the use of transcranial magnetic stimulation in cognitive ... · tms has nowadays become a standard...

R

Ts

Ma

Bb

Rc

d

e

a

ARRA

KTrTSfESN

C

N

0d

Neuroscience and Biobehavioral Reviews 35 (2011) 516–536

Contents lists available at ScienceDirect

Neuroscience and Biobehavioral Reviews

journa l homepage: www.e lsev ier .com/ locate /neubiorev

eview

he use of transcranial magnetic stimulation in cognitive neuroscience: A newynthesis of methodological issues

arco Sandrinia,b,∗, Carlo Umiltàc, Elena Rusconid,e

Human Cortical Physiology and Stroke Neurorehabilitation Section, National Institute of Neurological Disorders and Stroke, National Institutes of Health, 10 Center Drive,ethesda, MD 20892, USACenter for Neuroscience and Regenerative Medicine, Uniformed Services University of Health Sciences, Henry M. Jackson Foundation, 12725 Twinbrook Parkway,ockville, MD 20852, USADepartment of General Psychology, University of Padova, Via Venezia 8, 35131 Padova, ItalyFunctional Neuroimaging Laboratories, Center for Mind/Brain Sciences (CIMeC), University of Trento, Via delle Regole 101, 38060 Mattarello (TN), ItalyDepartment of Security and Crime Science, University College London, Torrington Place 2-16, London WC1E 7HN, UK

r t i c l e i n f o

rticle history:eceived 10 February 2010eceived in revised form 15 June 2010ccepted 17 June 2010

eywords:

a b s t r a c t

Transcranial magnetic stimulation (TMS) has become a mainstay of cognitive neuroscience, thus facingnew challenges due to its widespread application on behaviorally silent areas. In this review we willsummarize the main technical and methodological considerations that are necessary when using TMSin cognitive neuroscience, based on a corpus of studies and technical improvements that has becomeavailable in most recent years. Although TMS has been applied only relatively recently on a large scaleto the study of higher functions, a range of protocols that elucidate how this technique can be used to

MSTMSheta bursttate-dependencyMRI

investigate a variety of issues is already available, such as single pulse, paired pulse, dual-site, repetitiveand theta burst TMS. Finally, we will touch on recent promising approaches that provide powerful newinsights about causal interactions among brain regions (i.e., TMS with other neuroimaging techniques)and will enable researchers to enhance the functional resolution of TMS (i.e., state-dependent TMS). We

EGafetyumber processing

will end by briefly summarizing and discussing the implications of the newest safety guidelines.© 2010 Elsevier Ltd. All rights reserved.

ontents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5172. Neural mechanisms of TMS-induced effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5173. How to choose the most appropriate TMS protocol? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 518

3.1. Stimulation intensity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5183.2. TMS paradigms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5203.3. Stimulation frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521

4. Dependent variables and effect directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5215. Control conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5236. Site localization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5257. Combining TMS with other techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527

7.1. TMS and neuroimaging—off-line and on-line approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5277.2. TMS with concurrent PET or fMRI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5297.3. TMS with concurrent EEG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 529

8. State-dependent TMS paradigms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 530
9. Safety guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

∗ Corresponding author at: Human Cortical Physiology and Stroke Neurorehabilitationational Institutes of Health, 10 Center Drive, Bethesda, MD 20892, USA. Tel.: +1 301 496

E-mail address: [email protected] (M. Sandrini).

149-7634/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.oi:10.1016/j.neubiorev.2010.06.005

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 532. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 532

Section, National Institute of Neurological Disorders and Stroke,9986; fax: +1 301 402 7010.

dx.doi.org/10.1016/j.neubiorev.2010.06.005

http://www.sciencedirect.com/science/journal/01497634

http://www.elsevier.com/locate/neubiorev

mailto:[email protected]

dx.doi.org/10.1016/j.neubiorev.2010.06.005

Biobeh

1

omntdaidcwHeMstbbrTtecinpatciisstocsffl7iiTs1dsl2mhsmeacltfnpaod

M. Sandrini et al. / Neuroscience and

. Introduction

Cognitive neuroscience is concerned with the scientific studyf the neural substrates of mental processes and their behavioralanifestations. Over several years, different neuroimaging tech-

iques have provided correlational maps of cognitive processes inhe adult human brain at different levels of temporal and spatialetail. These imaging techniques have in common that the signalnd information that is gathered on brain function, such as changesn regional cerebral blood flow (rCBF), blood oxygenation levelependent (BOLD) signals or evoked potential changes (electroen-ephalography, EEG, and magnetoencephalography, MEG) covaryith the mental process of interest (Walsh and Cowey, 2000).owever, correlational approaches cannot differentiate betweenpiphenomenally and causally activated neuronal populations.oving beyond a merely correlational description of the relation-

hip between brain and behavior was the fresh approach offered byranscranial magnetic stimulation (TMS). A TMS-induced change inehavior (usually measured in reaction times, RTs, or accuracy) cane used to inform models of causal relations between specific brainegions and individual cognitive functions (Robertson et al., 2003).MS has nowadays become a standard stimulation technique forhe noninvasive investigation of cognitive function (Pascual-Leonet al., 2000), whereby neural tissue is stimulated by using the prin-iples of electromagnetic induction to generate electrical currentsn the brain (Barker et al., 1985). The key features of the tech-ique are that the TMS stimulator delivers a large current in a shorteriod of time and the current flowing in the TMS coil producesmagnetic field that lasts for only about a millisecond. Provided

hat appropriate stimulation parameters are selected, such rapidlyhanging magnetic field easily penetrates the scalp and skull andnduces an electrical field sufficient to stimulate neuronal activ-ty and change the pre-stimulus dynamics of neuronal firing in thetimulated region. Although its precise mechanisms of action aretill far from clear, TMS is thought to activate neuronal axons inhe cortex and subcortical white matter, rather than the cell bodiesf cortical neurons (Ridding and Rothwell, 2007). TMS focality isurrently expressed in square centimeters as a measure of corticalurface and can be optimized by combining two circular coils toorm a figure-of-eight or butterfly coil. A standard 70-mm butter-y coil (56 mm inside turn diameter, 90 mm outside turn diameter,3 mm mean diameter and nine turns of copper wire per winding)

s thought to maximally stimulate about 1–2 cm2 of cortex beneathts central junction. However, the behavioral spatial resolution ofMS effects ranges from 0.5 cm to 1.5 cm apart, depending on thepecific tissue that is being stimulated (e.g., Beckers and Hömberg,992; Brasil-Neto et al., 1992a; O’Shea and Walsh, 2007a). The rapidecline in magnetic field strength with distance depending on coilize and stimulation intensity is a critical issue in cortical stimu-ation and limits its application to areas that are no deeper than–3 cm from the skull surface. For a standard butterfly coil with aaximum field strength of 2 T, the induced field is more or less

emispherical, peaking at about 2.5 cm from the surface of thekull. Considering skull thickness and the presence of interposedembranes, this implies that most brain tissue cannot be directly

xcited by TMS. It might be, however, indirectly excited by prop-gation of activity from the region beneath the coil via synapticonnections. TMS technology is not suitable at present to stimu-ate deep brain structures such as amygdala, hippocampus or thehalamus. It is possible that with future technical improvements,ocality will be measured in units of volume. For example, the use of
onlinear coil materials (H-coils) for stimulation of deeper neuralathways has already been explored (Levkovitz et al., 2007; Roth etl., 2007), although the advantage of this coil with regard to depthf stimulation in comparison to the figure-of-eight coil has not beenemonstrated yet (Fadini et al., 2009).
avioral Reviews 35 (2011) 516–536 517

In this review we touch on some practical considerations thatare relevant for researchers wishing to utilize this technique tostudy cognitive processes and their neural basis. We will also alertthe reader to important technical and methodological considera-tions for the use of TMS in cognitive neuroscience. In doing this,we will mainly refer to evidence from studies on number pro-cessing (see e.g., Rusconi and Umiltà, 2008; Sandrini and Rusconi,2009), our main field of expertise. All these studies will exem-plify nicely how TMS can be used to investigate a variety ofissues in the cognitive neurosciences in general, including loca-tion, timing, lateralization and functional relevance of the neuralcorrelates underlying number processing. Finally, new researchdirections will be pointed out, that might guide the investigationof functional cortical interactions between two connected brainareas, the influence of TMS in remote brain areas effectively con-nected with stimulation site, the dynamics of cortical connectivityand how to enhance the functional resolution of the technique.Since methodological issues and technological advances cannotbe entirely separated from safety and ethical considerations, thecurrent synthesis will take into account the new safety guide-lines (Rossi et al., 2009), which are meant to update the ones(Wassermann, 1998) that regulated TMS application during its firstexciting decade of experimentation in laboratories all over theworld.

2. Neural mechanisms of TMS-induced effects

Despite the widespread usage of TMS in basic research, its exactmechanisms of action and interactions with ongoing neural activityremain unclear.

Data from animal studies using TMS in combination with directelectrophysiological recording (e.g., Allen et al., 2007; de Labra et al.,2007; Moliadze et al., 2003; Pasley et al., 2009), metabolic (Valero-Cabre et al., 2005) or with optical imaging techniques (Allen etal., 2007) have provided insight into TMS effects, for both primarymotor cortex (M1) and other neural structures, including primaryvisual cortex (V1). TMS has been shown to elicit distinct and com-plex episodes of enhanced and suppressed activity at the corticallevel (Moliadze et al., 2003; Allen et al., 2007) also depending onthe state of the stimulated area (Pasley et al., 2009). Allen et al.(2007) investigated effects induced by short TMS pulse trains at var-ious frequencies (1–8 Hz) on neural and hemodynamic responses ofspontaneous and visually evoked activity. For spontaneous activitytheir main result was a TMS-induced immediate increment last-ing up to 60 s. In a task-free context, TMS appears thus to enhanceactivity in the area underneath the coil. Conversely, the same TMSstimulus suppressed for more than 5 min the neural activity evokedby a visual stimulus (Allen et al., 2007).

The reason why this happens is not yet completely clear. Onepossibility is that TMS induces intracortical inhibition. It has beenindeed found that an induced electric field can increase GABA lev-els, which in turn suppresses activity (Mantovani et al., 2006). Analternative hypothesis is that TMS amplifies background activity,instead of reducing signal strength. In other terms, TMS mightintroduce random neural noise (Moliadze et al., 2003). It is also pos-sible that TMS acts by disrupting the temporal relation betweenneurons belonging to a more extended circuit activated by thetask (Pasley et al., 2009). Finally, it is important to point out thatthese effects were also dose-dependent, scaling with stimulationfrequency (1–8 Hz) and duration (1–4 s), as they became more
pronounced when longer TMS trains and higher stimulation fre-quencies were used.
In humans studies, TMS is generally applied with the aim to dis-rupt the neural activity that subserves cognitive processing (i.e., asa ‘virtual lesion’ method). However, the ‘virtual lesion’ hypothesis

5 Biobeh

dosaCdfaitvwTiaomi(tacp

dTaaetmaghsbTaSeithe

3

adme

3

aptnuuoes

18 M. Sandrini et al. / Neuroscience and

oes not provide insight about the precise mechanisms of actionf TMS (Miniussi et al., 2010). In fact, it is unclear whether TMSuppresses neural signals (Harris et al., 2008), adds random neuralctivity to the ongoing processing (Ruzzoli et al., 2010; Walsh andowey, 2000) or is a combination of each, with the exact balanceepending on stimulation intensity but also on the anatomo-unctional characteristics of the stimulated tissue (see Miniussi etl., 2010; Siebner et al., 2009a). For instance, Harris et al. (2008)nvestigated TMS-induced effects on V1 during a visual discrimina-ion task. The subjects were asked to discriminate the orientation ofisual grating while the level of image noise in the visual stimulusas concurrently manipulated. The effect of interaction between

MS and stimulus noise on the visual discrimination threshold wasnterpreted as showing that TMS decreased signal strength withoutffecting neural noise. Ruzzoli et al. (2010) investigated the effectf TMS (brief pulses) applied to V5/MT on the shape of the psycho-etric function in a visual motion-direction-discrimination task

n which the visual stimulus featured two elements: a visual signaldots that moved coherently in one direction) and visual noise (dotshat moved randomly in many directions). The results showed thatTMS train induced a decrement in the slope of the psychometric

urve and therefore seemed to add neural noise to the perceptualrocess.

