Repetitive Transcranial Magnetic Stimulation - Am MIT spezialisierte KI-Forschung

Repetitive Transcranial Magnetic Stimulation
Dissociates Working Memory Manipulation
from Retention Functions in the Prefrontal,
but not Posterior Parietal, Kortex

Bradley R. Postle, Fabio Ferrarelli, Massihullah Hamidi,
Eva Feredoes, Marcello Massimini, Michael Peterson,
Andrew Alexander, and Giulio Tononi

D
Ö
w
N
l
Ö
A
D
e
D

F
R
Ö
M

/
J

F
/

T
T

ich
T
.

:
/
/

D
H
T
Ö
T
w
P
N
:
Ö
/
A
/
D
M
e
ich
D
T
F
P
R
Ö
R
M
C
.
H
S
ich
P
l
v
D
e
ich
R
R
e
C
C
H
T
.
A
M
ich
R
e
.
D
C
u
Ö
M
Ö
/
C
J
N
Ö
A
C
R
N
T
ich
C
/
A
e
R
–
P
T
D
ich
C
1
l
8
e
1
–
0
P
1
D
7
F
1
/
2
1
8
1
/
9
1
3
5
0
6
/
0
1
3
7
1
Ö
2
C
/
N
1
2
7
0
5
0
5
6
9
1
4
8
8
/
1
J
0
Ö
1
C
7
N
1
.
2
2
P
0
D
0
6
B
.
j
1
G
8
u
.
e
1
S
0
T
.
Ö
1
N
7
0
1
7
2
S
.
e
P
P
D
F
e
M
B
B
j
e
R
G
2
u
0
e
2
S
3
T

/
J

Ö
N

1
8

M
A
j

2
0
2
1

Abstrakt

& Understanding the contributions of the prefrontal cortex
(PFC) to working memory is central to understanding the neu-
ral bases of high-level cognition. One question that remains
controversial is whether the same areas of the dorsolateral PFC
(dlPFC) that participate in the manipulation of information in
working memory also contribute to its short-term retention
(STR). We evaluated this question by first identifying, mit
functional magnetic resonance imaging (fMRT), brain areas
involved in manipulation. Nächste, these areas were targeted with
repetitive transcranial magnetic stimulation (rTMS) while sub-
jects performed tasks requiring only the STR or the STR
plus manipulation of information in working memory. fMRT
indicated that manipulation-related activity was independent

of retention-related activity in both the PFC and superior
parietal
lobule (SPL). rTMS, Jedoch, yielded a different
pattern of results. Although rTMS of the dlPFC selectively
disrupted manipulation, rTMS of the SPL disrupted manipu-
lation and STR to the same extent. rTMS of the postcentral
gyrus (a control region) had no effect on performance. Der
implications of these results are twofold. In the PFC, they are
consistent with the view that this region contributes more
importantly to the control of information in working memory
than to its STR. In the SPL, they illustrate the importance
of supplementing the fundamentally correlational data from
neuroimaging with a disruptive method, which affords stron-
ger inference about structure–function relations. &

EINFÜHRUNG

One perspective on the role of the prefrontal cortex (PFC)
is that it supports short-term retention (STR, alternatively
referred to as ‘‘storage’’ or ‘‘maintenance’’) functions (z.B.,
Courtney, 2004; Leung, Seelig, & Gore, 2004; Goldman-
Rakic & Leung, 2002; Pessoa, Gutierrez, Bandettini, &
Ungerleider, 2002; Constantinides, Franowicz, & Goldman-
Rakic, 2001; Funahashi, Chafee, & Goldman-Rakic, 1993),
and can thus be viewed as supporting the buffers of
a multiple-component working memory system such as
that proposed by Baddeley (2000) and Baddeley and Logie
(1999). An alternative perspective denies an important role
in STR for the PFC (z.B., Lebedev, Messinger, Kralik, &
Wise, 2004; Petride, 2000; D'Esposito & Postle, 1999),
ascribing these functions instead to activity in the non-
PFC regions that have evolved to accomplish sensory-,
representation-, and action-related functions (z.B., Postle,
2006; Jonides, Lacey, & Nee, 2005; Pasternak & Greenlee,
2005; Theeuwes, Olivers, & Chizk, 2005). One implica-
tion of the latter view is that working memory may be

University of Wisconsin

better understood as an emergent property of the mind
and brain, rather than as a system that can be readily
localized (as can, z.B., vision to the occipital cortex or
skeletomotor control to the precentral gyrus).

One way to operationalize working memory retention
processes is by varying the number of items that must
be retained on different trials—brain regions whose
delay-period activity is sensitive to such variations in
memory load are presumed to contribute to retention-
related processes. To date, the functional magnetic re-
sonance imaging (fMRT) literature on the load sensitivity
of the dorsolateral PFC (dlPFC) delay period activity
for verbal stimuli has been mixed, with some studies
finding evidence for (Narayanan et al., 2005; Zarahn,
Rakitin, Abela, Flynn, & Stern, 2005; Veltman, Rombouts,
& Dolan, 2003), and some finding evidence against
(Feredoes & Postle, 2005; Postle, Berger, & D'Esposito,
1999; Rypma & D'Esposito, 1999). The same study by
Postle et al. (1999) also reported evidence that some
load-insensitive regions of the dlPFC were nonetheless
sensitive to the requirement to reorder the contents
of working memory during the delay period—an oper-
im Gegenzug,
ationalization of manipulation. These results,

D 2006 Massachusetts Institute of Technology

Zeitschrift für kognitive Neurowissenschaften 18:10, S. 1712–1722

have been challenged by Veltman et al. (2003), WHO
reported that ‘‘maintenance’’ (what we refer to in this
article as retention) and manipulation ‘‘activate virtually
identical systems’’ (P. 247). Clearly, the neuroimaging
literature related to this issue is at an impasse.

The present study was intended to address the spe-
cific question of whether STR and manipulation func-
tions colocalize in the dlPFC and to do so with a method
that would support stronger inference than have the
neuroimaging studies performed to date. This approach
entailed supplementing fMRI with repetitive transcranial
magnetic stimulation (rTMS), a method that produces
the temporary disruption of the function of a local re-
gion of cortex. Although fMRI data are limited to testing
hypotheses about brain–behavior correlations, rTMS can
address hypotheses about the necessity of a brain area
to a particular aspect of cognitive performance ( Walsh
& Pascual-Leone, 2003). An additional goal of this study
was to broaden the investigation of manipulation in
working memory beyond the PFC, to reflect the general
consensus that most examples of high-level cognition, In-
cluding the control of working memory, are supported
by broadly distributed networks that extend beyond this
one brain region. Our two-step procedure entailed first,
acquiring fMRI data while subjects performed delayed
recognition with different loads and different manip-
ulation requirements, und zweitens, delivering rTMS to
fMRI-identified areas of the dlPFC and superior parietal
lobule (SPL) with rTMS while the same subjects per-
formed the same task. (The SPL has also been implicated
in executive control, z.B., Garavan, Ross, Li, & Stein, 2000,
and has shown manipulation-related activity in previous
Studien, unpublished observation). We predicted that
delay-period rTMS of manipulation-sensitive regions of
the dlPFC and SPL would produce selective deficits on
trials requiring the manipulation, but not the simple re-
Aufmerksamkeit, of items in working memory.

