonance
imaging (fMRI). Functional neuroimaging studies can,
however, only visualize brain areas that are involved in a
certain task, but give no information about how essential a
certain region is for the observed task performance.1This
gap can be bridged by transcranial magnetic stimulation
(TMS), a noninvasive method to examine neural processes
by delivering short magnetic pulses penetrating the skull
and thus inducing the depolarization of underlying cortical
neuronal assemblies.2,3Excitability of the cortex can be
either inhibited or facilitated depending on stimulation
parameters.4-7TMS can be applied using single or multiple
pulses and with different intensities and frequencies. The
intensity of stimulation is influenced by the strength of
the magnetic field applied and the excitability of the
cortex.8High-frequency repetitive TMS (. 5 Hz) increases
cortical excitability, whereas stimulation with frequencies
of 1 Hz or lower decrease excitability.9,10
Single TMS pulses are routinely used for neurophysio-
logic examinations, such as analysis of the central motor
conduction time, the cortical silent period, or the motor
threshold.3,11,12Repetitive TMS (rTMS) is commonly used
in cognitive neuroscience,2,8,13but also shows promising
results as a clinical therapeutic tool.7,12,14-16Several studies
have been conducted examining rTMS as a complimentary
therapy, for example, in depression or aphasia.15,17-19For
the latter purpose the investigation of language networks
in patients with left-hemispheric lesions is essential: mech-
anisms of brain plasticity can be evaluated according to
their functional relevance20-22and possible applications of
rTMS can be identified. In this context, also studies in
healthy subjects produce remarkable results.23,24
Accurate positioning of the magnetic coil in regions
where no overt responses such as muscle contractions can
be provoked can be particularly challenging.4,25Determina-
tion of the correct coil position on the subject’s head is
however crucial for reproducible effects, in such a way
that the maximum strength of the magnetic field is deliv-
ered at the target of interest, for example, Broca’s area.26
One of the main procedures currently being used is a opti-
cal tracking method (for example, Surgical Tool Navigator,
Carl Zeiss AG, Jena, Germany) based on frameless stereo-
taxy: a 3-dimensional (3D)-camera system detects light-
emitting diodes (LED) or infrared-reflecting beacons,
which are located on the subject’s head and the coil.4,27,28
A referencing procedure coregisters the positions of the
head and the coil with the cranial magnetic resonance im-
ages in a 3D coordinate system. Thus, the relative positions
of the freely moveable head and coil are registered and vi-
sualized in real time during stimulation.4,28Another
method for navigating the coil according to individual anat-
omy is the use of digitizers, which are based on radio sig-
nals.6This system (Polhemus Isotrak, Polhemus, Inc,
Colchester, Vermont) registers relative positions of land-
marks on the skull and the position of the stimulation
site, which can then be identified in the individual MRI. Be-
fore and after magnetic stimulation, the position of the coil
in relation to the brain is located with the digitizer and vi-
sualized in the MRI.4Alternatively, if magnetic stimulation
is carried out during functional imaging sessions, the coil
can be positioned and held by using integrated systems
that are based on structural images gathered during the
If the precise but expensive instruments necessary for
frameless stereotaxy and other sophisticated systems are
not available or are difficult to apply, different approaches
are necessary. One alternative is to use the International
10-20-system for identification of stimulation sites.29,30
Well known for localization of electroencephalogram
(EEG) electrodes by proportional skull distance measure-
ments, Pascual-Leone et al23described the International
10-20 system as a standardized procedure for coil place-
ment. The main drawback of the latter method is the prob-
lem of negative results: if no stimulation effect is observed,
one would readily conclude that the target is not an elo-
quent region for the given task. A false-negative interpreta-
tion would ensue, however, if the target of interest was
merely missed because of anatomic variability or deform-
ing brain lesions.
In this article, we propose a method based on surface
distance measurements (SDM) as an alternative approach to
the problem of target localization in TMS studies and
compare coil positioning over Broca’s area according to our
method with coil placement using an advanced stereotaxic
navigation system, which also allows the estimation of the
electric field strength at the target site (eXimia NBS,
Nexstim Ltd, Helsinki, Finland). Our method integrates
3D reconstructions of structural neuroimaging data to
94Weiduschat et al
derive individual SDM for localization of the stimulation
site on the subject’s head without the need for external
Methods and materials
Before each neuronavigation session, T1-weighted, high-
resolution magnetic resonance scans were obtained from
six volunteers (three women and three men) without any
history of neurologic or psychiatric disease. Informed
consent was obtained from all subjects. MRI was per-
formed on a 1.5 T Phillips gyroscan scanner (Phillips,
Eindhoven, The Netherlands) producing 160 contiguous
transaxial slices of 1-mm thickness.
was realized with an interactive data language (IDL)-based
visualization software (3D-Tool) that was developed at the
Max-Planck Institute for neurologic research in Cologne,
Germany.31First, 3D surface renderings of the subject’s skin
and brain were generated from the magnetic resonance
images. Precondition for surface reconstructions was that
anatomic structures were segmented as only the transitions
between these structures were included instead of using all
intensities of the volume. The segmentation of cerebral
and noncerebral structures and the visualization of brain
surface renderings was performed automatically based on
thresholding, connected component analysis and morpho-
logic operations. Surfaces were rendered by using volume-
based surface projection and ray-casting algorithms. The
reconstructed skin surface was reproduced as translucent
so that underlying anatomic structures were visible.