Although the results of these two studies, which were based onifferent rationales, and employed different paradigms, sites andMS parameters, are not in agreement, they suggest that throughpsychophysical approach it is possible to highlight the functionalctivation state of the target area and link in a specific way theffects of TMS on behavioral performance and its putative effects athe neural level. The interpretation of TMS effects and the establish-

ent of causal relations between activity in the targeted brain areand a given behavioral effect is thus much more complex than sug-ested by the virtual lesion metaphor. It is true that such metaphoras guided most of the published cognitive studies and provides ahortcut to predicting, describing and interpreting TMS effects onehavior, however, at least basic awareness of the main issues inMS physiology is necessary to devise a correct experiment designnd draw reliable conclusions (see e.g., Di Lazzaro et al., 2008;iebner et al., 2009a; Ziemann, 2010). Studies of the molecularffects of noninvasive stimulation in the human brain will becomencreasingly relevant from a cognitive neuroscience perspective ashey will start to draw systematic links between genetic variations,uman cortical plasticity and individual differences in behavioralffects (see e.g., Cheeran et al., 2010).

. How to choose the most appropriate TMS protocol?

The selection of stimulation parameters (intensity, frequencynd duration) and paradigm relies on a difficult and often arbitraryecision. Here below we will provide an essential overview of theost common issues that the cognitive neuroscientist is likely to

ncounter in the process of planning a TMS experiment.

.1. Stimulation intensity

Resting motor threshold (rMT) has been classically defineds the amount of TMS machine output (intensity) necessary toroduce a motor evoked potential (MEP) exceeding a defined peak-o-peak amplitude (usually 50 �V) 50% of the time in a finiteumber of trials (usually 10). Accurate estimation of rMT is of
tmost importance in both research and clinical studies as it is thenit most commonly used for TMS dosing (Wassermann, 2002). Theriginal guidelines for assessment of rMT by the International Fed-ration of Clinical Neurophysiology suggest that the TMS operatortarts with a subthreshold TMS intensity and increases in steps of
avioral Reviews 35 (2011) 516–536

2% or 5% of machine output until a stimulus strength is reached with“approximately” 50% successful MEPs in 10–20 consecutive stimuli(Rossini et al., 1994). In a revised proposal, Rothwell et al. (1999)stated that the procedure should start with a suprathreshold valuefrom which TMS intensity is decreased in steps of 2% or 5% until alevel is reached below which reliable responses disappear (wherethe definition of reliable response is that at least 50% successfulMEPs are observed in 10–20 stimuli). However, neither proceduredescribes in detail how the sub- or suprathreshold strength valueto initialize the procedure should be obtained.

Conventionally, hand muscles are chosen as a reference for thedetermination of the individual MT. Most researchers use the firstdorsal interosseous (FDI) as target muscle (Kansaku et al., 2007;Oliveri et al., 2004b). MT can be measured also in a tense mus-cle (active motor threshold, aMT). In this case it is the minimalintensity of stimulation required to produce an MEP having ampli-tude of about 150–200 �V on more than 5 out of 10 trials while thesubject is maintaining a voluntary contraction of about 10–20% ofmaximum contraction using visual and/or auditory feedback. Stud-ies have demonstrated high concordance between MT estimationsusing electromyogram (EMG) and visual inspection (Pridmore et al.,1998; Stokes et al., 2005). Therefore, in some studies applying TMSover non-motor areas, the MT is established as the lowest stimula-tion intensity that would result in a visible movement in a relaxedmuscle (Rusconi et al., 2005, Exp. 1; Sandrini et al., 2004) or in atense muscle (Göbel et al., 2001, 2006).

TMS protocols expressed as a percentage of MT allow stimula-tor output to be calibrated individually to an overt physiologicalresponse, even when applied to non-motor cortical regions (e.g.,Andres et al., 2005, 130% of rMT; Göbel et al., 2006, 110% of aMT;Sandrini et al., 2004, 110% of rMT; Oliveri et al., 2004b, 90% of rMT;Rusconi et al., in press, 110% of aMT).

Individual differences in MT are closely related to variations inthe cortical depth of M1. It is therefore reasonable to assume that,when the intensity of TMS applied to non-motor regions is basedon MT, variations in depth from the overlying scalp surface willresult in different levels of effective stimulation. Indeed, if targetedcortical sites vary in depth from the stimulating coil, the strengthof the magnetic field necessary to reach them, and thus the amountof cortical stimulation, will also vary. Nowadays, a simple methodfor adjusting MT to account for variations in cortical distance isavailable (Stokes et al., 2005, 2007), providing a more accurate cal-ibration than unadjusted MT for the effective application of TMS incognitive neuroscience. It shows that for every millimeter from thestimulating coil, an additional 2.5–3% of TMS output is required toinduce an equivalent level of brain stimulation to the motor cortex.Such gradient also depends strongly on the motor threshold itself(lower MT = lower gradient).

It is important to mention that MEP threshold varies widelyin the healthy populations. Wassermann (2002) estimated thatexperimental error and other unstable determinants of thresh-old may account for about 36% of the across subjects variabilityat rest and about 50% during voluntary activation. The remainingbetween-subjects variability in MEP threshold presumably resultsfrom relatively stable biological differences between individuals.

Most TMS studies did not find any differences in excitabilitythreshold of the primary motor cortex (M1) depending on age(Rossi et al., 2004; Oliviero et al., 2006). Other studies suggest, how-ever, a possible association between healthy aging and an increasein motor threshold (Rossini et al., 1992; Peinemann et al., 2001).Such discrepancies likely originate in technical and experimental
differences.
Stimulation over non-primary motor areas often does not pro-duce as readily an objective, quantifiable response (but see speecharrest on left inferior frontal gyrus (IFG)/Broca’s area and slowing ofcontraversive saccades during frontal eye fields (FEF) stimulation).

Biobeh

Ot(sett

up

oeiasMtiTnmoc

tcasinbtmti


ne important exception is, in most individuals, the visual cor-ex (e.g., V1, V4 and V5), whose stimulation can elicit phosphenesbright spots of light in the visual field) that can be reported byubjects but cannot be directly quantified by observers (Boroojerdit al., 2000). While phosphenes elicited by V1-TMS are stationary,hose evoked by TMS on V5 are often moving. Their characteristicsherefore reflect the perceptual specialization of a given area.

Analogous to MT, phosphene threshold (PT) is the lowest stim-lation intensity at which phosphenes (stable or moving) areerceived in at least 5 of 10 stimulations.

For example, Salillas Pérez et al. (2009) delivered TMS (110%f PT) over the ventral intraparietal sulcus (vIPS) to determineffects on performance in motion detection and number compar-son tasks. The rationale of this study is based on data supportinglink between representations of number and space (for reviews,

ee Fias and Fischer, 2005; Hubbard et al., 2005; Umiltà et al., 2009).ore precisely, because motion perception can influence visuospa-

ial imagery processes (Logie, 1995), it was predicted that it couldnfluence numerical representations as well. The finding that vIPS-MS results in impaired efficiency in both motion perception andumber comparison may suggest that these processes share a com-on neural substrate, although, given the limited spatial resolution

f TMS when used as a virtual lesion, this should be interpreted withaution.

Traditional methods of determining stimulation threshold haveo take also into consideration two parameters that have a signifi-ant impact on it: stimulus waveform (monophasic vs. biphasic)nd current direction. As for waveform, in the past most TMStudies on M1 were performed with monophasic waveforms,n particular those examining MT (Mills and Nithi, 1997). Theew generation of stimulators is however capable of producing
iphasic pulses, and studies on M1 suggest that, for a given ampli-ude of initial current, biphasic stimulation is more effective than
onophasic stimulation (Sommer et al., 2006). As for current direc-ion, in order to obtain responses by lower stimulus intensitiesn M1, induced currents from monophasic pulses should flow in

Fig. 1. Schematic illustration of different approac

avioral Reviews 35 (2011) 516–536 519

a posterior-to-anterior direction, and from biphasic pulses in theposterior-to-anterior direction during the second phase, and thus inthe opposite direction from the first phase (Kammer et al., 2001a).This suggests that the coil should be positioned in opposite direc-tions depending on whether a monophasic or a biphasic pulse isused for stimulation. In the visual cortex, PT is lower if the direc-tion of the induced current is oriented form lateral to medial inthe occipital lobe rather than vice versa (Kammer et al., 2001b).The seemingly reasonable assumption that a relevant proportion ofTMS threshold measures from different neocortical regions couldreflect a shared component of within-individual responsiveness toTMS has been criticized by studies that did not find any correlationbetween MT and PT. It should, however, be noted that such stud-ies did not take intra-individual variations in scalp–cortex distancebetween M1 and V1 in account (Antal et al., 2003; Boroojerdi et al.,2002; Gerwig et al., 2003; Stewart et al., 2001; but see Deblieck etal., 2008; see also Oliver et al., 2009, for the finding of a correlationbetween MT and the intensity level at which TMS is effective on aparietal site). As a consequence, it has become common practiceto use a fixed intensity defined as a percentage of the stimula-tor output (see Cappelletti et al., 2007; Cattaneo et al., 2009b;Cohen Kadosh et al., 2007a; Dormal et al., 2008; Knops et al., 2006;Rusconi et al., 2005, Exp. 2; 2007). This approach reduces experi-ment duration and limits the number of TMS pulses, by eliminatingthe preliminary phase of threshold hunting. Stimulation intensityis usually fixed in these studies at the lowest stimulation intensitythat can successfully affect behavior when TMS is delivered overthe area of interest, based on the literature and/or the researcher’sprevious experience (see Fig. 1 for a schematic summary regardingstimulation intensity).

On a conclusive note, it is worth mentioning that a compre-
hensive and individually tailored approach (encompassing not onlystimulation intensity but also brain-specific coil orientation, stim-ulus duration, etc.—see below) is not yet possible with currenttechnical facilities; it may be implemented, however, in a not-so-far future, when fine-grained information regarding individual
hes used to set the intensity of stimulation.

5 Biobeh

atmntf

3

asapcoaftia

odposno2gsuibflbibna+t

dh

eowca(sa

ieds

mtht


natomy and function data will be made available to neuronaviga-ion systems (Wagner et al., 2009). Ideally it will also be possible to

odel the current flow over different areas by taking into accountot only scalp–cortex distance but also the specific tissues, withheir conductivity characteristics, that fill such distance under dif-erent scalp regions.

.2. TMS paradigms

TMS can be applied as single pulses (spTMS) or as trains of pulsesnd in the latter case it is named repetitive transcranial magnetictimulation (rTMS; for review see Pascual-Leone et al., 2000). Bypplying spTMS at variable times during task execution (on-linearadigm, where both stimulation and task performance occur con-urrently), it is possible to investigate with a temporal resolutionf a few tens of milliseconds (ms) at what exact time point neuralctivity at the stimulation site is critical for successful task per-ormance (chronometry of functional relevance). In spTMS studies,he interval between pulses has to be sufficiently long to preventnteractions between consecutive pulses. However, an interval ofbout 6–7 s between pulses may be sufficient.

Andres et al. (2005) used spTMS to investigate the involvementf a portion of the bilateral IPS in number comparison, which isirectly related to the representation of number magnitude. TMSulses were delivered over left and right IPS, either unilaterallyr bilaterally (in which case two coils were made to dischargeimultaneously), while participants were engaged in a single-digitumber comparison task against a fixed standard (5). The timingf the TMS pulses was varied: stimulation could occur 150, 200, or50 ms after digit presentation. When the distance between tar-et and reference numbers was small, responses were significantlylower in unilateral left and bilateral stimulation conditions than innilateral right stimulation and control conditions. When numer-

cal distance was large, responses were significantly slower in theilateral stimulation only. Left and right parietal cortices, there-ore, showed a different resolution in single-digit comparison, theeft being necessary for fine discrimination, and both right and lefteing able to support coarser comparisons. What remains unclear

s whether, by having a larger range of TMS timings, effects coulde modulated. Event-related potentials (ERPs) studies indicate thatumber semantic processing from Arabic input occurs between 170nd 240 ms (Dehaene, 1996). Introducing TMS timings beyond the150/250 ms window might thus reveal a more complex picture ofhe interaction between left and right IPS.