METHODEN

Subjects

Der 12 Erwachsene (6 men and 6 Frauen; Durchschnittsalter = 22 Jahre,
SD = 2.7) whose data are presented here had no psy-
chiatric or neurologic disorders, as determined by phys-
ical examination, a structured psychiatric diagnostic
screening interview (Mini-International Neuropsychiatric
Interview; Sheehan et al., 1998), and a mood assessment
(Hamilton Depression Rating Scale; Hamilton, 1960), alle
administered by a psychiatrist.

Behavioral Task

The task, requiring delayed recognition of item position,
was identical to that used by Postle et al. (1999). Jede
trial began with the simultaneous presentation of two or
five consonant letters (all in a single row), followed by

instructions (‘‘forward’’ or ‘‘alphabetize’’), followed by
an 8-sec delay period, followed by a memory probe
comprising an item from the memory set and a digit.
On forward trials, subjects were to retain a memory of
the two or five letters in the order in which they were
vorgeführt. On these trials, the probe digit represented
(with p = .5) the ordinal position in which the probed
letter had appeared in the initial stimulus display. An
alphabetize trials, subjects were to reorder the letters
into alphabetical order. On these trials, the probe digit
represented (also with p = .5) the alphabetical position
of the probed letter with respect to the other four letters
in the memory set.

D
Ö
w
N
l
Ö
A
D
e
D

F
R
Ö
M

/
J

F
/

T
T

ich
T
.

:
/
/

/
J

Ö
N

1
8

M
A
j

2
0
2
1

Functional Magnetic Resonance Imaging

Data Acquisition and Preprocessing

Whole-brain images were acquired with a 3-T scanner (GE
Signa VH/I, Waukesha, WI). High-resolution T1-weighted
Bilder (30 axial slices, 0.9375 (cid:1) 0.9375 (cid:1) 4 mm) were ob-
tained in all participants, and a gradient-echo, echoplanar
sequence (TR = 2000 ms, TE = 50 ms) was used to
acquire data sensitive to the blood oxygen level depen-
dent (BOLD) signal (Kwong et al., 1992; Ogawa et al.,
1992) within a 64 (cid:1) 64 Matrix (30 axial slices coplanar
with the T1 acquisition, 3.75 (cid:1) 3.75 (cid:1) 4 mm, no skip).
Scans of the delayed-recognition task were preceded by
a scan in which we derived an estimate of the hemody-
namic response function (HRF) für jeden Teilnehmer.
During this scan, each participant performed a simple
reaction-time task that required a bimanual button press
once every 20 sec in response to a brief change in shape
of the fixation stimulus. A partial F test associated with a
Fourier basis covariate set ( Josephs, Turner, & Friston,
1997) was used to evaluate the significance of task-
correlated activity in each voxel of primary somatosensory
and motor cortical regions of interest (ROIs). An HRF
estimate was extracted from the suprathreshold voxels
of these ROIs by spatially averaging their time series, fil-
tering the resultant averaged fMRI time series to remove
hoch (> 0.244 Hz) and low (< 0.05 Hz) frequencies, adjusting it to remove the effects of nuisance covari- ates (Friston, Holmes, Poline, Heather, & Frackowiak, 1995), and trial averaging. The HRF characterizes the fMRI response resulting from a brief impulse of neural activity (Boynton, Engel, Glover, & Heeger, 1996) and can vary markedly across subjects (Handwerker, Ollinger, & D’Esposito, 2004; Aguirre, Zarahn, & D’Esposito, 1998). The subject-specific HRFs were used to convolve inde- pendent variables entered into the modified general linear model (GLM; Worsley & Friston, 1995) that we used to analyze the data from the scans of the working memory task. The eight scans of the working memory task each lasted 6 min 20 sec (6 min of task preceded by 20 sec of dummy pulses to achieve a steady state of tissue magnetization). Postle et al. 1713 Analyses The fMRI time series analysis modeled the signal change associated with each discrete epoch of the trial with a covariate comprised of a BOLD HRF shifted along the timeline of the task in order best model the trial epoch in question (Postle, Zarahn, & D’Esposito, 2000; Zarahn, Aguirre, & D’Esposito, 1997). The least-squares solu- tion of the GLM of the fMRI time series data yielded parameter estimates that were associated with each co- variate of interest. The smoothness of the fMRI response to neural activity allows fMRI evoked responses that arise from temporally dependent events to be resolved on the order of 4 sec (Zarahn et al., 1997). Load-sensitive and alphabetization-sensitive voxels were identified with the contrasts [DelayForward 5–DelayForward 2] and [DelayAlphabetize 5–DelayForward 5], respectively, thresh- olded at a mapwise level of p = .05, Bonferroni-corrected for multiple comparisons. Because the principal function of the fMRI data for this study was to provide activation maps that would guide the rTMS, the principal analyses were performed as single-subject analyses. (This first step was a precise replication of the procedure from Postle et al., 1999.) Transforming a subject’s anatomical and functional data into a ‘‘normalized’’ atlas space would not be appropri- ate with this approach for the simple reason that rTMS can only be applied to a subject’s brain in its ‘‘native’’ configuration (i.e., it is not possible to apply rTMS to a composite, group-normalized statistical volume; for a similar approach, see the work of Herwig et al., 2003). Before the rTMS session, the whole-brain alphabetiza- tion statistical map was coregistered and merged with a high-resolution T1-weighted anatomical scan. The three- dimensional reconstruction of this merged image would be used to guide rTMS. To provide a sense of aggregate trends in activity produced by our task, we also performed a spatial normalization-based group analysis. This analysis was performed by first warping unthresholded statistical volumes from each subject to a template in MNI space, smoothing them to 8-mm full width half maximum, then evaluating the reliability of these statistical maps across subjects with a ‘‘second-level’’ analysis implemented with a GLM that treated subject as a random variable. The resultant group maps were thresholded at p = .01, uncorrected for multiple comparisons. Transcranial Magnetic Stimulation Procedure The behavioral task used in the rTMS session only in- cluded Alphabetize 5 and Forward 5 trials that occurred with equal probability in a randomly determined or- der. An intertrial interval of 10 sec separated each trial.1 Orthogonal to the factor of instructions was that of rTMS (present, absent; each also occurring randomly with p = .5). An entire rTMS study comprised 12 twelve-trial blocks, with four consecutive blocks performed for each stimulation site: middle frontal gyrus (MFG) of the dlPFC, SPL, and postcentral gyrus (PCG). Order of stimulation site was counterbalanced across subjects. Within each block, the orthogonal factors of instructions (alphabetize, forward) and rTMS (present, absent) were randomized such that each trial type occurred three times during each block. Probe validity also varied independently of the two principal factors of interest, such that an equal number of valid and invalid probes occurred during each block, no more than two trials of any type (e.g., an ‘‘alphabetize, rTMS absent’’ trial) featured a valid probe within a single block, and an equal number of valid and invalid probes (i.e., three) had occurred for each trial type upon com- pletion of each even-numbered block. Apparatus TMS was delivered with a Magstim Standard Rapid mag- netic stimulator fit with a 70-mm figure-8 stimulating coil (Magstim, Whitland, Wales, UK). The first step of the TMS session was to determine the minimal intensity at which a single pulse through the TMS coil, positioned over the motor cortex, reliably produced a motor-evoked potential of (cid:2)50 AV in the abductor pollicis brevis in 5 of 10 successive stimuli. This ‘‘motor threshold’’ was the intensity at which the subsequent rTMS was performed.2 As stated in the Introduction section, the logic of the experiment was to target portions of the dlPFC and SPL that showed alphabetization sensitivity (i.e., DelayAlphabetize 5 > DelayForward 5) in the fMRI scan. Das
was accomplished via coregistration, with infrared-based
frameless stereotaxy (eXimia Navigated Brain Stimulation
[NBS]; Nexstim, Helsinki, Finland), of the subject’s head
with his/her MRI data. The TMS coil was also fitted with
infrared-reflecting beacons, thereby permitting us to tar-
get regions identified in the fMRI data with rTMS. NBS
works from the understanding that TMS preferentially
stimulates neurons located in the area where the induced
current is strongest (Thielscher & Kammer, 2002). Der
system displays the cortical area likely to be maximally
stimulated by TMS by displaying the electric field maxi-
mum in the cortex after calculating the estimated distri-
bution and strength of the intracranial electric field. Das
computation takes into account the exact shape of the
copper wiring inside the TMS coil, the three-dimensional
position and orientation of the coil, and the overall shape
of the head and the brain.