Secondly, the pars triangularis of the inferior frontal
gyrus (IFG) was marked by two raters as a target B on both
hemispheresusingcriteria basedon macroanatomy
(Figure 1A). In four subjects, this was performed for both
hemispheres thus yielding two target points for each of
them. In the other two subjects, the IFG could not be iden-
tified on the right hemisphere caused by individual anatomy
or MRI artifacts, respectively. Thus, in some subjects we
marked Broca’s area only, in some we marked both Broca’s
area and Broca’s homologue on the right hemisphere. In the
3D reconstruction, the IFG was identified as the most infe-
rior of the three frontal gyri running an anteroposterior
course on the lateral convexity of the frontal lobe.32Having
identified the pars triangularis, target point B was defined
on the cortical surface in the center of the gyrus. After
positioning the virtual head for a strictly lateral view, the
target point B was shifted onto the head surface along a
line perpendicular to the monitor plane toward the viewer.
Next, the lateral angle of the eye (A) and the center of
the tragus (T) were identified on the virtual skin’s surface
(Figure 1B) and a baseline between A and T was automat-
ically drawn onto the 3D rendering of the head surface that
followed the curvature of the skull.
Finally, the software constructed a line from and
perpendicular to the baseline AT that followed the head
surface above the target point B thus yielding intersection
point S (Figure 1C) and defining the distances AS, TS, and
BS. These connecting lines were projected onto the 3D-
rendered skin creating bent lines along the head surface.
The course of these curves was determined by connecting
all voxels along the distances of AT, AS, TS, and BS and
their length was calculated by adding up the Euclidean
distances between respective voxels along these lines
making use of the Z-buffer information.
With a picture of the virtual head on hand, we marked the
lateral angle of the eye (point A) and the center of the
tragus (point T) on the real subject’s head with a marker
pen. Next, we connected these two points with a line using
a measuring tape and compared the length of this line with
the length of AT on the reconstructed virtual head surface.
of the eye (A) and the tragus (T) were identified on the virtual skin surface (B). Next, the software calculated a line connecting A
and T along the surface and constructed a perpendicular line from the skin surface above B onto the baseline thus yielding inter-
section S (C).
Defining the target for rTMS: The triangular part of the IFG was marked as virtual target B (A). Then the lateral angle
TMS-coil positioning with surface distances95
In every case these lengths were identical. Then, we
measured point S on the connecting line between A and
T according to the virtual measurements AS and TS
(Figure 2). Setting at point S, a straight line perpendicular
to AT was marked onto the head surface. In a last step,
we defined the target point B on this line according to the
distance of BS on the virtual head.
For frameless stereotaxic monitoring (eXimia Navigated
Brain Stimulation NBS, Nexstim), each subject put on
tightly fitting goggles with infrared-reflecting beacons. The
TMS coil was also fitted with infrared-reflecting markers.
We coregistrated the subject’s head to the respective MRI
data in a common reference frame.
First, we defined the nasion and the crus helicis of both
ears as landmarks on the magnetic resonance scans. The
median sagittal line and the transaxial plane tangent to the
upper pole of the bulbi were projected onto the surface of
the forehead and the intersection point of these lines was
defined as the nasion. The auricular landmark was defined
as the point, at which the crus helicis merges in the concha.
Next, we registered the corresponding structures on the
subject’s head by pinpointing these landmarks with a
digitizer pen. Nine additional scalp points were digitized
and thus coregistered by pointing so-called guiding areas
disseminated over the head surface with the digitizer pen
providing a more accurate registration. By moving the
digitizer pen over the scalp and at the same time checking
from the NBS display that the tip of the pen moved
correspondingly over the scalp in the 3D head view, we
verified that the coregistration was coherent.
In one of the subjects registration of the nasion and the
auricular landmarks was unusually difficult, yielding in-
consistent readings of coil position and estimated electric
field. We thus decided not to include these data. As
mentioned before, in two other subjects, the right IFG on
the right hemisphere could not clearly be identified in the
3D reconstruction. Thus, eight measurements in five
subjects were used for analysis.