A nice demonstration, of how TMS can be used to differentiateistinct time points of functional relevance within a brain region,as been provided by Chambers et al. (2004).

The authors applied spTMS to the parietal cortices, at 12 differ-nt time points (30-ms intervals) starting at 30–360 ms after thenset of the target, to reveal which part of the parietal cortex, and athat time, is functionally relevant for spatial attention shifts. They

ould reveal that activity in the right angular gyrus (AG) is crucialt two distinct time points during reorienting of spatial attentionbetween 90 ms and 120 ms; and between 210 ms and 240 ms aftertimulus onset), probably reflecting receipt of information via twonatomically distinct visual pathways.

The possibility to cover a larger time window is offered by rTMSn which pulse trains with rates up to 50 Hz are delivered. Length-ning the duration of on-line rTMS is presumed to cause moreisruption by virtue of temporal summation of the effects of thetimulation.

Sandrini et al. (2004) found that rTMS applied over left supra-arginal gyrus (SMG) slowed down RTs when subjects were asked

o select the largest digit among a pair, whereas right-SMG rTMSad no effect. However, rTMS did not interact with numerical dis-ance. This result could therefore be interpreted as evidence that


the processing chain is interfered with before accessing magnitude(i.e., operations between input and semantics are slowed down).Another possible explanation is interference on motor responseselection, a process in which the left SMG is thought to play a fun-damental role (e.g., Rushworth et al., 2001). Rusconi et al. (2007)found an involvement of the anterior portion of posterior parietalcortex (PPC) in a typical stimulus–response (S–R) correspondence(Simon) effect, but not in the Spatial-Numerical Association ofResponse Codes (SNARC) effect, thought to be driven by numbermagnitude. In contrast, they found that the posterior portion ofPPC disrupted the numerical S–R correspondence effect. Likewise,Göbel et al. (2001) found interference, mostly on the left poste-rior PPC, when participants were performing a magnitude task andrTMS was delivered on stimulus presentation.

The spTMS-induced change in neuronal activity can last forroughly 40–60 ms (Amassian et al., 1989; Brasil-Neto et al., 1992b).If short burst of rTMS are applied, the influence of TMS on neu-ronal activity can be prolonged (Pascual-Leone et al., 2000). Thisimplies that temporal specificity of rTMS is much less than spTMSbut with the advantage to be more suited for inducing measurablebehavioral effect sizes in higher cognitive functions than spTMS.An interesting rTMS approach to get insights into the temporalcharacteristics of the involvement of a brain area during a cog-nitive process has recently been adopted by some studies (Rossiet al., in press; Schuhmann et al., 2009). For example, Schuhmannet al. (2009) investigated the temporal dynamics of Broca’s areainvolvement in picture naming using triple-pulse TMS (inter-pulseinterval of 25 ms) at various time points after picture presentationonset, namely at: (1) 150–175–200 ms, (2) 225–250–275 ms, (3)300–325–350 ms, (4) 400–425–450 ms and (5) 525–550–575 ms.The results showed an increase in RTs only when applied in the300-ms range after picture presentation. These data confirm therelevance of Broca’s area during picture naming and provide newinsight into the temporal characteristics of its specific contributionduring speech production.

RTMS can be also applied in an off-line paradigm, in which mag-netic stimulation and task performance are dissociated in time. Ina common approach, performance is measured initially and thenrTMS is usually applied over a site of interest for several minutes(from 10 to 25), generally at a low-frequency (≤1 Hz), then perfor-mance is measured again. This paradigm is based on data showingreduction of visual (Boroojerdi et al., 2000) and motor cortex (Chenet al., 1997) excitability beyond the duration of the rTMS applicationitself.

Previous studies show that the duration of the after-effectsis short-lasting, that is at least half the duration of the stimula-tion train (Robertson et al., 2003). However, these studies onlyreport behavioral measures and provide no information about theduration of the neurophysiological effect of off-line low-frequencyrTMS. Recently, Eisenegger et al. (2008) assessed the duration ofthis effect after stimulation of the right dorsolateral prefrontalcortex (DLPFC). Subjects underwent positron emission tomogra-phy (PET) scans after application of long-train low-frequency rTMS(1 Hz for 15 min). Immediately after the stimulation train, rCBFincreases were present under the stimulation site as well as in otherprefrontal cortical areas, including the ventrolateral prefrontal cor-tex (VLPFC) ipsilateral to the stimulation site. The mean increasesin rCBF returned to baseline within 9 min, showing that the dura-tion of rCBF changes in the prefrontal cortex is approximately 60%the duration of the stimulating train itself.

However, it should be kept in mind that the after-effect can vary
a lot in its duration, depending on the context. Off-line paradigmscan avoid nonspecific disruption of performance due to discomfort,noise, muscle twitches and intersensory facilitation associated withrTMS during the task. Moreover, compared to on-line paradigms,they require the use of a task that lasts no longer than the duration

Biobeh

otdsati

(fthecteirs2rblremblsc

3

TlfsMwemwmtanfpoiei2He

psoi(tput


f the after-effect and they miss any kind of specificity in terms ofemporal involvement of a cognitive process. Moreover, the totaluration of an experiment is longer because before proceeding totimulate a new site one has to wait for effects of rTMS to wash outnd sometimes the session has to be split in different days. As forhe interpretation of results, in the presence of facilitation effects,t is important to rule out learning effects.

Off-line rTMS at 1 Hz has been used extensively in recent yearsChen et al., 1997). It was applied to investigate a variety of higherunctions over different brain areas, such as parietal contributionso spatial attention (Hilgetag et al., 2001; Thut et al., 2005), spatialearing (Lewald et al., 2002) and number processing (Cappellettit al., 2007; Knops et al., 2006; Rusconi et al., 2005) or prefrontalontributions to working memory (Mottaghy et al., 2002), affec-ive processing (d’Alfonso et al., 2000) decision making (Knocht al., 2006) or supplementary motor area (SMA) to motor learn-ng (Perez et al., 2007; Tanaka et al., in press). Moreover, it wasecently applied to evaluate the role of anterior temporal lobe inemantic cognition (Pobric et al., 2009, 2010; Lambon Ralph et al.,009). The off-line protocol facilitated the use of rTMS over a brainegion where rTMS is generally more unpleasant than on otherrain areas, and it is therefore rather difficult to employ in an on-

ine high-frequency protocol. In some studies this issue has beenesolved using custom made butterfly coil with a smaller diam-ter (40 or 50 mm). These small coils with very short TMS trainsinimize discomfort from oral-facial muscle movement caused

y stimulation of brain areas such as temporal lobe or ventro-ateral prefrontal cortex (see also Rusconi et al., in press, for atimulation protocol applied to right IFG with 50-mm diameteroils).

.3. Stimulation frequency

When targeting the motor system using an off-line paradigm,MS effects have been fairly consistent and appear to causeong-lasting neurophysiologic effects. Generally speaking, low-requency rTMS (≤1 Hz) decreases whereas high-frequencytimulation (≥1 Hz) increases cortical excitability as assessed byEP (Fitzgerald et al., 2006; but see also Rothkegel et al., 2010 inhich breaks during 5 Hz rTMS are essential for facilitatory after-

ffects). When targeting cognitive functions, the picture becomesore complex. By assuming that suppression of cortical activityithin a specific cortical target can significantly impair perfor-ance (e.g., Robertson et al., 2003), one could erroneously expect

hat increased excitability will lead to a behavioral improvementnd decreased excitability will be detrimental to performance. Untilow we know of only one study that has used an off-line, high-

requency rTMS paradigm to convincingly demonstrate enhancederformance (Kim et al., 2005). In contrast, some studies applyingff-line low-frequency rTMS have actually shown improvementsn behavior (Andoh et al., 2006, 2008; Dräger et al., 2004; Hilgetagt al., 2001; Snyder, 2009). Similar results have been also foundn studies applying on-line high-frequency rTMS (Boroojerdi et al.,001; Cappa et al., 2002; Kirschen et al., 2006; Köhler et al., 2004;amidi et al., 2008; Luber et al., 2007; Mottaghy et al., 1999; Sparingt al., 2001).

Stimulation after-effects may also depend on the pattern of theulses applied. For instance, Huang et al. (2005) used a pattern oftimulation named theta burst stimulation (TBS), in which trainsf 50 Hz (3 pulses) are applied every 200 ms (frequency of 5 Hz,n the range of theta band) at low intensity (80% aMT) for 20–40 s
continuous TBS, cTBS). They demonstrated that this TBS protocolemporarily reduces the excitability of motor cortex outlasting theeriod of actual TMS (for 20 s of cTBS: up to 20 min, for 40 s of cTBS:p to 1 h). However, in the same study (Huang et al., 2005), if, ratherhan using one long period of TBS, each burst is applied only for 2 s
avioral Reviews 35 (2011) 516–536 521

followed by a pause of 8 s and then repeated for 190 s (intermittentTBS, iTBS), the effect becomes facilitatory.

The cTBS protocol, aimed at exerting long-lasting disruptiveeffects, has been used in few off-line cognitive studies (Catmuret al., 2009; Cho et al., in press; Davare et al., 2010; Galea et al.,2010; Kalla et al., 2009; Rosenthal et al., 2009; Vallesi et al., 2007;Wilkinson et al., 2009) (see Fig. 2a). Some other studies have usedmodified TBS protocols. For instance, Nyffeler et al. (2006, 2008)applied a variant (three pulses at 30 Hz, separated by an interval of100 ms resulting in a repetition frequency of pulses at 6 Hz) to rightfrontal eye field before an oculomotor task (precisely, 200 bursts at80% of rMT for a total duration of 33 s; Nyffeler et al., 2006) andto right posterior parietal cortex before visual exploration behav-ior (precisely, 267 bursts at 90% of rMT for a total duration of 44 s;Nyffeler et al., 2008; see Fig. 2b). Such specific TBS protocol hasbeen shown to cause behavioral changes that last for at least 30 min.Silvanto et al. (2007) also used a customized protocol (8 pulses at40 Hz, separated by 1800 ms for a duration of 50 s with a stimula-tion intensity of 60% of maximum stimulator output) over V1–V2during a motion-direction-discrimination task (see Fig. 2c).

In studies seeking to compare the effects of different frequencyor patterns of pulses, it is fundamental to dose the number of pulsesin the two protocols. Thus, in protocol of continuous TMS (low-frequency or cTBS) pulses are applied in a continuous train, whereasin protocols of intermittent TMS (high-frequency and iTBS) shortperiod of rTMS is separated by periods of no stimulation (see Fig. 2for a schematic summary regarding TBS protocols).

Relatively short-term effects, of the order of seconds to a fewminutes, could be due to changes in neural excitability caused byshifts in ionic balance around populations of active neurons oreven the electrical capacitative effects storing charge induced bythe stimulus. On the other hand, longer-lasting effects presumablyinvolve changes in the effectiveness of synapses between corticalneurons (long-term depression, LTD, and long-term potentiation,LTP, of synaptic connections; Ridding and Rothwell, 2007).

Conventional ‘off-line’ rTMS needs a longer application time(between 10 and 25 min) and produces shorter-lasting effects (i.e.,in the range of 5–20 min) than patterned off-line rTMS. It thereforeseems plausible that, for future studies, TBS techniques may be theelective option for ‘off-line’ rTMS studies on higher functions.

4. Dependent variables and effect directions

A TMS-induced change in behavior (in the form of RTs, accu-racy or signal detection measures) can be used to inform models ofcausal relations between specific brain regions and individual cog-nitive functions. TMS is generally used in cognitive neurosciencewith the aim to disrupt some specific neural activity that subservescognitive processing (i.e., as a ‘virtual lesion’). However, in sev-eral studies, certain cognitive tasks have been in fact enhancedby TMS, thus revealing the potential of TMS-induced paradoxicalfunctional facilitation (e.g., interhemispheric or intrahemisphericinteractions; Sack and Linden, 2003; Théoret et al., 2003; Wardand Cohen, 2004; facilitation is defined paradoxical as it is gen-erally attributed to disinhibition of connected distant areas ratherthan actual improvement of the computation of the targeted site).This interpretation is based on animal studies and a few humancase reports which suggest that direct or indirect neural ‘damage’to specific areas in the central nervous system may result in facilita-tion of behavioral functions (see Kapur, 1996). Causality, therefore,
is always assumed when a significant effect occurs (whether facili-tative or inhibitory). Most TMS studies predict changes in behavior,but are rarely explicit about their direction (if they are, it is proba-bly a post hoc statement disguised as an a priori hypothesis). Thisis due to the fact that for cognitive functions, the way in which

522 M. Sandrini et al. / Neuroscience and Biobehavioral Reviews 35 (2011) 516–536

F ursts( terval1

Ttdbapiq(ettes2tinid

iea

todorstbit

ig. 2. Schematic illustration of different TBS protocols. (a) Protocol in which short biTBS) train; (b) protocol in which after the burst (3 pulses at 30 Hz) there is an in800 ms.