Target Selection

In each subject, only one site was stimulated in each of
three regions: dlPFC, SPL, and a control site in the PCG.
We opted to target the ‘‘hotspots’’ from each individual
subject’s fMRI data, rather than areas defined by compos-
ite, group-averaged statistical maps, because of growing

1714

Zeitschrift für kognitive Neurowissenschaften

Volumen 18, Nummer 10

D
Ö
w
N
l
Ö
A
D
e
D

F
R
Ö
M

/
J

F
/

T
T

ich
T
.

:
/
/

/
J

Ö
N

1
8

M
A
j

2
0
2
1

evidence for high levels of intersubject topographical
variability in many domains of cognition (z.B., Swallow,
Mutiger, Snyder, Speer, & Zacks, 2003; Tsao, Freiwald,
Knutsen, Mandeville, & Tootell, 2003; Miller et al., 2002),
including the STR of information (Feredoes & Postle,
2005), but relatively lower intrasubject variability over time
(d.h., good test–retest reliability, Feredoes & Postle, 2005;
Peeling & Downing, 2005; Tsao et al., 2003; Miller et al.,
2002). (One manifestation of high intersubject topograph-
ic variability in the present data set was that the mean
alphabetization effect size from the dlPFC was an order
of magnitude larger in the single-subject analyses, 2% sig-
nal change, Figur 1, than in the group-averaged analysis,
0.33 % signal change, averaged across the two hemi-
spheres; Tisch 1.) Based on this, it may be that a single-
subject statistical map provides a better estimate of the
true anatomical location of task-related activity for that
subject than would a map derived from a group average.
(If it were the case that we were applying rTMS to sub-
jects for whom we did not have fMRI data, Jedoch, unser
best estimate would come from group-averaged data.)

In some instances, there were multiple foci of alpha-
betization-sensitive activity in a particular region, und in
these cases the focus selected was the one whose stim-
ulation was judged by the experimenters to be most
likely to be tolerated by the subject. Insbesondere, target-
ing regions of the MFG located relatively ventrally and
anteriorly increases the likelihood of stimulating the su-
perior auricularis muscle, which can produce involuntary
wincing and discomfort. In such instances, if a different
focus of alphabetization sensitivity were located in a more

dorsal and/or posterior portion of the MFG, we would
target this latter focus. Insbesondere, we found that stim-
ulation sites near or overlapping the inferior bank of
the superior frontal sulcus were well tolerated by sub-
jects.3 Another constraint was that, for each individual
Thema, all three stimulation sites were in the same hemi-
Kugel. This meant that, for each subject, the hemi-
sphere to be stimulated was determined by the location
of alphabetization-sensitive activity in the dlPFC and SPL.
PCG was chosen as the control region for this study be-
cause it was presumed to have no direct involvement in
any component of the working memory task, and thus
could serve as a region to control for nonspecific effects
of cortical rTMS. daher, PCG stimulation sites were
expressly selected for the absence of either alphabetiza-
tion- or load-sensitive activity. At each stimulation site,
the stimulating coil was oriented with the handle point-
ing posteriorly with respect to the subject’s head, Und
roughly parallel to the midline, so as to induce current in
the brain in the posterior-to-anterior direction.

Stimulation Parameters

On stimulation-present trials, the system was programmed
to deliver 30 equally spaced pulses during a 6-sec epoch,
beginning 2 sec after the offset of the instructions and
lasting for the remaining 6 sec of the delay period (d.h., A
6-sec-long train of 5 Hz rTMS). Jedoch, a programming
error was detected after the experiment was completed,
and inspection of the stimulation logs indicated that for
three subjects, Die 30 pulses were actually delivered over

Figur 1. Group load and
alphabetization effects from
the fMRI data. Effects were
extracted in a two step
Verfahren. Erste, load-sensitive
and alphabetization-sensitive
voxels were identified in each
of the three types of region.
Zweite, the magnitude of load
and alphabetization sensitivity
was determined for the voxels
identified in Step 1. Der
topography of load sensitivity
varied across subjects
(see text), and the effects
displayed here were drawn
from the functionally defined
load sensitive region of
each subject. Because load
sensitivity was only observed
in two subjects in the dlPFC
(Subjects 16 Und 21) Und
in two subjects in the SPL
(Subjects 21 Und 27), diese
results are plotted as single
data points, rather than as
means. Error bars represent
95% confidence intervals.

D
Ö
w
N
l
Ö
A
D
e
D

F
R
Ö
M

/
J

F
/

T
T

ich
T
.