Placement of the coil on the target marked on the
subject’s skin is always challenging when no neuronaviga-
tion is used. We used a figure-eight-shaped coil, encased
with an even and laterally straight coating and with a
upward handle positioned in the center of the coil, perpen-
dicular to the plane of the coil as well as a fluorescent
marker on the center of the coil plane (Nexstim Focal
Monopulse 50 mm). We then placed the coil with the lower
edge on the subject’s head below the temple with the
marker over the target point and then slowly tilted the coil
so that its entire surface was positioned flat against the
temple. This technique and the special position of the coil
handle allowed us to verify the correct position of the
marker relative to the target as long as possible during the
To assess the accuracy of the SDM method, the 3D
position of the coil relative to the head (SDM position) was
monitored in real time by the eXimia tracking system. At
the same time, the estimated strength of the induced electric
field in the target area was recorded (SDM field). NBS is
based on the assumption that transcranial magnetic stimu-
lation preferentially influences neurons located in the area
in which the induced current is strongest.33The system
displays the cortical area likely to be maximally stimulated
by a specified TMS pulse at the respective coil position and
estimates the strength of the virtually induced electric field
at the target.
Finally, the 3D-defined coil position with the, according
to the NBS-system, maximum obtainable strength of the
electrical field within the target region was recorded (NBS
The coil position was adjusted so that the estimated
strength of the virtual electric field within the target region
was maximized (NBS field).
To assess the accuracy of our procedure, the Euclidean
distance between the SDM and the NBS positions was
calculated. Furthermore, the difference between estimated
field intensities (SDM and NBS fields) was determined, as
well as the intrarater and interrater reliability for the
landmark definition process within the SDM for two
independent raters. Statistical analysis was performed
with the use of SPSS 14.0 (SPSS, Inc, Chicago, Illinois).
Across all subjects the distance ranged from 3.78-15.5
mm with an average distance of 8.31 mm (Table 1). The
surface distance measurements are applied onto the real subject.
This photograph shows how the virtually derived
96Weiduschat et al
interquartile range was 1.98. Figure 3 gives an exemplary
illustration of the relation between the SDM and the
NBS positions in the subject with the smallest distance
(3.78 mm) and Figure 4 shows this relation in the patient
with the largest difference (15.5 mm) (Figures 3 and 4).
The mean difference of estimated electric field intensity
for all subjects and measurement points was 7.37 V/m (with
an interquartile range of 6) corresponding to 4.01% of
maximum field strength (Table 1). As expected, the esti-
mated electric field was more intense when using the
NBS method compared with the SDM method in all
subjects but one. In this subject, the coil could not be
positioned and tilted optimally for NBS because of the
Intrarater reliability was calculated for the definition of
targets and landmarks on the virtual skin and brain surface.
theaxes could be
Mean variability of target definition was 0.63 mm (stan-
dard deviation [SD] 0.90) for the angle of the eye, 0.63
mm (SD 0.79) for the center of the tragus, and 0.48 mm
(SD 0.63) for the center of the triangular part of the IFG,
The mean of the differences between raters (interrater
reliability) was 1.44 mm (SD 0.82) for the eye, 1.52 mm
(SD 0.53) for the tragus, and 0.86 mm (SD 0.62) for
Comparing the distances between the NBS and the SDM
positions, no significant difference could be assessed for
targeting Broca’s area versus targeting Broca’s homologue.
To elucidate whether an increased spatial distance to the
target correlates positively with a decrease of the electric
distance measurements (SDM)
Average distances and electric field differences between navigated brain stimulation (NBS) and software-based individual surface
MeasurementEuclidean distance (mm)
Difference of electric
field intensity (V/m)
Difference of electric
field intensity (%)
Mean 6 SD
8.31 6 3.43
7.37 6 11.13
4.01 6 6.12
SD 5 standard deviation.
and the NBS positions in the subject with the smallest Euclidean
distance (3.78 mm). The anterior pin stands for the coil position
and orientation with the maximally obtainable electric field inten-
sity in the target region, the posterior pin represents the SDM
position. (Measurement 3, Table 1.)
This figure illustrates the relation between the SDM
SDM point and the optimal coil position and orientation in a pa-
tient with the largest difference (15.5 mm). The anterior pin rep-
resents the coil placement at SDM position, the posterior pin the
coil position with the maximum electric field in the target region
(measurement 7, Table 1). On the bottom of the figure one sees
how the maximum electrical field induced (Max E-field), the elec-
trical field at the defined target (E-field at target), and the metric
distance between the maximum electrical field and the target
structure (distance to MRI target) are presented when using the
This 3D reconstruction shows the relation between
TMS-coil positioning with surface distances97
field strength, we conducted a correlation analysis, but
statistical significance was not reached.
With the increasing importance of rTMS as a diagnostic
and therapeutic tool, the precise localization of the target
point of interest (for example, Broca’s area) is essential.
There are three factors that determine the accuracy of
tracking, and target
One of the most critical steps in neuronavigational
approaches is a careful subject-image registration.34,35
Thus, man-made errors seem to be the most common pitfall
in neuronavigational systems.36Reference structures can be
either markers attached to the skin, anatomic landmarks,
the geometric surface of the face, or a combination thereof.