MS modulates behavior (i.e., if it will exert facilitatory or disrup-ive effects) depends upon a number of variables, such as intensity,uration, frequency, and pattern of pulses applied, as well as possi-le interactions between them. In addition, other factors may playn important role, such as brain region and the cognitive task beingerformed (Walsh et al., 1998). Regarding stimulation parameters

t is important to mention that in some TMS studies different fre-uency of stimulation on the same site and task led to similarHilgetag et al., 2001; Jin and Hilgetag, 2008) or opposite (Knocht al., 2005) effects, whereas in other studies identical stimula-ion parameters (frequency, duration and intensity) and site ledo opposite results in different tasks (Cappa et al., 2002; Sandrinit al., 2003; Rossi et al., 2001). Moreover, studies show on theame task opposite effects between different areas (Dräger et al.,004) or between homologous areas of the human cerebral cor-ex (Cappelletti et al., 2007). For example, Cappelletti et al. (2007)nvestigated the causal role of IPS in processing both two-digitumbers and dots: in either case, performance was slower follow-

ng left-IPS rTMS, and faster following right IPS rTMS (attributed toisinhibition of the homologous region in the left hemisphere).

Such evidence suggests that different protocols may induce sim-lar effects on the same area or same protocols may induce oppositeffect, and that no simple inference may be drawn from one corticalrea to another or from one cognitive function to a related one.

Another important point that should be taken into considera-ion is that TMS can induce a range of behavioral effects dependingn the initial activation state of the stimulated region (“state-ependency” see Siebner et al., 2009a; Silvanto et al., 2008). Inn-line paradigms, these effects can be either facilitatory or dis-uptive depending on the time point of stimulation. For example,
pTMS usually disrupts cognitive functions when applied duringhe time period in which the stimulated area contributes to the task,ut a number of studies have reported facilitation effects when TMS
s delivered early in the time course of a trial – before activation inhe region was expected (Grosbras and Paus, 2003; Pulvermüller

of 50 Hz (3 pulses) are repeated every 200 ms as a continuous (cTBS) or intermittentof 100 ms; (c) protocol in which short bursts of 40 Hz (8 pulses) are separated by

et al., 2005; Stoeckel et al., 2009; Töpper et al., 1998). When TMSis delivered before the onset of a cognitive process, neural pop-ulations are less influenced by central factors than when TMS isapplied during a cognitive process. This difference in the initialneural activation state could explain why the effects of TMS areopposite in the two circumstances. However, further studies arenecessary to evaluate the validity of this interesting hypothesis.Moreover, although the neural mechanisms of long-lasting effectsof off-line rTMS are different from those induced by spTMS or shortbursts of rTMS, they have also been shown to be state-dependent(Andoh et al., 2008; Brighina et al., 2002; Lang et al., 2004; Siebneret al., 2004). Understanding state-dependency is thus crucial foran accurate interpretation of TMS studies (Pasley et al., 2009). Inconclusion, the available evidence shows a very complex picture inrelation with the application of TMS in cognitive studies. Factorssuch as experimental design, TMS parameters, type of task, stim-ulated brain area and its neural activation state are all important,as well as their possible interactions, which makes sometimes dif-ficult to generate clear predictions about the direction of the TMSeffect (facilitation vs. interference).

Moreover, the precise neurological mechanisms underlyingfacilitatory effects of rTMS on behavior are far from understood andeven more obscure than the mechanisms of interference (Rossi andRossini, 2004; Sack and Linden, 2003, but see Miniussi et al., 2010for a fresh interpretative approach).

Perhaps less ambiguously, TMS can be used to measure func-tional connections between cognitive processing and activity inthe visual cortex as well as the motor cortex. This can be doneby measuring the change in the PT or in the amplitude of MEPsunder various TMS experimental conditions and cognitive tasks. For
example, it is possible to use PT to investigate whether a given taskmodulates visual cortical excitability. A recent study by Cattaneoet al. (2009d) showed that number priming modulates excitabil-ity of the visual cortex in a topographic fashion: small numbers,associated with left side space, increase the excitability of the right

Biobeh

evTtvuabt

ttFnaailrimicettmicr

T(1atbsm6t7ottTloirpvfa

tipoovctspS


arly visual cortex and decrease the excitability of the left earlyisual cortex. The opposite pattern was found for large numbers.hese data suggest that the attentional shifts induced by the men-al number line can be traced down to the earliest cortical stages ofisual processing. Regarding the motor cortex, when a task mod-lates MEPs in a region of motor cortex, this provides evidence offunctional link between the task and the motor cortex. MEPs cane recorded in the motor cortex that represents the hand, the arm,he leg, the foot and even the tongue (Fadiga et al., 2002).

Two studies (Andres et al., 2007; Sato et al., 2007) investigatedhe relation between numbers and hand/finger motor representa-ions. Andres et al. (2007) employed TMS over M1 and found rightDI MEPs increase during serial counting, regardless of the use ofumbers or letters to keep track of items, but no increase in armnd foot MEPs. They suggested that hand motor circuits are activebove baseline whenever a correspondence is to be drawn betweentems and ordered series. Sato et al. (2007) measured abductor pol-icis brevis (APB) and abductor digit minimi (ADM) MEPs from bothight and left hands in a parity judgment task, and found an increasen MEPs for the right hand only during small number processing, a

odulation they relate to their participants’ counting habits. It ismportant to know that a modulation of the motor system can alsoonsist of an amplitude decrease of the recorded MEPs (Buccinot al., 2005). Note that TMS is used in this fashion to investigatehe causal role of a given cognitive task on the activation state ofhe motor cortex rather than to investigate the causal role of the

otor cortex in that task. Change in motor cortex excitability dur-ng counting shows that this cognitive process can recruit motorortex – it does not tell whether motor cortex plays a necessaryole in counting.

Since changes in resting MEPs, as measured with single pulseMS are not an unequivocal measure of activity in cortical areasDi Lazzaro et al., 2001), the use of paired pulse TMS (Kujirai et al.,993) should be preferred in future studies. In paired-pulse TMS,conditioning stimulus (CS) below the threshold intensity needed

o elicit an MEP is followed at short interstimulus intervals (ISIs)y a suprathreshold test stimulus (TS). The effects of the CS on theize of a test MEP depend on stimulus intensity and ISIs. Maxi-um inhibitory effects are found at short ISIs of 1–4 ms and CS of

0–80% of the rMT (Kujirai et al., 1993; Schäfer et al., 1997). Facili-atory effects of the CS on the test MEP can be observed at intervals–20 ms (Kujirai et al., 1993; Ziemann et al., 1996). The magnitudef the intracortical inhibition and facilitation vary depending onhe amplitude of the test MEPs and the degree of contraction ofhe target muscle, a critical variable to control for in paired-pulseMS studies. This modulation of MEP size takes place at the corticalevel and is thought to reflect the activation of separate populationsf inhibitory and excitatory cortical interneurons without affect-ng spinal circuits (see Di Lazzaro et al., 2008; Reis et al., 2008, foreviews about physiological mechanisms of this technique). Henceaired-pulse TMS provides a reliable index of motor cortical acti-ation (see Oliveri et al., 2004a). Such method was first introducedor the study of motor cortex but can also be applied to non-motorreas (Oliveri et al., 2000).

This technique has also been used to study functional interac-ions between two connected brain areas. In a creative protocol,n which one pulse is applied over each of two brain areas, it isossible to investigate functional cortical interactions on a subsec-nd timescale (dual-site paired-pulse TMS). In the motor system,ne can test the causal influence of a connected area on the acti-ation state of M1 by first stimulating that area, and measuring
onsequent changes in the amplitude of MEPs. A CS is first usedo activate putative pathways to the motor cortex from the site oftimulation, whereas a second TS delivered over M1 a few ms laterrobes any changes in excitability that are produced by the input.ome studies have been conducted (either at rest or during tasks)
avioral Reviews 35 (2011) 516–536 523

with the CS delivered over contralateral M1 (Ferbert et al., 1992),cerebellum (Ugawa et al., 1995), pre-SMA (Mars et al., 2009), pre-motor cortex (Baumer et al., 2006; Boorman et al., 2007; Buch etal., 2010; Davare et al., 2008, 2009, 2010; Koch et al., 2006, 2007a;O’Shea et al., 2007c) and parietal cortex (Koch et al., 2007b, 2008,2009), confirming the existence of these pathways in humans.

Analogous studies have been conducted in the visual system,where PT evoked by TMS over V1 or V5 has been used to indexcortical excitability. These experiments have tested functional con-nectivity in the visual system by measuring how a prior TMSpulse over another area changes PT over V1/V5 (Pascual-Leone andWalsh, 2001; Silvanto et al., 2006, 2009). The particular advantageof this protocol is that it can reveal with subsecond resolution howactivity changes in one brain area causally impact on activity in con-nected areas (see for reviews Koch and Rothwell, 2009; O’Shea etal., 2008). Finally, interaction studies need not be limited to primarymotor and sensory areas. Performance modulations after dual-sitecompared to single-site TMS, can also provide a marker of causalinteraction (Ellison et al., 2007) (see Fig. 3 for a schematic summaryregarding TMS protocols).

Finally, we will conclude this section with a mention that gen-der and aging may modulate TMS-induced effects. As for gendereffects in TMS studies, only few studies reported different effectsbetween male and female participants (De Gennaro et al., 2004;Knops et al., 2006). Evidence is instead more robust for age effects.For instance, traces of elementary motor-related memories, reflect-ing synaptic adaptation properties, can be revealed by TMS. Somestudies showed that the ability to encode such motor-related mem-ories seems to be age-dependent, as it progressively diminisheswith age (Celnik et al., 2006; Sawaki et al., 2003). Furthermore, neu-roimaging studies suggest that lateralization of prefrontal cortexactivation associated with episodic memory declines with aging.The effects of rTMS applied to the left or right DLPFC on a visu-ospatial recognition memory task were compared in a populationof healthy subjects divided into two classes of age (<45 and >50years). In younger subjects, rTMS of the right DLPFC interfered withretrieval more than left DLPFC stimulation. The asymmetry of thiseffect was reduced with aging, and bilateral interference effectswere found on memory performance. These data suggest that thebilateral engagement of the DLPFC may have a compensatory roleon episodic memory across the life span (Rossi et al., 2004).

5. Control conditions

Different control conditions can be used to try and ensure thatchanges in performance be ascribed to TMS effects upon a specificbrain area. One of the most common strategies is the use of shamstimulation. TMS is indeed associated with a number of sensoryperceptions that can nonspecifically interfere with task perfor-mance. For instance, the discharging coil produces a click soundthat may induce arousal, thereby modulating task performance,irrespective of the experimental demands (i.e., via intersensoryfacilitation; see Hershenson, 1962; Marzi et al., 1998). Sham rTMSstimulation is generally carried out by tilting the coil away from thescalp (Sandrini et al., 2004), so that both sound and scalp contactare roughly similar to those experienced during active stimulation,whereas the magnetic field does not reach cortical neurons or cuta-neous receptors or superficial muscles. Nowadays, also sham-coilsare commercially available. They still produce the same sound asduring stimulation, but no magnetic field is generated, so that they
can rest tangential to the scalp surface exactly as they are dur-ing active stimulation (see Cappelletti et al., 2007; Cohen Kadoshet al., 2007a). However, neither of the above sham conditions isfully satisfactory because none of them produces exactly the samesensations as a real TMS application. The introduction in the mar-


ocols s

kip(tostbw

TslRvedbsTwitoar

Fig. 3. Schematic illustrations of TMS prot

et of a new placebo tool, which can simulate the scalp sensationnduced by real TMS, offers a more satisfactory solution. For exam-le, Rossi et al. (2007) developed a Real Electro Magnetic PlaceboREMP) device, which is built in compact wood and it is attachedo the real coil by Velcro strips. It has two main functions: the firstne is to act as a physical barrier between the real coil and thecalp and the latter is to house a bipolar electrical stimulator withinhe surface that is in contact with the scalp. Such placebo TMS haseen demonstrated to be superior to the available sham procedureshen applied on subjects naïve to TMS.