:
/
/

/
J

Ö
N

1
8

M
A
j

2
0
2
1

Postle et al.

1715

Tisch 1. Tabulation of Activity Identified in Spatial Normalization-based Group Analyses

Region of Activation

MNI Coordinates (mm)

Effect Size (Mean Percent
Signal Change)

Volume of Activation
(mm3)

Alphabetization contrast

Right PFC

Left PFC

Anterior cingulate cortex

Superior frontal gyrus

Right SPL

Right intraparietal sulcus

Left SPL

Left inferior parietal lobule

31.88

26.3

(cid:3)2

7.7

3.8

33.8

(cid:3)15

(cid:3)7.5

(cid:3)37.5

Right head of the caudate nucleus

11.25

Left head of the caudate nucleus

(cid:3)11.25

Load contrast

Left inferior frontal gyrus/MFG

Right central sulcus

Right PCG

Right SPL

Right inferior frontal gyrus

Right fusiform gyrus

Right amygdala

Right medial temporal lobe

(cid:3)45

18.75

26.25

67.5

48.75

22.5

18.75

22.5

28.1

10.5

1.9

(cid:3)63.8

(cid:3)60

(cid:3)56.25

(cid:3)60

(cid:3)45

11.25

7.5

(cid:3)7.5

(cid:3)41.25

(cid:3)33.75

(cid:3)71.25

7.5

(cid:3)45

(cid:3)3.75

(cid:3)18.75

(cid:3)24

(cid:3)20

(cid:3)16

0.37

0.29

0.34

0.46

0.63

0.42

0.39

0.38

0.47

0.45

0.13

0.14

0.21

0.23

0.73

0.42

0.41

0.43

3825.0

2137.5

562.5

956.3

900.0

506.3

281.3

900.0

1012.5

5.6

56.3

The coordinates reported here indicate the centers of clusters of activity identified within each anatomical region. Identification of anatomical
regions was confirmed via conversion of MNI coordinates to Talairach coordinates with the mni2tal Matlab routine of Matthew Brett (http://
www.mrc-cbu.cam.ac.uk/Imaging/mnispace.html) and looking them up in the atlas of Talairach and Tournoux (1988).

a 6.3-sec epoch (daher, at a rate of 4.8 Hz), und für
the remaining nine subjects, Die 30 pulses were deliv-
ered over a 6.8-sec epoch (daher, at a rate of 4.3 Hz).
The result was that, for each subject, rTMS was delivered
during the final 6 sec of the delay period and continued
during the first few hundred msec after the onset of the
probe. This ‘‘spillover’’ of rTMS into the probe portion
of the trial, although unfortunate, does not complicate
our interpretation of our results, for reasons summa-
rized in the Results section.

ERGEBNISSE

Functional Magnetic Resonance Imaging

Single-subject Analyses

Alphabetization-sensitive voxels were identified in the
dlPFC in 11 subjects and in the SPL in all 12 Fächer.

Load-sensitive voxels were identified in 11 Fächer, im
following regions (with n corresponding to the number
of subjects in which it was detected):
left superior
temporal gyrus and/or inferior parietal lobule, n = 5;
right inferior parietal lobule, n = 1; SPL, n = 2; left dorsal
extrastriate cortex, n = 1; left central sulcus, n = 2; Rechts
central sulcus, n = 1; anterior cingulate gyrus, n = 1; links
dlPFC, n = 2; right temporal pole, n = 1. To accommo-
date this topographical variability, the ‘‘load sensitivity
region’’ was defined exclusively from functional proper-
ties and permitted to vary topographically across sub-
Projekte. To address the question of whether alphabetization
sensitivity and load sensitivity are seen in the same vox-
els, we extracted estimates of these two effects (sehen
Methods section, Functional Magnetic Resonance Imag-
ing, Analyses) from alphabetization-sensitive voxels found
in the three ROIs (load-sensitive, dlPFC, and SPL) sowie
as from load-sensitive voxels found in these ROIs (z.B.,

1716

Zeitschrift für kognitive Neurowissenschaften

Volumen 18, Nummer 10

D
Ö
w
N
l
Ö
A
D
e
D

F
R
Ö
M

/
J

F
/

T
T

ich
T
.