If anatomic landmarks are used for coregistration, charac-
teristic points such as the tragus are identified in image
data as well as on the real patient.37In a study by Scho ¨n-
feldt- Lecuona et al,35the intrasession stability and inter-
session repeatability pooled over all anatomic landmarks
were 1.6 mm and 2.5 mm, respectively. With an application
accuracy of about 3-5 mm,38,39the accuracy of anatomic
The neuronavigational system that was used in this study
does not accept a mean registration mismatch of more than
4 mm, thus the root mean square distance between the
virtual scalp surface and single digitized points on the
subject’s head must be less than 8 mm (manufacturer
The corresponding step to coregistration when using the
SDM method is the correct application of the virtually
measured surface distances onto the real subject’s head
using a marker and a tape measure.
The modern navigation systems offer the advantage to
monitor the 3D position of the unfixed head and the freely
movable coil in real time during TMS application. Refer-
encing and tracking are realized either by using magnetic or
radiofrequency fields, or optical or mechanical systems.
Optical tracking systems measure the 3D locations of
infrared LEDs or infrared-reflecting beacons attached to
the subject‘s head and the coil.40The simultaneous use of
two cameras recording the infrared beams allows the defi-
nition of the object’s 3D position by triangulation.41Eval-
uation of magnetic field (Compass Cygnus-PFS system;
Compass International, Rochester, New York) and optical
tracking (Stealth Station; Medtronic SNT, Louisville, Ken-
tucky) showed a calculated accuracy of about 1-2 mm for
In the NBS system, accuracy of the recorded coil
location with respect to the head tracker is typically below
1.5 mm and is tested within the production process for
every system. The accuracy of the infrared position sensor
unit is below 0.35 mm, accuracy of coil tracker elements,
and digitizer pen and head tracker below 0.25 mm (man-
Maintenance of the coil position during stimulation is
crucial for the correct application of the SDM method as
in contrast to frameless stereotaxy, the coil position is not
constantly monitored. Options for facilitating coil main-
tenance during stimulation when using the SDM method,
are fixation of the coil (for example, with a tripod) and
restraining movement of the subject’s head. In several
PET studies, we maintained the coil position during
stimulation using a lockable pivot arm on a tripod22,43-46
and restrained the subject’s head in the head rest of
the scanner bed with foam padding. In addition, during
and at the end of the session, digital photographs can
be taken of the subject to control for variations of the
Within the NBS stereotaxic system used in this study,
definition of the target point is straight forward and is
simply accomplished by selecting a target voxel by brows-
ing the individual coregistered MRI volume using the
mouse of the NBS computer.
Within the SDM method the process of target definition
does not only comprise the definition of the true target point
(for example, center of pars triangularis of the IFG), but
also the definition of the reference line, determined by the
two reference points (angle of the eye lid and the tragus).
The accuracy with which these three points are defined on
the MRI is obviously crucial for the accuracy of the SDM
For this definition of target and reference points, the
values for intrarater and interrater reliability obtained in
this study range from 0.48-1.52 mm thus demonstrating a
very good reproducibility. Given the fact that the resolution
of the MRI scan is 1 mm in all spatial directions the error
introduced by the target definition process is approximately
of the order of magnitude of the spatial resolution of the
For both methods, target definition can be enhanced by
combining functional imaging data with the structural MRI.
In this study we did not include fMRI or PET for localizing
the TMS target, but referred to cortical anatomy only. For
our purpose, this is not relevant as any arbitrarily chosen
point would have been sufficient as long as it is well
defined. Nevertheless, functional data can be used for target
definition with the SDM method.
98 Weiduschat et al
Accuracy of the SDM method in comparison with
the NBS system
There are three possible parameters that can be used to
compare the accuracy of coil positioning: the metric
distance between target points defined within both systems,
the strength of the induced electrical field at the target
points, and the TMS effect as a more global parameter of
The metric accuracy of the navigation system that was
used in this study lies within several millimeters, which is
compatible to precision of similar systems as reviewed
previously.4,47Comparison of this state-of-the-art frameless
stereotaxy system with our method showed a mean Euclid-
ean distance of 8.31 mm (interquartile range, 1.98) for
positioning the coil over Broca’s area and Broca’s homo-
logue, respectively. Taking into account that the stimulated
brain area is supposed to have a diameter of several centi-
meters depending on stimulation intensity,4,33,48-50the
accuracy of SDM in terms of metric distance to the target
is sufficiently precise.