An alternative way that is routinely used in the cognitiveMS literature is vertex stimulation because the auditory andomatosensory activations caused by vertex TMS can be equiva-ent to those of real TMS (Dormal et al., 2008; Knops et al., 2006;usconi et al., in press). Of course, the underlying assumption is thatertex TMS does not affect the cognitive network active during taskxecution. There are also studies in which rTMS is delivered ran-omly over the experimental site on half of the trials. This seems toe the ideal condition to control for carry-over and practice effects,ince the experimental and control condition (i.e., trials withoutMS) are in fact performed at the same time during a session, andith the coil in exactly the same position. Whereas the procedure of

nterleaving stimulation and no stimulation trials controls for unin-ended “spill-over” effects of rTMS, because any residual effectsf rTMS would influence control trials as well as stimulation tri-ls (Göbel et al., 2001, 2006; Rusconi et al., 2005), the possibilityemains that TMS exerts after-effects from one trial to the other if a

howing how TMS or rTMS can be applied.

short intertrial interval (ITI) is used. Rusconi et al. (2007), for exam-ple, employed a 1000-ms interval (ITI), and found a block ratherthan a trial-specific effect. The general assumption in on-line rTMSstudies is that the effect of rTMS is restricted to the stimulation win-dow (see e.g., Göbel et al., 2001, who also employed a 1000-ms ITI inmixed blocks but found no spill-over). It is possible, however, thatdepending on the importance and the state of a stimulated area, ablock rather than a trial effect will be found. Therefore the intra-block manipulation of rTMS condition has to be used with somecaution. Other approaches can be employed to ensure that changesin performance are indeed specific for TMS over a given brain area.One of these is to stimulate more than one brain site (Cappellettiet al., 2007; Cattaneo et al., 2009b; Göbel et al., 2001; Kansaku etal., 2007; Rusconi et al., 2005, 2007). If the effects of stimulationare observed exclusively at one site, then this gives some reassur-ance that the differences across sites are due to the specific effectsof neuronal disruption. Within this approach, one possibility is tostimulate the homologous area in both hemispheres in particularduring experimental designs regarding the lateralization of cogni-tive processes (Andres et al., 2005; Cattaneo et al., 2009b; Dormal etal., 2008; Sandrini et al., 2004; Rusconi et al., 2005, 2007). For exam-ple, Cattaneo et al. (2009b) used a priming paradigm with a line
bisection task to assess the bias in spatial allocation of visual atten-tion induced by exposure to either small (16–24) or large (76–84)ends of the mental number line. They applied rTMS over the left andright AG. In the small number prime condition, in which attentionis presumably biased to the left side of visual space, TMS applied

Biobeh

oatloinRoevr

csnsaoedaodcadtoociaTicrnapi

2sTtogmsdaofneiobfiabniT


ver the right AG abolished the effect of number priming, whereaspplication of TMS over the left AG had no significant effect. In con-rast, both left and right AG TMS modulated the effect of priming byarge numbers (which presumably shift attention to the right sidef visual space). These findings show that the AG plays a critical rolen the allocation of visuospatial attention modulated by the mentalumber line. Another possibility is to stimulate different brain sites.usconi et al. (2007) reduced the SNARC effect by applying rTMSver left and right posterior, but not anterior portion, of PPL. Theffect was attributed to disruption of the link between numbers andisuospatial attention rather than to interference with core numberepresentations.

Relatively small changes in position can cause substantialhanges in the sensory effects of stimulation. Moreover, even takingpatial specificity for granted, interference with a single task doesot contain enough information to allow testing specific hypothe-is about the cognitive operation that is being interfered with. Asconsequence, observing behavioral dissociations induced by TMSn a given area across distinct tasks (Cappelletti et al., 2007; Dormalt al., 2008; Kansaku et al., 2007) or even in the same task but forifferent markers (e.g., Rusconi et al., 2007) may be as importants observing behavioral dissociations between stimulated sites. Inther words, the inclusion of appropriate control tasks and/or con-itions within tasks in the experimental design can enhance theognitive resolution of TMS studies. A recent study (Dormal etl., 2008) tested a possible dissociation between numerosity anduration processing in the parietal cortex using a task in which par-icipants had to compare the numerosity of flashed dot sequencesr the duration of single dot displays. RTMS was applied over the leftr right IPS or the vertex, chosen as a control site. Compared to theontrol site, performance was slowed down only for the numeros-ty comparison task after left-IPS stimulation, whereas it was notffected in the duration comparison task for either parietal sites.hese data, together with Cappelletti et al.’s study, confirm recentmaging studies in humans and cell-recording in primates indi-ating that, besides comparison of symbolic stimuli, the parietalegions are also involved in quantity processing of non-symbolicumerosities (e.g., Castelli et al., 2006; Piazza et al., 2007; Niedernd Miller, 2004). Moreover, they show that at least one of thearietal areas that are critically involved in numerosity processing

s not crucial for duration processing.Following up on their parietal stimulation study (Rusconi et al.,

007), Rusconi et al. (2008, in press) moved the exploration of theubstrate of number–space interaction forwards, in the frontal lobe.hey found that suppression of the SNARC effect is dependent onask and number magnitude with right FEF-rTMS and is dependentn task but generalized to all magnitudes with right IFG-rTMS. Ineneral, when simple dissociations between tasks are found (i.e.,odulation of performance in one task but not in the other when

timulating the same site), it is important to evaluate whether theifferential effect of TMS may be attributed to global task char-cteristics such as relative difficulty rather than specific mentalperations. It is indeed possible that the manipulation is power-ul enough to modulate performance in the more difficult task butot in the easier one, which can still be performed at baseline levelsven under modified neural conditions. This caveat is most popularn the neuropsychological literature. One other issue that may beften overlooked is that a reversed type of argument needs also toe taken into account in TMS studies. Rusconi et al. (2008, in press),or example, found a selective suppression of the SNARC effect dur-ng the easier task, that is magnitude comparison, which was on
verage 50-ms faster than parity judgment. In that case, therefore,ased on behavioral performance the task difficulty objection couldot be applied. Moreover, in Rusconi et al.’s study, the difference
n terms of RTs between the two tasks was relatively small and theMS effectiveness window could be assumed to cover all the time

avioral Reviews 35 (2011) 516–536 525

from stimulus to response selection in either task (Wassermann etal., 2008). In theory, cases where a similar dissociation is present,which escapes the difficulty objection since an index in the eas-ier and not in the more difficult task is modulated by TMS, mighthowever be detecting spurious effects. If the critical process that isaffected by TMS occurs quite some time after stimulus presentation,it is possible that only the faster task reveals an effect because theTMS protocol covers all of the important processing stages, whereasit fails to cover all processes in the slower task. It is therefore veryimportant, before deciding on a specific TMS protocol, to carry outa task analysis and specify what processing stages are expected tobe involved and show sensitivity to the TMS manipulation.

In conclusion, experimenters have to choose carefully theappropriate experimental design to use in order to test theirhypothesis and their theoretical framework. The use of more thanone control condition is crucial for determining the contributionmade by a cortical area to a specific behavior (see Fig. 4 for aschematic summary regarding control conditions).

6. Site localization

The spatial resolution of TMS depends on the geometry of thecoil, stimulus intensity, stimulus configuration and the electricalproperties of the cortex under the coil. Its temporal resolutiondepends on the duration of TMS and the duration of an area’sinvolvement in the task. In studies on cognitive functions, the infor-mation about the when and where of TMS application is derivedfrom previous studies using complementary brain mapping tech-niques [ERPs, MEG, PET, functional magnetic resonance imaging(fMRI)] or neuropsychological data with the task of interest. Evenwhen exact temporal and spatial information are available, theprecise and reliable positioning of the TMS coil is not a trivialissue.

Some areas of interest in cognitive TMS studies, when stimu-lated, have a signature output (for instance, M1 and V1). However,many others are behaviorally silent. A possible approach consistsof locating brain areas relative to those that have a reasonably cer-tain position. For example, DLPFC can be defined as 5 cm anteriorof the region from which the most prominent motor response ofthe muscle APB can be recorded (Pascual-Leone et al., 1996) or thelateral cerebellum can be defined 1 cm under and 3 cm to the left orto the right of the inion (Torriero et al., 2004; Théoret et al., 2001).

A different approach of localizing the TMS target is the applica-tion of TMS in relation to specific individual scalp landmarks, forexample, based on the standard 10–20 electrode scalp position-ing system (Jasper, 1958). Locations (such as P3, P4, sites of the10–20 EEG positions for left and right posterior parietal cortex) canbe determined in relation to landmarks on the scalp such as thevertex (e.g., van Donkelaar and Müri, 2002), which is itself foundby finding the crossing point between the nasion–inion line andthe line connecting the pre-auricular points. Oliveri et al. (2004b)showed that when comparing numerical intervals, normal subjectsoverestimate the difference between the middle number and theouter number positioned at its left side. The right parietal cor-tex plays a major role in such mental space “deformation,” as itstransient disruption by means of rTMS over P4 abolishes the left-ward bias. Similar results were obtained by Göbel et al. (2006) in amental number bisection task when the stimulation site was local-ized both functionally and with reference to individual anatomicalscans.

With the above methods, coil positioning can be related tocertain cortical areas with an estimated resolution in the rangeof centimeters. Although it may be appropriate for some pur-poses, in general this approach accounts for neither interindividualdifferences in the correspondence between scalp landmarks and


F try anm nd con

ufe

dafTao

psitibsbtrttidntIadltwfStts

ig. 4. Schematic illustration of which possible control conditions can be used toultiple options may be implemented in the same design (e.g., sham stimulation a

nderlying brain anatomy nor for interindividual differences in theunctional organization of the brain (Herwig et al., 2003; Okamotot al., 2004).

Coil position relative to individual anatomy can be preciselyefined post hoc by marking the stimulation sites with a locallyttached MR contrast marker (e.g., vitamin E capsules) and per-orming a structural magnetic resonance scan (Terao et al., 1998).he limit of this method is that localization has to be done onlyfterwards and does not allow the selection of the stimulated siten neuroanatomical basis in advance.

Instead of using anatomical information, it is also possible toosition the TMS coil by using a function-guided approach: if thetimulation effect on the performance of a given behavioral tasks known, this task can serve as a “functional” probe to posi-ion the coil for the subsequent investigation of another task. Fornstance, Göbel et al. (2001) applied the “hunting procedure” usedy Ashbridge et al. (1997) for determining coil position on a visualearch task. Thus, to select the region of parietal cortex that woulde stimulated in the main task, the experimenter asked subjectso carry out a visual conjunction search task first, while receivingTMS in the area of the P4 electrode site (right posterior parietal cor-ex). Sites on and around this location were probed for blocks of 20rials until one region was found that consistently and markedlyncreased mean RTs on the search task. The site at which rTMSisrupted visual search was first established for each subject. Theumber task was then carried out while rTMS was delivered onhat same site. The location of the AG/adjacent posterior part of thePS was later confirmed with contrast marker in an MRI scan. Theuthors found increased RTs when rTMS was applied over this siteuring comparison of two-digit numbers to 65. RTs increase fol-

owing left hemisphere stimulation was specific to numbers closeo and larger than 65, whereas a trend for a more generalized effectas observed after right hemisphere stimulation. No effect was

ound when participants were stimulated on either left- or right-MG and it was suggested that the AG/adjacent posterior part ofhe IPS might play a fundamental role in the spatial representa-ion of number magnitude, with the left hemisphere containing apatiotopic representation (i.e., the mental number line).

d ensure that the effects of TMS upon a brain area are indeed specific. Note thattrol sites).