:
/
/

/
J

Ö
N

1
8

M
A
j

2
0
2
1

from alphabetization-sensitive voxels in the dlPFC, Wir
extracted an estimate of the alphabetization sensitivity
and of the load sensitivity.) Inspection of mean effect
sizes and their associated 95% confidence intervals indi-
cated that in none of the ROIs did alphabetization-
sensitive voxels display significant load sensitivity or, In
any ROI, did load-sensitive voxels demonstrate alphabet-
ization sensitivity (Figur 1). In der Tat, in the dlPFC and
SPL, alphabetization-sensitive voxels showed reliably neg-
ative load effects (mean and 95% confidence interval < 0). These results replicate and extend the findings of Postle et al. (1999) and stand in contrast to the conclu- sions of Veltman et al. (2003). Group Analyses Alphabetization-sensitive activity identified by the spatial normalization-based group analyses was extensive and spanned large extents, bilaterally, of the PFC, anterior cingulate cortex, superior frontal cortex, SPL, and the caudate nucleus (Table 1). In both hemispheres of the PFC, these voxels showed a significant negative effect of load [right: t(11) = (cid:3)3.4, p < .01; left: t(11) = (cid:3)4.4, p < .005]. Load-sensitive activity identified by this analy- sis was markedly more sparse, consisting only of indi- vidual voxels in the right and left frontal cortex, right central sulcus and parietal cortex, and three locations in the right temporal lobe (Table 1). In neither hemisphere of the PFC did these load-sensitive voxels show a sig- nificant alphabetization effect [right: t(11) = (cid:3)0.5, ns; left: t(11) = 1.5, ns]. Therefore, the results of the group analyses also demonstrated a dissociation of retention from manipulation effects (as with the single-subject analyses, consistent with Postle et al., 1999, and incon- sistent with Veltman et al., 2003). Repetitive Transcranial Magnetic Stimulation rTMS was performed in the left hemisphere in seven subjects and in the right hemisphere in five (Figure 2). Analyses of variance (ANOVAs) indicated that rTMS had its greatest effects on Alphabetize 5 performance at the dlPFC and SPL sites and on Forward 5 performance at the SPL (Figure 3). An initial omnibus ANOVA found no effect of hemisphere of stimulation, F(1,10) = 0.1, ns, and so all subsequent analyses collapsed across this varia- ble. Omnibus ANOVA of the accuracy data revealed main effects of trial type, F(1,11) = 11.2, p < .005, and rTMS, F(1,11) = 7.6, p < .05, but not of region, F(2,22) < 1.0, ns, interactions Region (cid:1) rTMS, F(2,22) = 4.9, p < .05, and Trial Type (cid:1) rTMS, F(1,11) = 5.42, p < .05 (no other interactions achieved significance, Fs < 1.0). ANOVA of the accuracy data from dlPFC alone confirmed a selective effect of rTMS on alphabetization performance, with sig- nificant main effects of trial type, F(1,11) = 10.4, p < .01, and rTMS, F(1,11) = 9.5, p < .05, and a Trial type (cid:1) rTMS interaction, F(1,11) = 5.1, p < .05. In contrast, ANOVA of the accuracy data from the PCG control region did not show evidence for comparable selectivity of rTMS. In- stead, it revealed a main effect of trial type, F(1,11) = 5.7, p < .05, but no main effect of rTMS or Trial type (cid:1) rTMS interaction (Fs < 3.5). To confirm the differential effects of rTMS on these two regions, ANOVA directly comparing the dlPFC with PCG was performed. It revealed a main effect of trial type, F(1,11) = 10.6, p < .001, Region (cid:1) rTMS interaction, F(1,11) = 7.1, p < .05, and Trial Type (cid:1) rTMS interaction, F(1,11) = 9.3, p < .05. In contrast with the results from the dlPFC, ANOVA of the accuracy data from SPL stimulation revealed main effects of trial type, F(1,11) = 5.1, p < .05, and rTMS, F(1,11) = 5.1, p < .05, but no Trial Type (cid:1) rTMS interaction, F(1,11) < 1, ns. The absence of this inter- action indicated that rTMS effects on Alphabetize 5 and Forward 5 performance were comparable. Analyses of RT data (Figure 3) confirmed that there were no effects of interest in these data (including no effect of hemisphere of stimulation; Fs (cid:4) 2.0, ns). Note that the mean RT, collapsed across all trial types and regions, was roughly double that of the 800-msec intru- sion of rTMS, experienced by nine of the subjects, into the probe epoch of the trial. This, and the fact that in no region was RT sensitive to rTMS, gives us confidence that the error effects reported above were because of the disruption of processes engaged during the delay period. DISCUSSION The results from the dlPFC are clear: Alphabetization- related fMRI activity is independent of (indeed, perhaps negatively correlated with) load-related activity, and delay-period rTMS of loci of alphabetization-related activity produces a decrement on Alphabetize 5 trials, but not on Forward 5 trials. This portion of the results therefore confirms our hypothesis that disruption of manipulation-sensitive regions of the dlPFC would disrupt manipulation-dependent, but not retention- dependent, performance. (Note that these dlPFC results cannot be attributed to the disparity in difficulty be- tween the two tasks because of the results from the SPL.) Our preferred interpretation of this aspect of our results is that they were produced by rTMS disruption of delay-period manipulation processes. We cannot rule out, however, the possibility that the spillover of rTMS into the first 300–800 msec of the probe epoch may have interfered with probe perception and/or evaluation- related processes. Nonetheless, this ambiguity does not lessen the result of principle theoretical import, which is that rTMS manipulation-sensitive regions of the dlPFC did not significantly disrupt the STR of information. The results from the SPL are more complex and, indeed, illustrate the value that disruptive techniques Postle et al. 1717 D o w n l o a d e d f r o m l l / / / / / j f / t t i t . : / / D h t o t w p n : o / a / d m e i d t f p r o r m c . h s i p l v d e i r r e c c h t . a m i r e . d c u o m o / c j n o a c r n t i c / a e r - p t d i c 1 l 8 e 1 - 0 p 1 d 7 f 1 / 2 1 8 1 / 9 1 3 5 0 6 / 0 1 3 7 1 o 2 c / n 1 2 7 0 5 0 5 6 9 1 4 8 8 / 1 j 0 o 1 c 7 n 1 . 2 2 p 0 d 0 6 b . y 1 g 8 u . e 1 s 0 t . o 1 n 7 0 1 7 2 S . e p p d f e m b b y e r g 2 u 0 e 2 s 3 t / j t . . . . . f o n 1 8 M a y 2 0 2 1 D o w n l o a d e d f r o m l l / / / / / j f / t t i t . : / / D h t o t w p n : o / a / d m e i d t f p r o r m c . h s i p l v d e i r r e c c h t . a m i r e . d c u o m o / c j n o a c r n t i c / a e r - p t d i c 1 l 8 e 1 - 0 p 1 d 7 f 1 / 2 1 8 1 / 9 1 3 5 0 6 / 0 1 3 7 1 o 2 c / n 1 2 7 0 5 0 5 6 9 1 4 8 8 / 1 j 0 o 1 c 7 n 1 . 