Because the stimulated brain area is of a certain spatial
extent, it seems obvious that the difference in the strength
of the induced electrical field at the target points can also
serve as an estimate of localization accuracy. The NBS
system used for this study performs the calculation of
strength and distribution of the intracranial electric field on
the basis of the spherical head model.51,52It uses 4000
spheres that are adjusted virtually to the individual head,
producing a geometrically approximated shape of the
head and the brain. Generally, the sphere is an accurate
volume-conductor model of the head if adapted to the
individual curvature of the brain near the cortical area of
interest52as it was performed in this study. In the NBS sys-
tem, the stimulation intensity, coil parameters, and selected
stimulator are also taken into account when estimating the
electric field strength.47However, whether the location of
the stimulating magnetic field is congruent with the neural
tissue affected most by the stimulation remains unclear no
matter which method for coil positioning or electric field
calculation is used. On the basis of this model for estima-
tion of the electrical field strength at the respective targets,
the Euclidian distance of 8.31 mm translates into an esti-
mated loss of electrical field strength of only 4.01%
when positioning the coil over Broca’s area and Broca’s
The third measure of accuracy, which is the more global
parameter of stimulation efficiency, was not the primary
focus of this study, in which we sought to quantify metric
differences and differences in the virtually induced electric
field between the two methods. However, the feasibility and
effectiveness of the SDM method for inducing virtual brain
lesions with rTMS has recently been demonstrated.22,43-46,53
These studies focused on Broca’s area and Broca’s homo-
logue and demonstrated the applicability and effectiveness
of the SDM method in healthy subjects as well as in patients
with brain lesions caused by stroke or glioma. Typical
effects of rTMS over Broca’s area are the induction of a
speech arrest23or increase of verb-generation latencies.53
rTMS over Broca’s homologue may be used as a future ther-
apeutic approach in aphasic patients.14,15,54
Surface distance measurements in comparison
with other methods for localization
Among the nonstereotactic methods for localization of
stimulations sites for TMS the International 10-20-system
has widely been used. The International 10-20 system is
based on bony landmarks of the skull, is easily applicable,
and is cost-effective. However, the individual cortical
anatomy in relation to these landmarks is variable.
With a mean 3D deviation of less than 2 cm to the
designated cortical target,29the accuracy of the Interna-
tional 10-20 system might be sufficient to yield reproduc-
ible TMS interference in healthy subjects, but whenever
one assumes diverging cerebral or cranial anatomy, more
sophisticated systems are needed.29
A recent study compared the positioning accuracy of the
SDM method with the International 10-20 system in
healthy subjects.55The mean distances to the target that
were aimed for were 6.02 mm for the SDM method and
14.13 mm for the International 10-20 system. The latter
result is the same as the results of similar studies in the
literature.29,30,56Thus, it seems that there can be a consid-
erable gain when using the SDM method compared with
the International 10-20 system, which is probably even
more pronounced when patients with brain-deforming
lesions are examined.
Strategies that use frameless stereotaxic neuronavigation
rely on individual structural data, on individual functional
data or group functional data (probabilistic approach). The
latter uses group averages of functional imaging data (PET
and fMRI).28,57This is advantageous in cases in which no
statistically significant cortical activation can be obtained
from single-subject data. A recent study directly compared
several approaches in relation to their distance to the center
of gravity of motor-evoked potentials (MEP-CoG) in the
primary motor area within a 2D grid.56For coregistration,
anatomic landmarks were used. Coil positioning according
to the 10-20 system showed a mean distance to the MEP-
CoG of 12.3 mm (SD 4.4). Neuronavigation based on indi-
vidual structural and functional data resulted in deviations
of 7.8 mm (SD 3.4) and 6.3 mm (SD 2.5), respectively.
The difference between positioning accuracy according to
individual functional data and the probabilistic approach
with a mean distance of 8.6 mm (SD 2.2) was not signifi-
cant. These results, however, were obtained in healthy sub-
jects. In case of pathologic anatomy or pathologically
altered activation patterns, the difference may well become
TMS-coil positioning with surface distances99
Various sophisticated systems have been developed to
improve positioning of TMS coils. However, because of
their specialization, some methods such as stereotaxic
navigation or robotic devices58are quite expensive. The
economically priced integrated system by Bohning et al25
is well designed for fMRI studies and might also be adapted
for PET; however, it is only applicable for studies, in which
magnetic stimulation is conducted during functional
imaging sessions. In studies in which TMS is carried out
frequently and in different locations (for example, in mul-
ticenter trials examining potential therapeutic effects of
TMS), systems providing easy assistance with coil position-
ing without the need for parallel imaging are needed.
Therefore, future developments might aim at more cost-
effective and practical solutions for placement and mainte-
nance of TMS coil positions.
The method of SDM can be seen as an improvement to
conventional non-navigated approaches as it is based on
individual imaging data. In our approach, stereotaxic
neuronavigation is replaced by surface distances, derived
from individual imaging data. Even in healthy subjects this
can be advantageous over function-guided approaches.59
Effects of anatomic variability are fully respected, as dem-
onstrated by a previous study.46
Relevant advantages of the SDM method are its general
practicability and its applicability during trains of rTMS
within the PET or MRI scanner. Independent from equip-
ment other than the stimulator and a measuring tape, the
SDM method can be applied easily in different locations
and situations, once that the SDMs are determined using a
standard PC. Moreover, from an economic point of view,
the SDM method appears to be a reasonable alternative to
cost-intensive neuronavigation systems. Because of the
consideration of individual imaging data, it can even be
applied in patients with brain-deforming lesions.
Compared with high-tech (NBS) and low-tech (Interna-
tional 10-20 system) approaches, the SDM method yields a
sufficient accuracy for TMS coil placement over Broca’s
area and Broca’s homologue. It must, however, be kept in
mind that the result is favored by the fact that for this
special target, the coil can be placed flat against the temple
of the subject and does not need to be tilted. The use of this
method for targeting areas in different cortical regions in
which the skull is more convex may yield different results.