Although influential, the hunting procedure of Ashbridge et al.(1997) is time-consuming and relies on RT differences that donot undergo any statistical tests. An alternative or supplementarylocalizer has been developed by Oliver et al. (2009). The authorsintroduced a procedure for right parietal TMS effects upon visu-ospatial performance, in a task where signal detection theory couldbe applied and sensitivity measures, such as d′, computed. Assess-ment of the TMS intensity required to produce the phenomena ofinterest found this was linearly related to individuals’ rMT overhand M1. In general, the hunting method has limited applicationsbecause it enables the localization of only very specific regions forwhich a TMS localizer is available. In addition, different laboratoriestend to develop their own localizer that better matches their theo-retical assumptions, empirical interests, and previously employedparadigms (see Ro et al., 1999; Ryan et al., 2006).

With the introduction of frameless stereotaxic systems, the TMScoil can be navigated to target specific anatomical sites based onindividual subjects’ structural brain images. These systems provide“on-line” information about the location of the coil (Paus, 1999;Sparing et al., 2010). They achieve a virtual linkage between MRIimages and real anatomy, and allow three-dimensional (3D) orien-tation by interactive visual navigation. Sterotaxic neuronavigationcan be based on subject’s structural MRI (e.g., see Cappelletti et al.,2007; Rusconi et al., 2005). For instance, in Cappelletti et al.’s paper,the horizontal segment of the IPS was defined as middle part of thesulcus in the dorsal parietal lobe that intersects the postcentral sul-cus whereas the AG was defined as the gyrus on the lateral surfaceof the parietal lobe curving around the posterior end of the supe-rior temporal sulcus. However, although this method accounts forinterindividual differences, reliable anatomical landmarks exist forvery few target sites and individual differences of gyral folding andcortical layering have to be taken into account. Another approachexploits functional neuroimaging data (“probabilistic approach”,
Paus et al., 1997). Talairach coordinates could be generated from:(1) a previous fMRI study with the same task paradigm on differentsubjects (Cappelletti et al., 2009; Kansaku et al., 2007); (2) a studyinvestigating similar cognitive processes (Cattaneo et al., 2009b;Dormal et al., 2008; Rusconi et al., 2007); (3) a meta-analysis of sev-

Biobeh

eewacgiutre

divownm(pi(etpsraltlnpawwattdtal

epwsb(McbtarrfTwTtnrw


ral studies (Rusconi et al., 2008, in press). For example, Kansakut al. (2007) used the same counting task in two experiments – oneith fMRI and the other with TMS. fMRI was utilized to localize

n area (or areas) additionally recruited for large vs. small numberounting, and TMS would be applied to disrupt the function of tar-et area(s) to determine whether they play an indispensable rolen the exact large number counting process. It turned out that thepper part of the left ventral premotor cortex (PMv) was preferen-ially activated and its subsequent stimulation confirmed its causalole in the process as it selectively disrupted exact large numbernumeration.

This method considers both structural and functional imagingata relevant to the cognitive task to be used, but it does not take

nto account individual structure–function differences. Finally, aery efficient method consists of identifying target areas basedn individual functional activation maps. For example, to clarifyhether there is a portion of the IPS that is specifically selective forumber or whether the IPS supports all analog magnitude judg-ents (e.g., size, weight, density, brightness), Cohen Kadosh et al.

2007a) combined fMRI and neuronavigated TMS in a size congruityaradigm, a Stroop-like task in which numerical values and phys-

cal sizes were varied independently. In line with previous studiesCohen Kadosh et al., 2007b; Pinel et al., 2004), the size congruityffect activated the bilateral IPS. In order to accurately stimulatehe functionally defined region of interest along the IPS for eacharticipant in the TMS experiment, the authors analyzed the IPSize congruity effect activation for each individual separately. Theesults revealed considerable interindividual differences in the IPSctivation both in extent and in layout and, therefore, in the stimu-ated coordinates. The results showed that TMS to the right (but noto the left) IPS is sufficient to produce a size congruity pattern (i.e.,ack of facilitation), independent of whether subjects compared theumerical value (and ignored the physical size), or compared thehysical size (and ignored the numerical value). Importantly, annalogous behavioral pattern was also obtained when participantsith developmental dyscalculia performed the same task (albeitithout TMS) (Rubinsten and Henik, 2005, 2006). Therefore, the

utomatic processing of numerical value or physical size seemso be equally impaired due to structural abnormality, or stimula-ion, of the right IPS. This approach thus accounts for individualifferences both in brain anatomy and in the functional architec-ure of the brain (see for a comparison of these methods, Sack etl., 2009; Sparing et al., 2008; see Fig. 4 for a flow chart on siteocalization).

The same authors (Sack et al., 2009) investigated whether theffects of right parietal TMS on size congruity revealed in theirrevious study (Cohen Kadosh et al., 2007a) could be obtainedhen using other approaches to determine the target stimulation

ite. They repeated their original study testing the same num-er of participants (n = 5) while basing TMS localization either on:1) individual fMRI-guided TMS neuronavigation; (2) individual

RI-guided TMS neuronavigation; (3) group functional Talairachoordinates; or (4) 10–20 EEG position P4. They quantified theehavioral effects induced by TMS using each approach, calculatedheir standardized effect sizes, and conducted a statistical powernalysis in order to calculate the optimal sample size required toeveal statistical significance. Their results indicate that when theight IPS is stimulated in the location indicated by the individualMRI results, the size congruity effect (SCE) decreases significantly.his effect was found to be significant with a sample of only n = 5hen using fMRI-guided TMS neuronavigation to approach the

MS target site. Although all TMS approaches showed the samerend as the fMRI-guided TMS neuronavigation, the effect size wasot large enough to reach statistical significance. Power analysesevealed the number of necessary participants increased to n = 9hen employing MRI-guided neuronavigation, to n = 13 in case of a

avioral Reviews 35 (2011) 516–536 527

site localization based on group Talairach coordinates, and to n = 47when applying TMS over P4.

It may be that for other brain regions or other cogni-tive tasks, the interindividual differences in brain anatomyand structure–function relationships are less pronounced, thusreducing the difference between fMRI-guided, MRI-guided, andprobabilistic approaches, making these latter two approaches anacceptable compromise between stimulation accuracy and pre-experimental effort. However, there are also data showing aqualitative difference among different approaches. A paper byFeredoes et al. (2007) used fMRI to localize TMS sites to disrupt theshort-term storage of verbal information during a delayed letter-recognition task. Results indicated that rTMS targeting regionsidentified by single-subject analysis (SS, based on fMRI) in left peri-sylvian and sensorimotor cortex impaired performance, whereasrTMS targeting regions identified by the spatially normalizedgroup-averaged (SNGA, based on structural MRI) in left caudal PFChad no effect on performance. This seems to suggest that the brainbasis of some cognitive functions, in which there is considerableinterindividual variability, may be better detected by SS than bySNGA approaches to fMRI data analysis (see Postle and Feredoes,2010).

In conclusion, it seems incontestable that individual fMRI-guided TMS offers the highest experimental power. It howeverlimits the range of possible targets to those that can be identi-fied by fMRI (see Fig. 5 for a schematic summary regarding sitelocalization).

7. Combining TMS with other techniques

7.1. TMS and neuroimaging—off-line and on-line approaches

It is worth mentioning that progress in cognitive neuroscienceoften depends on the convergence of evidence from multiple meth-ods. Since every single technique has its own limitations, there isa clear theoretical advantage in combining different approaches.The timing of TMS relative to neuroimaging (PET, fMRI, EEG)defines which questions can be tackled using a combined TMS-neuroimaging approach. TMS can be applied “off-line” before orafter neuroimaging. We have already discussed the importance ofavailable temporal information (e.g., EEG, MEG) to define the pre-cise time window during which TMS should be applied and thespatial information (e.g., PET or fMRI) that can be used to guidethe positioning of the TMS coil over the cortical target area. Neu-roimaging techniques can also map the spatio-temporal pattern offunctional reorganization induced in the brain by TMS (De Gennaroet al., 2007; Huber et al., 2007; Siebner and Rothwell, 2003) orthe lasting functional impact of TMS on task-related neural activ-ity at the systems level (O’Shea et al., 2007b; Solé-Padullés et al.,2006).

As a rule, neuroimaging should start as quickly as possible afterrTMS to ensure that short-lasting after-effects are captured. Oneexcellent way to detect the effects of rTMS on neural activity is tocompare fMRI task-related activation before and after stimulation.It does pose technical challenges, however.

In TMS-fMRI off-line studies safety procedures and technicalissues are the main constraints on timing. Necessary actions, whichcause a delay between TMS administration and fMRI recording, canbe summarized as follows:

(1) Move subject into the MRI room.(2) Position subject inside the scanner. With a high field scan-

ner (e.g., 4T) positioning takes about 50 s, as vertigo inducedwhen the gradient of the field, the gradient-field product, or thegradient-velocity product experienced by the subject’s head are


aches

(((

Tad

htan2ctOcgshwTtqcPr

Fig. 5. Schematic illustration of appro

high. In lower field (up to 3T) the bed speed can be increasedthus reducing the positioning time to 10–15 s.

3) Let personnel out (about 10 s).4) Localizer (about 50 s).5) Pre-EPI acquisition (distortion correction before each echo-

planar image takes about 45 s, when there is not enough timeit can be done off-line).

herefore, total time before the fMRI experiment can take place isround 4 min. This is an important aspect to keep in mind wheneciding on the duration and type of rTMS protocol.

For instance, a mechanism of rTMS effect on brain activityas been shown supporting the hypothesis that some brain func-ions operate in a state of interhemispheric compensation aftervirtual brain lesion, that is, a form of adaptive plasticity in theondominant hemisphere for functional recovery (O’Shea et al.,007b). These plastic-adapting changes of the intact hemisphereould occur rapidly (>4 min after the end of TMS) and be specifico functions that are normally mediated by the stimulated area.’Shea et al. (2007b) directly demonstrated the interhemisphericompensation by testing the hypothesis that, if the pattern of reor-anization causes recovered behavior, then recovery performancehould breakdown when activity is disrupted in the compensatoryemisphere. Subject underwent two fMRI sessions, one of whichas preceded by 15 min of 1 Hz TMS to dorsal premotor (PMd) area.

he results showed a compensatory increased in fMRI response in
he right PMd and the connected medial premotor areas. Subse-uent TMS of the “reorganized” right PMd disrupted performance,onfirming that the pattern of functional reorganization of the rightMd made a causal contribution in preserving behavior after neu-onal interference.
used for the localization of brain sites.

In the TMS–EEG off-line study, the only point to consider iswhether TMS should be applied with electrodes being attached tothe scalp. The decision might depend on the experimental designas well as number and montage of electrodes. Standard electrodesincrease the distance between the TMS coil and the cortex requir-ing a higher intensity of stimulation to induce a stimulation effect.However, flat electrodes have become available to reduce this prob-lem.

This off-line approach, in which TMS and neuroimaging are sep-arated in time, is technically easier to implement than the “on-line”approach, in which TMS and neuroimaging overlap in time withTMS having the possibility to adversely affect data acquisition dur-ing neuroimaging (for a review about the combination of TMS withneuroimaging see Siebner et al., 2009b). The potential of this on-line approach is the possibility of using TMS to test how focal cortexstimulation acutely modifies the activity and connectivity in thestimulated neural circuits (see for a review Driver et al., 2009). Ifthe goal of the experiment is to investigate the spatial pattern ofTMS-induced changes in the brain activity, an on-line combinationof TMS with PET or fMRI allows one to visualize the causal impactof target neural interference on connected brain structures. Con-versely, if temporal aspects of neuronal processing are the mainfocus, the combination of TMS with EEG provides a methodol-ogy to investigate causal and dynamic relations between brainareas. Finally, it is important to remember that a complementaryapproach to on-line TMS-neuroimaging combinations (TMS–PET,
TMS–fMRI and TMS–EEG) is feasible just using a TMS protocol(dual-site paired pulse TMS), in which behavioral effects of TMSapplied to one cortical area (e.g., hand twitches for M1-TMS, orphosphenes for V1-TMS) can be modulated by preceding ‘condi-tioning’ TMS pulses to a second, potentially interconnected cortical

Biobeh

riwow

buopphatpced

7

lAc

TesTnatsr

ttagd2vflstud

2sitttc(fcaciida


egion. The advantage of this protocol is the possibility to collectnformation about neural processing on a milliseconds timescale,

hile its limit is due to its restriction to pairs of cortical sites with-ut offering the possibility to record neural activity changes in thehole-brain.