2 2 p 0 d 0 6 b . y 1 g 8 u . e 1 s 0 t . o 1 n 7 0 1 7 2 S . e p p d f e m b b y e r g 2 u 0 e 2 s 3 t / j t . f . . . . o n 1 8 M a y 2 0 2 1 Figure 2. (A) Illustration of an rTMS experiment, from Subject 14 as displayed by the NBS system. The brain is displayed as though looking down from above, with the nose at the top of the image. The right hemisphere appears on the right side of the image. Skin and bone have been removed to below the level of the Sylvian fissure, and the cortex has been ‘‘peeled’’ to the depth that best displays the fMRI information (white blobs indicate regions showing alphabetization sensitivity). Each red sphere indicates the location on the scalp at which an rTMS train was delivered; the corresponding yellow spike indicates the orientation of the induced magnetic field for that stimulation train; the corresponding orange arrow indicates the estimated direction of current induction. Purple spheres indicate the targeted portions of the dlPFC and SPL. For this subject, PCG was the last region stimulated, and the splash of purple color at the end of the red spike indicates the estimated area of maximal intensity of the single rTMS train that is captured in this image. (B–H) Analogous images, highlighting rTMS of the dlPFC, are presented for seven additional subjects. Images for the rTMS sessions of the remaining four subjects were lost during a software upgrade. (B) Illustration of the left dlPFC rTMS of Subject 3. Display conventions are the same as those in (A). (C) Illustration of the right dlPFC rTMS of Subject 8. Display conventions are the same as those in (A). (D) Illustration of the three rTMS targets of Subject 16, including the left dlPFC. Display conventions are the same as those in (A). 1718 Journal of Cognitive Neuroscience Volume 18, Number 10 D o w n l o a d e d f r o m l l / / / / / j f / t t i t . : / / D h t o t w p n : o / a / d m e i d t f p r o r m c . h s i p l v d e i r r e c c h t . a m i r e . d c u o m o / c j n o a c r n t i c / a e r - p t d i c 1 l 8 e 1 - 0 p 1 d 7 f 1 / 2 1 8 1 / 9 1 3 5 0 6 / 0 1 3 7 1 o 2 c / n 1 2 7 0 5 0 5 6 9 1 4 8 8 / 1 j 0 o 1 c 7 n 1 . 2 2 p 0 d 0 6 b . y 1 g 8 u . e 1 s 0 t . o 1 n 7 0 1 7 2 S . e p p d f e m b b y e r g 2 u 0 e 2 s 3 t / j t . f . . . . o n 1 8 M a y 2 0 2 1 Figure 2. (continued ) (E) Illustration of frontal-lobe rTMS target of Subject 20. This is the subject mentioned in footnote 3, who could not tolerate MFG rTMS. Display conventions are the same as those in (A). (F ) Illustration of the left dlPFC rTMS of Subject 25. Display conventions differ from those in (A) in that an orange sphere represents the targeted portion of cortex and information relating to the NBS system’s estimates of stimulation parameters is not shown. (G) Illustration of the left dlPFC rTMS of Subject 27. Display conventions are the same as those in (F), except that the line emanating from the targeted portion of cortex illustrates the maximal energy vector of the rTMS-induced magnetic field. (H) Illustration of the left dlPFC rTMS of Subject 17. Display conventions are the same as those in (G). Postle et al. 1719 D o w n l o a d e d f r o m l l / / / / / j f / t t i t . : / / D h t o t w p n : o / a / d m e i d t f p r o r m c . h s i p l v d e i r r e c c h t . a m i r e . d c u o m o / c j n o a c r n t i c / a e r - p t d i c 1 l 8 e 1 - 0 p 1 d 7 f 1 / 2 1 8 1 / 9 1 3 5 0 6 / 0 1 3 7 1 o 2 c / n 1 2 7 0 5 0 5 6 9 1 4 8 8 / 1 j 0 o 1 c 7 n 1 . 2 2 p 0 d 0 6 b . y 1 g 8 u . e 1 s 0 t . o 1 n 7 0 1 7 2 S . e p p d f e m b b y e r g 2 u 0 e 2 s 3 t / j t . . . . . f o n 1 8 M a y 2 0 2 1 Figure 3. Results from rTMS session. Error bars indicate SEM. can bring to systems-level analyses of cognitive func- tions. As with the dlPFC, fMRI data from the SPL indicated that alphabetization-related activity was inde- pendent of retention-related activity in the SPL. The rTMS data, however, led to the opposite conclusion: Disruption of activity at alphabetization-sensitive sites in the SPL yielded comparable levels of impairment on retention-requiring trials as on manipulation-requiring trials. There are at least two possible explanations for this divergence of the rTMS from the fMRI data. One is that the fMRI techniques that we employed are not sufficiently sensitive to detect retention-related effects in the SPL that are nonetheless sensitive to rTMS, a possibility that would suggest that manipulation- and retention-related functions are not independent in the SPL. This is difficult to reconcile, however, with the fact that the SPL load effects in alphabetization-sensitive voxels were not simply ‘‘not different from 0,’’ but were actually significantly less than 0. A second explanation, is that the SPL voxels identified by which we favor, fMRI as alphabetization-specific are just that, but that the SPL sites that we stimulated were more proximal to retention-related areas than were the analogous dlPFC sites. The proximity invoked here might be either topographical or synaptic. By the former account, the spread of the rTMS-induced electric field might be such that rTMS targeting the SPL also has disruptive effects on topographically adjacent areas, such as the angular and/or supramarginal gyri of the inferior parietal lobule, that may, themselves, support the retention of phonologically encoded information in working mem- ory. By the latter account, the region(s) supporting retention may be only one or two synapses distant from the SPL. It is known, for example, that TMS produces PET activation in regions that are mono- and disynapti- cally connected to the region that is directly targeted with TMS (Ferrarelli et al., 2004). The absence of disruptive effects of the PFC rTMS on Forward 5 performance is consistent with the previ- ously existing literature. Most notably, Herwig et al. (2003) used a delayed letter recognition task quite sim- ilar to our Forward 5 trials, and a similar method for guiding rTMS with fMRI data. They found no effect with a memory load of six items of delay-period rTMS target- ing the dlPFC or parietal cortex (SPL or inferior parietal lobule, depending on the subject). With a load of seven items, however, delay-period rTMS targeting both the lat- eral premotor and parietal cortex and had disruptive ef- fects. (They did not stimulate the dlPFC at load 7.) Other studies that have disrupted working memory perform- ance with rTMS of the PFC have used tasks that leave ambiguous the type of information and/or process that is being disrupted. For example, delay-period activity during spatial delayed response (Brandt, Ploner, Meyer, Leistner, & Villringer, 1998; Pascual-Leone & Hallett, 1994) might be sensory, attentional, motoric, or some combination of these. The n-back task (Mottaghy, Gangitano, Krause, & Pascual-Leone, 2003; Mottaghy, Doring, Muller-Gartner, Topper, & Krause, 2002; Mottaghy et al., 2000) requires the simultaneous engagement of several retention- and control-related operations. The absence of laterality effect in our rTMS results is also consistent with