Also, in some other regions different landmarks should be
used as increasing distance to the chosen reference points
also increases the inaccuracy of the surface measurements
on the subject’s head, for example, replacing the eye-tragus
line by the line nasion-inion, and using a perpendicular line
on this midsagittal baseline basically allows access to any
region of the skull.
The coregistration step of stereotaxic navigation can be
very challenging, if magnetic resonance images are
distorted by artifacts. This implies that subsequently the
reliability of the navigation is impaired (as was the case in
one of the subjects in this study).
When using the SDM method, artifacts can also
hamper the definition and reliability of target and refer-
ence points. But as the SDM method is focused on a
smaller area of the head, this will only be the case, if
structures in close proximity to the target or the landmarks
are concerned. In nonimage-based approaches, these
issues do not play a role as they are independent of any
The time needed to localize a target with the SDM
method consists of two steps. The 3D rendering of struc-
tural MRI and subsequent surface distance measurements
take less than 5 minutes. Application of the surface
distances onto the real head requires approximately 60
When positioning the coil using stereotaxic navigation,
the first step is to define (in the reconstructed images) the
reference points for coregistration and the targets on
the cortical surface, which will take less than 5 minutes.
The actual coregistration procedure, however, usually
requires much more time than the analogous step of the
SDM method. The duration depends on the quality of the
MRI (especially the congruence with the whole head of
the subject as the whole head is accounted for during
coregistration) and the experience of the person perform-
ing the procedure. In an optimal case the duration will be
in the range of 3-5 minutes, but can easily become
As the SDM method requires structural images just as
frameless stereotaxy systems do, it is particularly suitable
for studies in which nonimage-based approaches would not
be applicable, such as studies with patients who have brain-
Another advantage of our method is the fact that the
SDMs, once they are determined, can be applied in
different sessions and locations (such as rehabilitation
facilities, scanners, cooperating hospitals) independently
from any devices, except a measuring tape and the stim-
ulator. This is particularly suitable for multicenter studies.
When using NBS, the tracking devices must be available
every time and in each location.
The SDM method is easy enough to learn during one
introductory training session. For using NBS, more time
has to be invested as the single steps are technically more
challenging. In addition, positioning the target onto an
optimal position (defined by proximity to the target or
maximization of the electric field at the target) can be time-
In summary, the SDM method can be more easily
applied and takes less time for implementation than ster-
eotaxic systems and may yield a sufficient accuracy in coil
positioning for most applications of rTMS.
We thank all the subjects for their participation.
100Weiduschat et al
1. Price CJ, Mummery CJ, Moore CJ, Frakowiak RS, Friston KJ. Delin-
eating necessary and sufficient neural systems with functional imaging
studies of neuropsychological patients. J Cogn Neurosci 1999;11(4):
2. Jahanshahi M, Rothwell J. Transcranial magnetic stimulation studies
of cognition: an emerging field. Exp Brain Res 2000;131(1):1-9.
3. Kobayashi M, Pascual-Leone A. Transcranial magnetic stimulation in
neurology. Lancet Neurol 2003;2(3):145-156.
4. Herwig U, Schonfeldt-Lecuona C, Wunderlich AP, et al. The naviga-
tion of transcranial magnetic stimulation. Psychiatry Res 2001;108(2):
5. Pascual-Leone A, Tormos JM, Keenan J, Tarazona F, Canete C,
Catala MD. Study and modulation of human cortical excitability
with transcranial magnetic stimulation. J Clin Neurophysiol 1998;
6. Wassermann EM. Risk and safety of repetitive transcranial magnetic
stimulation: report and suggested guidelines from the International
Workshop on the Safety of Repetitive Transcranial Magnetic Stimula-
tion, June 5-7, 1996. Electroencephalogr Clin Neurophysiol 1998;
7. Wassermann EM, Lisanby SH. Therapeutic application of repetitive
transcranial magnetic stimulation: a review. Clin Neurophysiol 2001;
8. Robertson EM, Theoret H, Pascual-Leone A. Studies in cognition: the
problems solved and created by transcranial magnetic stimulation.
J Cogn Neurosci 2003;15(7):948-960.
9. Berardelli A, Inghilleri M, Rothwell JC, et al. Facilitation of muscle
evoked responses after repetitive cortical stimulation in man. Exp
Brain Res 1998;122(1):79-84.
10. Maeda F, Keenan JP, Tormos JM, Topka H, Pascual-Leone A. Modu-
lation of corticospinal excitability by repetitive transcranial magnetic
stimulation. Clin Neurophysiol 2000;111(5):800-805.