An elegant advancement in the application of this protocol haseen demonstrated recently by Davare et al. (2010). The authorssed a triple coil TMS approach, and investigated the consequencesf interfering with cTBS over anterior intraparietal area (AIP) onhysiological interactions between PMv and M1 at rest or duringreparation to grasp objects with either a precision grip or a whole-and grasp. They found that task-related interactions, but not thoset rest, were disrupted following AIP cTBS. Behaviorally, disrup-ion of AIP was also associated with a relative loss of grasp-specificattern of digit muscle activity. These data suggest that AIP is criti-al in processing context- and grasp-dependent information, whichnables PMv to bias excitability levels of M1 hand representationuring preparation for an upcoming grasp.

.2. TMS with concurrent PET or fMRI

The successful combination of TMS with PET was reported in theate 1990s (Paus et al., 1997; Siebner et al., 1998; Fox et al., 1997).t around the same time Bohning et al. (1997) reported the firstombined TMS–fMRI study.

The earliest TMS–PET (see for a review Paus, 2005) andMS–fMRI studies (see for reviews Bestmann et al., 2008a; Rufft al., 2009a) measured the effects of focal brain stimulation whileubjects were at rest, that is, in the absence of any cognitive task.hese data show that the neural consequences of focal TMS wereot restricted to the site of stimulation but also modulated thectivity of distal brain areas. Another important observation washat the impact of stimulation was dose-dependent, as strongertimulation intensities evoked larger activity changes in thoseegions.

Related to this distal connective spread, some studies usedhe combination of TMS and ligand PET to study neurotransmit-er system function, showing that cortical stimulation (prefrontalnd motor cortices) led to an activation of specific cortico-striatallutamatergic pathways, which in turn, induced local release ofopamine at their terminations in the striatum (Strafella et al.,001, 2003). Moreover, a recent study (Cho and Strafella, 2009) pro-ided the first evidence of extrastriatal brain dopamine modulationollowing acute rTMS of the left (but not right) DLPFC with its effectimited to the medial part of prefrontal cortex, including pregenual,ubgenual anterior cingulate cortex and medial orbitofrontal cor-ex. These studies of neurochemical functions with rTMS may beseful for exploring hemispheric differences and the effects of theseifferences on striatal and extrastriatal dopaminergic systems.

Moreover, a recent series of studies (Ruff et al., 2006, 2008,009b; Sack et al., 2007) has moved towards another domain,howing the potential of this approach to investigate interregionalnteractions and their possible functional consequences for percep-ion and cognition. The data from these studies seem to converge onhe following conclusions: (1) focal TMS applied to a particular cor-ical area has both local and remote neural effects in the brain; (2)ortical interactions depend on the functional state of connectionsstate-dependency). For example, Ruff et al. (2008) investigatedunctional interactions between fronto-parietal and visual corti-al areas. The stimulation of right IPS, but not right FEF, elicitspattern of activity changes in visual cortices that depends on

urrent visual context (i.e., affecting the BOLD signal in V1–V4 dur-ng the absence of visual input whereas affecting the BOLD signaln V5/MT only when moving stimuli were present). These state-ependent TMS effects on cortico-cortical interactions have beenlso demonstrated in the motor system (Bestmann et al., 2008b); (3)

avioral Reviews 35 (2011) 516–536 529

the state-dependent remote neural effects of TMS may be function-ally relevant for behavior. For instance, Ruff et al. (2006) reportedthat TMS to the right FEF leads to BOLD increases for peripheralvisual field presentation whereas BOLD decreases for the centralvisual field. Based on these data, the authors predicted that FEFstimulation should enhance peripheral vision, relative to central,in both visual hemifields. This new prediction was confirmed ina psychophysical experiment testing the effects of FEF-TMS uponcontrasts judgment for Gabor gratings presented in the central vs.peripheral visual field. The results showed that perceived contrastwas enhanced for peripheral relative to central stimuli whereasvertex stimulation (control site) was ineffective. Remote neuralconsequences of TMS can thus be used to generate new predictionsfor behavioral TMS effects outside the MR scanner.

If this is true, however, can behavioral TMS studies still be con-sidered crucial to verify the functional necessity of a stimulatedcortical brain region?

Although it seems reasonable to interpret significant remoteneural effect of TMS as also functionally relevant, the causalinvolvement of such remote effects in behavioral changes needsto be demonstrated. In the absence of empirical proofs it is sensibleto keep attributing the primary cause to changes at the site wherestimulation is maximal (Sack, 2010).

Physical modulation of a targeted cortical region, as with TMS,is not the only approach to test for interplay between humanbrain regions. In recent years, functional neuroimaging has movedbeyond conventional brain mapping and is now challenging themonopoly on causality analysis ascribed to functional interferencetechniques in humans (i.e., transcranial direct current stimulationand TMS). For fMRI especially, new techniques are increasinglyapplied providing mathematical models that can be used to assesspossible changes in “effective connectivity (EC)” between brainregions under different conditions (Friston et al., 2003; Roebroecket al., 2005). For example, Granger causality mapping (GCM) is adata-driven EC method with the aim to identify, with reference to aseed region Y, which other brain regions (e.g., X) engage in directedinteractions with region Y (Roebroeck et al., 2005). Therefore, usingspTMS neuronavigation to stimulate fMRI EC-identified clusters atsingle-subject level, it is possible to test behavioral relevance of anfMRI EC functional network underlying a cognitive process (see deGraaf et al., 2009). Moreover, the combination of different methodsfor studying EC can allow new hypotheses to be tested that other-wise would be difficult to address. Finally, since several technicalproblems make such combination very challenging, until now onlyfew groups worldwide have been able to realize simultaneous TMSand fMRI. All these studies used 1.5 or 3T scanner, however sometechnical and safety aspects has been already tested in 4T scanner(Ferrari et al., 2010).

7.3. TMS with concurrent EEG

At around the same time of the first combination of TMS withPET and fMRI, Ilmoniemi et al. (1997) also showed that the com-bination of TMS with EEG was feasible, thus offering an approachto investigate whole-brain causal dynamics with excellent tempo-ral resolution. As other TMS-neuroimaging approaches, the earliestTMS–EEG studies (see for reviews Ilmoniemi and Kicic, 2010;Komssi and Kähkönen, 2006; Miniussi and Thut, 2010; Taylor et al.,2008) measured the effects of focal brain stimulation while sub-jects were at rest. With this approach some studies investigatedthe electrical responses evoked by TMS not only in the stimulated
area (reactivity) but also in the connected sites (connectivity). It hasbeen shown that the reactivity of the cortex is related to TMS inten-sity (Komssi et al., 2004; Kähkönen et al., 2005) and coil orientation(Bonato et al., 2006) and the measures of cortical activity and con-nectivity are state-dependent. For example, Massimini et al. (2005)

5 Biobeh

aTss

stptoar

wwFFntivtaw

taliM

tstt(tsto

8

itvdbaersiuAs–ttntTwt


pplied spTMS over PMd during EEG. The ERPs data showed remoteMS effects in awake subjects that became more restricted duringleep, suggesting that cortical functional connectivity changes withubjects’ level of arousal.

In recent studies, however, the efforts to combine TMS withimultaneous ERP recording have been successful in studying theemporal and functional impact of TMS interference on cognitiverocesses. The value of this combination lies in the fact that the fineemporal resolution of EEG allows one to make an on-line measuref the effects of TMS at different stages of processing (e.g., sensorynd post-perceptual), within brain regions which are anatomicallyemote from the area impaired by TMS.

Therefore, by using this approach, it is possible to investigatehere, when and how TMS interacts with task performance on ahole-brain scale (Fuggetta et al., 2006, 2009; Taylor et al., 2007).

or instance, Taylor et al. (2007) tested whether the stimulation ofEFs had a causal impact on distal visual cortex during the endoge-ous orienting of visuospatial attention. When TMS was deliveredo the right FEFs, immediately after presentation of an arrow cuendicating where to allocate attention (i.e., prior to the onset of theisual stimulus), induced a change in the baseline EEG signal and inhe event-related cortical response recorded over posterior visualreas. The TMS-induced effect occurred at the same time as ERPsere shown to be modulated by visuospatial attention.

Finally, another possible approach is the use of TMS withinhe broader perspective on brain rhythms that EEG provides. Thispproach can be used to investigate how rhythmic brain stimu-ation can be used to modify brain functions as well as how TMSnteracts with oscillatory brain activity (see for a review Thut and

iniussi, 2009).One interesting characteristic that should not go unnoticed is

hat the result of these “on-line” approaches differs from the simpleum of the two single techniques, and researchers should thereforeake extra care in designing experiments (i.e., appropriate controlask, sham or control site, intensity of stimulation), collecting datai.e., use of neuronavigation system, coil orientation, adopt carefuliming of interleaved TMS stimulus and echo-planar images acqui-ition), and interpreting data (i.e., to take into account nonspecificask effects such as auditory and somatosensory responses, choicef artifact correction strategy during data analysis).

. State-dependent TMS paradigms

As mentioned earlier, the neural impact of an external stimuluss determined not only by the properties of that stimulus but also onhe initial state of the activated brain region. In accordance with thisiew of state-dependent effects of TMS, new methods have beenevised. The first one, called TMS-adaptation paradigm (TMSA) isased on the hypothesis that TMS acts differentially on neuronsccording to their initial neural activation state, with attributesncoded by the adapted neural populations within the stimulatedegion being more susceptible to the effects of TMS. An adaptingtimulus, presented for a long time (usually 40–60 s), is used tonduce habituation in a subset of cells that code particular stim-lus features, therefore making them a selective target for TMS.ccording to this paradigm, if TMS delivered over a cortical area –upposedly containing neurons that code for the adapting stimulusimproves behavioral performance for these stimuli, this indicates

hat neurons in that area were indeed adapted by and tuned tohe adapting stimulus. Thus, by using adaptation to manipulate
eural activation states prior to the application of TMS, one can con-rol which neural populations are preferentially facilitated by TMS.he validity of this TMSA paradigm has been confirmed in studiesith stimulation of visual areas such as V1/V2 and V5/MT in which
he receptive properties are relatively well known (Silvanto and


Muggleton, 2008). Moreover, recent studies confirmed the validityof this paradigm in the language (Cattaneo et al., 2009a, 2009c),number (Cohen Kadosh et al., in press) and motor acts observationdomains (Cattaneo et al., in press). With an elegant design, Cattaneoet al. (in press) were able to exploit the specificity of this paradigmto investigate the neural representation of observed motor behav-ior in the inferior parietal lobule (IPL), PMv and in the cortex aroundthe superior temporal sulcus (STS). Participants were shown adapt-ing movies of a hand or a foot acting on different objects and wereasked to compare to the movie a motor act shown in test pictures.The invariant features between adapting and test stimuli fitted a2 × 2 design: same or different action made by the same or differ-ent effector. TMS over the left and right PMv and over the left IPLinduced a selective shortening of RTs to stimuli showing a repeated(adapted) action, regardless of the effector performing it. In a sec-ond experiment TMS applied over the left STS induced shorteningof RTs for adapted actions but only if also the effector was repeated.In conclusion, these results indicate that observed motor behavioris encoded unitarily with the body part that performs it in the tem-poral lobe. A hierarchically higher level of representation is carriedby neural populations in the parieto-frontal regions, where acts areencoded as their meaning.

Another psychophysical method for manipulating the initialactivation state is priming. Preliminary data suggest that when TMSis applied after priming, TMS facilitates target detection when thetarget has not been primed, whereas performance is unaffected ifthe target has been primed (Campana et al., 2002; Cattaneo et al.,2008, 2010). For instance, Cattaneo et al. (2010) investigated therole of left PMd and PMv in semantic processing. Priming to a cat-egory name (either “Tool” or “Animal”) was used to modulate theinitial activation state of these brain areas prior to application ofTMS and the presentation of the target stimulus. When the tar-get word was an exemplar of the “Tool” category, the effects ofTMS over PMv (but not PMd) interacted with priming history byfacilitating RTs on incongruent trials but not on congruent trials.This TMS by congruency interaction suggests that the “Tool” and“Animal” primes had a distinct effect on the initial activation stateof the left PMv and implies that this cerebral region is one neurallocus of category-specific behavioral priming for the “Tool” cate-gory. TMS delivered over PMv had no behavioral effect when thetarget stimulus was an exemplar of the “Animal” category, regard-less of whether the target word was congruent or incongruent withthe prime. In conclusion, results demonstrate the causal role of cat-egory specificity in the left PMv in the encoding of words describinggraspable tools.