the existing literature, which documents disruption of verbal working memory performance with right hemisphere as well as left hemisphere rTMS (e.g., Herwig et al., 2003; Mottaghy et al., 2002). This may be because of the complex and 1720 Journal of Cognitive Neuroscience Volume 18, Number 10 distributed nature of the networks that support verbal working memory (e.g., Mottaghy et al., 2002, 2003) and/ or that rTMS of a right hemisphere target is expected to have strong effects on the homologous contralateral re- gion (Ferrarelli et al., 2004; Bestmann, Baudewig, Siebner, Rothwell, & Frahm, 2003). The results of the present study are consistent with a model of segregation of executive control from STR functions in the PFC. They are also consistent with the idea that posterior regions may be more important for the STR of information than is the PFC. The contribu- tions of the PFC to working memory may be via general- purpose control processes that are not specialized for working memory (Postle, in press). Acknowledgments We thank Steve Esser, Olufunsho Faseyitan, Cassandra Jones, and Christopher Jordan for technical and experimental assist- ance. This work was supported by NIH MH064498 (B. R. P.) and NARSAD (G. T.). Reprints requests should be sent to Bradley R. Postle, 1202 West Johnson Street, Madison, WI 53726, USA, or via e-mail: postle@wisc.edu. Notes 1. The logic of this portion of the study was that Forward 2 trials had only been included in the fMRI study to permit eval- uation of whether alphabetization-sensitive voxels (i.e., voxels significant for the [DelayAlphabetize 5–DelayForward 5] contrast) also showed load sensitivity (as assessed with the contrast [DelayForward 5–DelayForward 2]). For rTMS, however, Forward 5 trials were sufficient, because if Forward 5 performance were disrupted by delay-period rTMS, one could infer that processes necessary for the STR of 5 items were disrupted. (Note that this inference would hold whether or not Forward 2 performance was disrupted by rTMS, a fact that makes clear why Forward 2 trials were not needed for the rTMS portion of this study.) The converse would be true if Forward 5 performance were not disrupted by delay-period rTMS. 2. The motor threshold offers a means of normalizing stim- ulation intensity across subjects, because stimulation of the (contralateral) motor cortex and PFC at motor threshold pro- duces positively correlated evoked responses (Kahkonen et al., 2004). 3. One subject (Subject 20) could not tolerate rTMS of the MFG, because of excessive stimulation of the superior auricularis muscle. For this subject, a cluster of alphabetization-sensitive voxels along the midline, in a region judged to correspond to Brodmann’s area 8, was selected as the alternative target for dlPFC stimulation (Figure 2E). The remaining two regions for this subject were stimulated in the right hemisphere. REFERENCES Aguirre, G. K., Zarahn, E., & D’Esposito, M. (1998). The variability of human, BOLD hemodynamic responses. Neuroimage, 8, 360–369. Baddeley, A. D. (2000). The episodic buffer: A new component of working memory? Trends in Cognitive Sciences, 4, 417–423. Baddeley, A. D., & Logie, R. H. (1999). Working memory: The multiple-component model. In A. Miyake & P. Shah (Eds.), Models of working memory (pp. 28–61). Cambridge, UK: Cambridge University Press. Bestmann, S., Baudewig, J., Siebner, H. R., Rothwell, J. C., & Frahm, J. (2003). Subthreshold high-frequency TMS of human primary motor cortex modulates interconnected frontal motor areas as detected by interleaved fMRI-TMS. Neuroimage, 20, 1685–1696. Boynton, G. M., Engel, S. A., Glover, G. H., & Heeger, D. J. (1996). Linear systems analysis of functional magnetic resonance imaging in human V1. Journal of Neuroscience, 16, 4207–4221. Brandt, S. A., Ploner, C. J., Meyer, B.-U., Leistner, S., & Villringer, A. (1998). Effects of repetitive transcranial magnetic stimulation over dorsolateral prefrontal and posterior parietal cortex on memory-guided saccades. Experimental Brain Research, 118, 197–204. Constantinides, C., Franowicz, M. N., & Goldman-Rakic, P. S. (2001). The sensory nature of mnemonic representation in the primate prefrontal cortex. Nature Neuroscience, 4, 311–316. Courtney, S. M. (2004). Attention and cognitive control as emergent properties of information representation in working memory. Cognitive, Affective, & Behavioral Neuroscience, 4, 501–516. D’Esposito, M., & Postle, B. R. (1999). The dependence of span and delayed-response performance on prefrontal cortex. Neuropsychologia, 37, 1303–1315. Feredoes, E., & Postle, B. R. (2005). Does size matter? Quantitatively and qualitatively discrepant results yielded by single-subject and spatial normalization approaches to fMRI group analysis. Poster presented at the annual meeting of the Society for Neuroscience 2005. Ferrarelli, F., Haraldsson, H. M., Barnhart, T. E., Roberts, A. D. F., Oakes, T. R., Massimini, M., Stone, C. K., Kalin, N. H., & Tononi, G. (2004). A [17F]-fluoromethane PET/TMS study of effective connectivity. Brain Research Bulletin, 64, 103–113. Friston, K. J., Holmes, A. P., Poline, J.-B., Heather, J. D., & Frackowiak, R. S. J. (1995). Analysis of fMRI time-series revisited. Neuroimage, 2, 45–53. Funahashi, S., Chafee, M. V., & Goldman-Rakic, P. S. (1993). Prefrontal neuronal activity in rhesus monkeys performing a delayed anti-saccade task. Nature, 365, 753–756. Garavan, H., Ross, T. J., Li, S.-J., & Stein, E. A. (2000). A parametric manipulation of central executive functioning. Cerebral Cortex, 10, 585–592. Goldman-Rakic, P. S., & Leung, H.-C. (2002). Functional architecture of the dorsolateral prefrontal cortex in monkeys and humans. In D. T. Stuss & R. T. Knight (Eds.), Principles of frontal lobe function (pp. 85–95). Oxford: Oxford University Press. Hamilton, M. (1960). A rating scale for depression. Journal of Neurology, Neurosurgery, and Psychiatry, 23, 56–62. Handwerker, D. A., Ollinger, J. M., & D’Esposito, M. (2004). Variation of BOLD hemodynamic responses across subjects and brain regions and their effects on statistical analyses. Neuroimage, 21, 1639–1651. Herwig, U., Abler, B., Schonfeldt-Lecuona, C., Wunderlich, A., Grothe, J., Spitzer, M., & Walter, H. (2003). Verbal storage in a premotor-parietal network: Evidence from fMRI-guided magnetic stimulation. Neuroimage, 20, 1032–1041. Jonides, J., Lacey, S. C., & Nee, D. E. (2005). Processes of working memory in mind and brain. Current Directions in Psychological Science, 14, 2–5. Josephs, O., Turner, R., & Friston, K. (1997). Event-related fMRI. Human Brain Mapping, 5, 243–248. Postle et al. 1721 D o w n l o a d e d f r o m l l / / / / / j f / t t i t . : / / D h t o t w p n : o / a / d m e i d t f p r o r m c . h s i p l v d e i r r e c c h t . a m i r e . d c u o m o / c j n o a c r n t i c / a e r - p t d i c 1 l 8 e 1 - 0 p 1 d 7 f 1 / 2 1 8 1 / 9 1 3 5 0 6 / 0 1 3 7 1 o 2 c / n 1 2 7 0 5 0 5 6 9 1 4 8 8 / 1 j 0 o 1 c 7 n 1 . 