11. Krings T, Chiappa KH, Foltys H, Reinges MH, Cosgrove GR,
Thron A. Introducing navigated transcranial magnetic stimulation as
a refined brain mapping methodology. Neurosurg Rev 2001;24(4):
12. Rossini PM, Rossi S. Transcranial magnetic stimulation: diagnos-
tic, therapeutic, and research potential. Neurology 2007;68(7):
13. Walsh V, Cowey A. Transcranial magnetic stimulation and cognitive
neuroscience. Nat Rev Neurosci 2000;1(1):73-79.
14. Martin PI, Naeser MA, Theoret H, et al. Transcranial magnetic stim-
ulation as a complementary treatment for aphasia. Semin Speech Lang
15. Naeser MA, Martin PI, Nicholas M, et al. Improved picture naming in
chronic aphasia after TMS to part of right Broca’s area: an open-
protocol study. Brain Lang 2005;93(1):95-105.
16. Liepert J. Transcranial magnetic stimulation in neurorehabilitation.
Acta Neurochir Suppl 2005;93:71-74.
17. Avery DH, Isenberg KE, Sampson SM, et al. Transcranial magnetic
stimulation in the acute treatment of major depressive disorder: clini-
cal response in an open-label extension trial. J Clin Psychiatry 2008;
18. O’Reardon JP, Solvason HB, Janicak PG, et al. Efficacy and safety of
transcranial magnetic stimulation in the acute treatment of major
depression: a multisite randomized controlled trial. Biol Psychiatry
19. Stern WM, Tormos JM, Press DZ, Pearlman C, Pascual-Leone A. An-
tidepressant effects of high and low frequency repetitive transcranial
magnetic stimulation to the dorsolateral prefrontal cortex: a double-
blind, randomized, placebo-controlled trial. J Neuropsychiatry Clin
20. Herholz K, Thiel A, Wienhard K, et al. Individual functional anatomy
of verb generation. Neuroimage 1996;3(3 Pt 1):185-194.
21. Thiel A, Herholz K, Koyuncu A, et al. Plasticity of language networks
in patients with brain tumors: a positron emission tomography activa-
tion study. Ann Neurol 2001;50(5):620-629.
22. Winhuisen L, Thiel A, Schumacher B, et al. Role of the contralateral
inferior frontal gyrus in recovery of language function in poststroke
aphasia: a combined repetitive transcranial magnetic stimulation and
positron emission tomography study. Stroke 2005;36(8):1759-1763.
23. Pascual-Leone A, Gates JR, Dhuna A. Induction of speech arrest and
counting errors with rapid-rate transcranial magnetic stimulation.
24. Devlin JT, Watkins KE. Stimulating language: insights from TMS.
Brain 2007;130(Pt 3):610-622.
25. Bohning DE, Denslow S, Bohning PA, Walker JA, George MS. ATMS
coil positioning/holding system for MR image-guided TMS inter-
leaved with fMRI. Clin Neurophysiol 2003;114(11):2210-2219.
26. Andoh J, Artiges E, Pallier C, et al. Modulation of language areas with
functional MR image-guided magnetic stimulation. Neuroimage 2006;
27. Ettinger GJ, Leventon ME, Grimson WE, et al. Experimentation with
a transcranial magnetic stimulation system for functional brain
mapping. Med Image Anal 1998;2(2):133-142.
28. Paus T. Imaging the brain before, during, and after transcranial mag-
netic stimulation. Neuropsychologia 1999;37(2):219-224.
29. Herwig U, Satrapi P, Schonfeldt-Lecuona C. Using the international
10-20 EEG system for positioning of transcranial magnetic stimula-
tion. Brain Topogr 2003;16(2):95-99.
30. Okamoto M, Dan H, Sakamoto K, et al. Three-dimensional probabilis-
tic anatomical cranio-cerebral correlation via the international 10-20
system oriented for transcranial functional brain mapping. Neuro-
31. von Stockhausen HM, Pietrzyk U, Herholz K. Techniken zur Visual-
isierung funktioneller tomographischer Daten in der klinischen For-
schung. Forschung und wissenschaftliches Rechnen. Go ¨ttingen:
Gesellschaft fu ¨r wissenschaftliche Datenverarbeitung 1996;49-60.
32. John JP, Wang L, Moffitt AJ, Singh HK, Gado MH, Csernansky JG.
Inter-rater reliability of manual segmentation of the superior, inferior
and middle frontal gyri. Psychiatry Res 2006;148(2-3):151-163.
33. Thielscher A, Kammer T. Linking physics with physiology in TMS: a
sphere field model to determine the cortical stimulation site in TMS.
34. Steinmeier R, Rachinger J, Kaus M, Ganslandt O, Huk W,
Fahlbusch R. Factors influencing the application accuracy of neurona-
vigation systems. Stereotact Funct Neurosurg 2000;75(4):188-202.
35. Schonfeldt-Lecuona C, Thielscher A, Freudenmann RW, Kron M,
Spitzer M, Herwig U. Accuracy of stereotaxic positioning of transcra-
nial magnetic stimulation. Brain Topogr 2005;17(4):253-259.
36. Spetzger U, Hubbe U, Struffert T, et al. Error analysis in cranial neuro-
navigation. Minim Invasive Neurosurg 2002;45(1):6-10.
37. Eggers G, Muhling J, Marmulla R. Image-to-patient registration tech-
niques in head surgery. Int J Oral Maxillofac Surg 2006;35(12):
38. Woerdeman PA, Willems PW, Noordmans HJ, Tulleken CA, van der
Sprenkel JW. Application accuracy in frameless image-guided neuro-
surgery: a comparison study of three patient-to-image registration
methods. J Neurosurg 2007;106(6):1012-1016.
39. Wolfsberger S, Rossler K, Regatschnig R, Ungersbock K. Anatomical
landmarks for image registration in frameless stereotactic neuronavi-
gation. Neurosurg Rev 2002;25(1-2):68-72.
40. Paus T, Wolforth M. Transcranial magnetic stimulation during PET:
reaching and verifying the target site. Hum Brain Mapp 1998;
41. Willems PW, van der Sprenkel JW, Tulleken CA, Viergever MA,
Taphoorn MJ. Neuronavigation and surgery of intracerebral tumours.
J Neurol 2006;253(9):1123-1136.
42. Mascott CR. Comparison of magnetic tracking and optical tracking by
simultaneous use of two independent frameless stereotactic systems.
Neurosurgery 2005;57(4 Suppl):295-301.. discussion 295–301.
TMS-coil positioning with surface distances101
43. Thiel A, Schumacher B, Wienhard K, et al. Direct demonstration of
transcallosal disinhibition in language networks. J Cereb Blood
Flow Metab 2006;26(9):1122-1127.
44. Winhuisen L, Thiel A, Schumacher B, et al. The right inferior frontal
gyrus and poststroke aphasia: a follow-up investigation. Stroke 2007;
45. Thiel A, Habedank B, Herholz K, et al. From the left to the right: how
the brain compensates progressive loss of language function. Brain
46. Thiel A, Haupt WF, Habedank B, et al. Neuroimaging-guided rTMS of
the left inferior frontal gyrus interferes with repetition priming. Neuro-
47. Hannula H, Ylioja S, Pertovaara A, et al. Somatotopic blocking of sen-
sation with navigated transcranial magnetic stimulation of the primary
somatosensory cortex. Hum Brain Mapp 2005;26(2):100-109.
48. Fox PT, Narayana S, Tandon N, et al. Column-based model of elec-
tric field excitation of cerebral cortex. Hum Brain Mapp 2004;22(1):
49. Thielscher A, Kammer T. Electric field properties of two commercial
figure-8 coils in TMS: calculation of focality and efficiency. Clin Neu-
50. Wagner TA, Zahn M, Grodzinsky AJ, Pascual-Leone A. Three-
dimensional head model simulation of transcranial magnetic stimula-
tion. IEEE Trans Biomed Eng 2004;51(9):1586-1598.
51. Ravazzani P, Ruohonen J, Grandori F, Tognola G. Magnetic stimula-
tion of the nervous system: induced electric field in unbounded, semi-
infinite, spherical, and cylindrical media. Ann Biomed Eng 1996;
52. Ruohonen J, Ilmoniemi RJ. Physical principles for transcranial mag-
netic stimulation. In: Pascual-Leone A, Davey NJ, Rothwell J,
Wassermann EM, Puri BK, editors. Handbook of Transcranial Mag-
netic Stimulation. London: Arnold; 2002. p. 18-29.
53. Thiel A, Habedank B, Winhuisen L, et al. Essential language function
of the right hemisphere in brain tumor patients. Ann Neurol 2005;
54. Naeser MA, Martin PI, Nicholas M, et al. Improved naming after TMS
treatments in a chronic, global aphasia patientdcase report. Neuro-
55. Habedank B, Thiel A, Schuster A, Von Stockhausen HM, Klein JC.
Finding the target for rTMS: an alternative approach to frameless
stereotaxic systems. Neuroimage 2004;22(supplement):57.
56. Sparing R, Buelte D, Meister IG, Paus T, Fink GR. Transcranial mag-
netic stimulation and the challenge of coil placement: a comparison of
conventional and stereotaxic neuronavigational strategies. Hum Brain
Mapp 2007.. xxx.
57. Paus T, Jech R, Thompson CJ, Comeau R, Peters T, Evans AC. Trans-
cranial magnetic stimulation during positron emission tomography:
a new method for studying connectivity of the human cerebral cortex.
J Neurosci 1997;17(9):3178-3184.
58. Lancaster JL, Narayana S, Wenzel D, Luckemeyer J, Roby J, Fox P.
Evaluation of an image-guided, robotically positioned transcranial
magnetic stimulation system. Hum Brain Mapp 2004;22(4):329-340.
59. Denslow S, Bohning DE, Bohning PA, Lomarev MP, George MS. An
increased precision comparison of TMS-induced motor cortex BOLD
fMRI response for image-guided versus function-guided coil place-
ment. Cogn Behav Neurol 2005;18(2):119-126.
102Weiduschat et al