2 2 p 0 d 0 6 b . y 1 g 8 u . e 1 s 0 t . o 1 n 7 0 1 7 2 S . e p p d f e m b b y e r g 2 u 0 e 2 s 3 t / j t . . . . . f o n 1 8 M a y 2 0 2 1 Kahkonen, S., Wilenius, J., Komssi, S., & Elmoniemi, R. J. Pessoa, L., Gutierrez, E., Bandettini, P. A., & Ungerleider, (2004). Distinct differences in cortical reactivity of motor and prefrontal cortices to magnetic stimulation. Clinical Neurophysiology, 115, 583–588. Kwong, K. K., Belliveau, J. W., Chesler, D. A., Goldberg, I. E., Weisskoff, R. M., Poncelet, B. P., Kennedy, D. N., Hoppel, B. E., Cohen, M. S., Turner, R., Cheng, H. M., Bradly, T. J., & Rosen, B. R. (1992). Dynamic magnetic resonance imaging of human brain activity during primary sensory stimulation. Proceedings of the National Academy of Sciences, U.S.A., 89, 5675–5679. Lebedev, M. A., Messinger, A., Kralik, J. D., & Wise, S. P. (2004). Representation of attended versus remembered locations in prefrontal cortex. PloS Biology, 2, 1919–1935. Leung, H.-C., Seelig, D., & Gore, J. C. (2004). The effect of memory load on cortical activity in the spatial working memory circuit. Cognitive, Affective, & Behavioral Neuroscience, 4, 553–563. Miller, M. B., Van Horn, J. D., Wolford, G. L., Handy, T. C., Valsangkar-Smyth, M., Inati, S., Grafton, S., & Gazzaniga, M. S. (2002). Extensive individual differences in brain activations associated with episodic retrieval are reliable over time. Journal of Cognitive Neuroscience, 14, 1200–1214. Mottaghy, F. M., Doring, T., Muller-Gartner, H. W., Topper, R., & Krause, B. J. (2002). Bilateral parietofrontal network for verbal working memory: An interference approach using repetitive transcranial magnetic stimulation. European Journal of Neuroscience, 16, 1627–1632. Mottaghy, F. M., Gangitano, M., Krause, B. J., & Pascual-Leone, A. (2003). Chronometry of parietal and prefrontal activations in verbal working memory revealed by transcranial magnetic stimulation. Neuroimage, 18, 565–575. Mottaghy, F. M., Krause, B. J., Kemna, L. J., Topper, R., Tellmann, L., Beu, M., Pascual-Leone, A., & Muller-Gartner, H. W. (2000). Modulation of the neuronal circuitry subserving working memory in healthy human subjects by repetitive transcranial magnetic stimulation. Neuroscience Letters, 280, 167–170. Narayanan, N., Prabhakaran, V., Bunge, S. A., Christoff, K., Fine, E. M., & Gabrieli, J. D. (2005). The role of prefrontal cortex in the maintenance of verbal working memory information: An event-related fMRI analysis. Neuropsychology, 19, 223–232. Ogawa, S., Tank, D. W., Menon, R., Ellermann, J. M., Kim, S. G., Merkle, H., & Ugurbil, K. (1992). Intrinsic signal changes accompanying sensory stimulation: Functional brain mapping using MRI. Proceedings of the National Academy of Sciences, U.S.A., 89, 5951–5955. Pascual-Leone, A., & Hallett, M. (1994). Induction of errors in a delayed response task by repetitive transcranial magnetic stimulation of the dorsolateral prefrontal cortex. NeuroReport, 5, 2517–2520. L. G. (2002). Neural correlates of visual working memory: fMRI amplitude predicts task performance. Neuron, 35, 975–987. Petrides, M. (2000). Dissociable roles of mid-dorsolateral prefrontal and anterior inferotemporal cortex in visual working memory. Journal of Neuroscience, 20, 7496–7503. Postle, B. R. (2006). Working memory as an emergent property of the mind and brain. Neuroscience, 139, 23–38. Postle, B. R., Berger, J. S., & D’Esposito, M. (1999). Functional neuroanatomical double dissociation of mnemonic and executive control processes contributing to working memory performance. Proceedings of the National Academy of Sciences, U.S.A., 96, 12959–12964. Postle, B. R., Zarahn, E., & D’Esposito, M. (2000). Using event-related fMRI to assess delay-period activity during performance of spatial and nonspatial working memory tasks. Brain Research Protocols, 5, 57–66. Rypma, B., & D’Esposito, M. (1999). The roles of prefrontal brain regions in components of working memory: Effects of memory load and individual differences. Proceedings of the National Academy of Sciences, U.S.A., 96, 6558–6563. Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., Hergueta, T., Baker, R., & Dunbar, G. C. (1998). The Mini-International Neuropsychiatric Interview (MINI): The development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. Journal of Clinical Psychiatry, 59, 22–33. Swallow, K. M., Braver, T. S., Snyder, A. Z., Speer, N. K., & Zacks, J. M. (2003). Reliability of functional localization using fMRI. Neuroimage, 20, 1561–1577. Talairach, J., & Tournoux, P. (1988). Co-planar stereotaxic atlas of the human brain. New York: Thieme. Theeuwes, J., Olivers, C. N. L., & Chizk, C. L. (2005). Remembering a location makes the eyes curve away. Psychological Science, 16, 196–199. Thielscher, A., & Kammer, T. (2002). Linking physics with physiology in TMS: A sphere field model to determine the cortical stimulation site in TMS. Neuroimage, 17, 1117–1130. Tsao, D. Y., Freiwald, W. A., Knutsen, T. A., Mandeville, J. B., & Tootell, R. B. H. (2003). Faces and objects in macaque cerebral cortex. Nature Neuroscience, 6, 989–995. Veltman, D. J., Rombouts, S. A., & Dolan, R. J. (2003). Maintenance versus manipulation in verbal working memory revisited: An fMRI study. Neuroimage, 18, 247–256. Walsh, V., & Pascual-Leone, A. (2003). Transcranial magnetic stimulation: A neurochronometrics of mind. Cambridge: MIT Press. Worsley, K. J., & Friston, K. J. (1995). Analysis of fMRI time-series revisited—again. Neuroimage, 2, 173–182. D o w n l o a d e d f r o m l l / / / / / j f / t t i t . : / / D h t o t w p n : o / a / d m e i d t f p r o r m c . h s i p l v d e i r r e c c h t . a m i r e . d c u o m o / c j n o a c r n t i c / a e r - p t d i c 1 l 8 e 1 - 0 p 1 d 7 f 1 / 2 1 8 1 / 9 1 3 5 0 6 / 0 1 3 7 1 o 2 c / n 1 2 7 0 5 0 5 6 9 1 4 8 8 / 1 j 0 o 1 c 7 n 1 . 2 2 p 0 d 0 6 b . y 1 g 8 u . e 1 s 0 t . o 1 n 7 0 1 7 2 S . e p p d f e m b b y e r g 2 u 0 e 2 s 3 t / j t . . f . . . Pasternak, T., & Greenlee, M. W. (2005). Working memory Zarahn, E., Aguirre, G. K., & D’Esposito, M. (1997). A trial-based in primate sensory systems. Nature Reviews Neuroscience, 6, 97–107. Peelen, M. V., & Downing, P. E. (2005). Within-subject reproducibility of category-specific visual activation with functional MRI. Human Brain Mapping, 25, 402–408. experimental design for fMRI. Neuroimage, 6, 122–138. Zarahn, E., Rakitin, B., Abela, D., Flynn, J., & Stern, Y. (2005). Positive evidence against human hippocampal involvement in working memory maintenance of familiar stimuli. Cerebral Cortex, 15, 303–316. o n 1 8 M a y 2 0 2 1 1722 Journal of Cognitive Neuroscience Volume 18, Number 10 Repetitive Transcranial Magnetic Stimulation image

PDF Herunterladen