A differential TMS effect when the target stimulus is primed vs.unprimed (or adapted vs. nonadapted) can reveal neural selectivitybecause it shows that the stimulated brain area contains a neuralrepresentation that was affected by the initial state manipulation.

Although the fact that preconditioning modulates the effects ofsubsequent TMS (see for a review Silvanto et al., 2008) and that alot of state-dependency ideas in the form of adaptive coding (seeDuncan, 2001) have been known for a long time, we believe thatthese paradigms are an important and useful step forward for thestudy of higher level cognitive processes.

These new methods could reveal something that ‘virtual lesion’TMS would not (naturally considering only situations where theycan be used to ask the same question). Imagine the case where aregion contains functionally overlapping populations of neuronsthat are selectively crucial for function X vs. Y. Subjects receivea task that measures both X and Y. Normal TMS reveals a deficit
of both tasks, which does not tell us whether it is a single popu-lation of neurons that do both functions, or intermingled specificpopulations. But by preconditioning subjects to X or Y (and thenshowing that the TMS effect differs for the two functions), somedegree of specificity is revealed in a way that normal TMS cannot

Biobeh

aoela

usss

9

etstal

ttphopndwdsmcTsl

istaplso

bAtseaIcviiolaisms


ccess. Therefore, in these new paradigms the functional resolutionf TMS is increased, allowing differential interventions on distinct,ven spatially overlapping, neural populations within the stimu-ated brain region (see for reviews Silvanto et al., 2008; Silvantond Pascual-Leone, 2008).

Finally, we do not believe that these new paradigms will be lessseful to cognitive neurosciences because the physiological basis oftate-dependent effects are unknown (Siebner et al., 2009a), as theame problem applies to other more common paradigms (includingtandard ‘virtual lesion’ TMS, fMRI adaptation, etc.).

. Safety guidelines

Methodological issues and technological advances cannot bentirely separated from safety and ethical considerations; we willhus briefly summarize and discuss the implications of the newestafety guidelines (Rossi et al., 2009). The new guidelines are meanto update the ones (Wassermann, 1998) which have regulated TMSpplication during its first exciting decade of experimentation inaboratories all over the world.

The last decade has seen a rapid increase in the use of TMSo study cognitive functions, brain–behavior relation and in thereatment for a variety of psychiatric and neurological disorders. Inarticular, new protocols have been introduced, changes in devicesave been implemented, TMS is being increasingly combined withther neuroimaging techniques and many normal subjects andatients are being investigated in longer stimulation sessions. Theecessity to revisit the safety guidelines, update the recommen-ations of practice and improve the discussion of ethical aspectsas the motivation behind the consensus conference in Certosai Pontignano (Italy) on March 7–9, 2008. As in the NIH consen-us conference (June 5–7, 1996 see Wassermann, 1998), the 2008eeting brought together some of the leading scientists who are

urrently using TMS in basic research and for clinical applications.he 2008 consensus conference led to the development of newafety guidelines, which build, whenever possible, on the old guide-ines (Wassermann, 1998).

Because the aim of our review is about research experimentsn normal subjects, we will only summarize the safety guidelinestrictly related to studies regarding cognitive functions. Accordingo Rossi et al.’s (2009) classification, these studies belong to class 3,s they provide only indirect benefits for the subject (no immediateersonal relevance) and the risk to which the subject is exposed is

ow. Given the inevitable presence of (a minimal) risk, these studieshould be carried out when they can be expected to yield data ofutstanding scientific or clinical value on brain physiology or safety.

In general, risk is defined as the product of threats, vulnera-ilities and possible consequences implied by an action (see e.g.,ven, 2008). In the TMS environment, the most severe consequence

hat may be caused by stimulation on healthy participants is acuteeizure induction. Based on the 9 new cases reported in the lit-rature after the publication of Wassermann’s (1998) guidelines,cute seizure during TMS experiments is an extremely rare event.n addition, a few of the reported episodes are dubious, and theirlinical description matches with syncope more than a seizure (seeideo available at www.brainstimjrn.com/content/mmc library). Its, however, clear that pro-epileptogenic drugs (in patient stud-es) and the use of TMS parameters exceeding safety guidelinesr experimenting with new TMS protocols for which no guide-ines are available (in patient and healthy subjects) increase threats
nd/or individual vulnerability, and therefore amplify risk. Follow-ng these considerations and years of research experience fromeveral leading laboratories in the world, Rossi et al. (2009) esti-ate a probability of less than 0.01 of inducing a seizure in healthy
ubjects when using established high-frequency rTMS parameters.

avioral Reviews 35 (2011) 516–536 531

The wide hands-on experience of the consensus group guaranteesthat such estimate takes into account the possible hidden figureof cases that did not make it into the relevant literature. Consid-erations regarding the following main points should be consultedwhen designing a TMS or an rTMS study – they are all exhaustivelydiscussed in section 7 of Rossi et al.’s (2009) guidelines:

(1) Ethical and regulatory issues.Research application must be governed by three fundamen-

tal ethical and legal requirements: informed consent, wherebyparticipants receive all the relevant information about potentialrisks and discomfort with a terminology they can understand,and decide voluntarily to take part to the study; risk-benefitratio, which has to be evaluated beforehand whether rele-vant data can be obtained and the scientific question can beanswered without exposing any subjects to risks; equal distri-bution of the burdens and benefits of research, which preventsto conduct research on categories of individual made particu-larly vulnerable by economic, social or physical conditions.

(2) Stimulation parameters.Conditions of increased or uncertain risk of inducing epileptic

seizures are:Any “novel paradigm” (i.e., that is not a conventional method

of high–low-frequency rTMS performed with a flat butterfly coiland biphasic stimulation) or TMS applied on more than a singlebrain region.

Conventional high-frequency protocol with parameters(intensity, frequency, train length or ITI) exceeding the safetylimits reported in Rossi et al.’s Table 4; to prevent seizure occur-rence when experimenting with new protocols and parameters,some precautions should be taken, such as continuous EMGmonitoring, direct visual/video monitoring of the subjects bya qualified individual, and long-term neuropsychological mon-itoring.

(3) Setting, TMS team and management of potential adverseeffects.

Class 3 studies that are conducted with known protocols andwith parameters respecting the guidelines can be conductedin non-medical setting by expert personnel (not necessarily alicensed physician). Appropriate procedures should be planned,however, in case of medical emergency; the personnel shouldbe adequately trained and instructed by a licensed physician onhow to react in such emergencies (see Rossi et al., 2009, Section7.2.6).

Finally, researchers combining TMS–fMRI studies have to knowthat if TMS is applied in the MR scanner, there are potential safetyconcerns which are related to the static magnetic field of the MRscanner, the radio frequency pulses and gradients applied duringscanning, and the mechanic interaction between the TMS systemand static magnetic field of the scanner (see Bestmann et al., 2008a;Siebner et al., 2009b). It will therefore be necessary to screen par-ticipants for additional exclusion factors (see Section 7.2.8 in Rossiet al., 2009).

In conclusion, this overview highlights the importance forexperimenters to carefully screen their participants for potentialexclusion factors and carefully consult the new safety guidelineswhen planning and designing a future TMS or rTMS study.

10. Conclusions

In this review we summarized the main technical and method-ological considerations needed for the use of TMS in cognitiveneuroscience, based on the corpus of studies and technicalimprovements that have become available in most recent years. All

http://www.brainstimjrn.com/content/mmc_library

5 Biobeh

otivpchscalitladttbsnpcnr

dmicbnpwm2sga

A

tPfm

R

A

A

A

A

A

A

A


f these aspects exemplified nicely how TMS can be used to inves-igate a variety of issues in the cognitive neurosciences in general,ncluding location, timing, lateralization and the functional rele-ance of the neuronal correlates underlying cognitive functions. Werovide crucial information for experimenters who want to choosearefully the most appropriate experimental design to test theirypotheses within a given theoretical framework. Moreover, wehowed how the use of more than one control condition is cru-ial for determining the contribution made by a cortical area tospecific behavior. We also discussed the possible approaches to

ocate “behaviorally silent” areas and the complexity of the effectsnduced by TMS stimulation on these brain regions. This is dueo the fact that, for this domain, the way in which TMS modu-ates behavior (facilitatory or disruptive effects), depends upon

number of variables such as stimulation parameters (intensity,uration, frequency and pattern of pulses applied), site of stimula-ion and type of task. Moreover, we should bear in mind that thathe neural impact of an external stimulus is not determined onlyy the properties of that stimulus but also on the initial activationtate of the brain. Based on the state-dependent effects of TMS,ew paradigms have been proposed (TMS-adaptation and TMS-riming) differing from the standard “virtual lesion” TMS whichan reveal the necessity of a cortical region but cannot discrimi-ate between functionally distinct neuronal representations in thategion.

In addition, the introduction of a causal (i.e., interventional)imension into noninvasive neuroimaging, shows that the effectsay be related to the direct change of activity in the areas

mmediately underlying the stimulation site, or upon remote inter-onnected regions. This opens up a new field for research on therain basis of cognition, in terms of interregional influences withinetworks. Finally, the increasing number of applications for newrotocols (e.g., theta burst), the concurrent combination of TMSith other neuroimaging techniques and the investigation of nor-al subjects and patients (see for a reviews Dimyan and Cohen,

010; Ridding and Rothwell, 2007; Miniussi et al., 2008) with longertimulation sessions led to the development of the new safetyuidelines which must be consulted when planning and designingfuture TMS or rTMS study.

cknowledgments

We would like to thank Leonardo Cohen and Chris Chambers forheir insightful comments on a previous version of the manuscript,aolo Ferrari for discussions about the combination of TMS withMRI, Manuela Ruzzoli for useful suggestions about the neural

echanisms of TMS and two anonymous reviewers.

eferences

llen, E.A., Pasley, B.N., Duong, T., Freeman, R.D., 2007. Transcranial magnetic stim-ulation elicits coupled neural and hemodynamic consequences. Science 317,1918–1921.

massian, V.E., Cracco, R.Q., Maccabee, P.J., Cracco, J.B., Rudell, A., Eberle, L., 1989.Suppression of visual perception by magnetic coil stimulation of human occipitalcortex. Electroencephalogr. Clin. Neurophysiol. 74, 458–462.

ndoh, J., Artiges, E., Pallier, C., Rivière, D., Mangin, J.F., Cachia, A., Plaze, M., Paillère-Martinot, M.L., Martinot, J.L., 2006. Modulation of language areas with functionalMR image-guided magnetic stimulation. Neuroimage 29 (2), 619–627.

ndoh, J., Artiges, E., Pallier, C., Rivière, D., Mangin, J.F., Paillère-Martinot, M.L., Mar-tinot, J.L., 2008. Priming frequencies of transcranial magnetic stimulation overWernicke’s area modulate word detection. Cereb. Cortex 18 (1), 210–216.

ndres, M., Seron, X., Olivier, E., 2005. Hemispheric lateralization of number com-
parison. Cogn. Brain Res. 25 (1), 283–290.
ndres, M., Seron, X., Olivier, E., 2007. Contribution of hand motor circuits to count-ing. J. Cogn. Neurosci. 19 (4), 563–576.

ntal, A., Nitsche, M.A., Kincses, T.Z., Lampe, C., Paulus, W., 2003. No correlationbetween moving phosphene and motor thresholds: a transcranial magneticstimulation study. Neuroreport 15, 297–302.


Ashbridge, E., Walsh, V., Cowey, A., 1997. Temporal aspects of visual search studiedby transcranial magnetic stimulation. Neuropsychologia 35 (8), 1121–1131.

Aven, T., 2008. Risk Analysis. Assessing Uncertainties Beyond Expected Values andProbabilities. John Wiley & Sons, Chichester.

Barker, A.T., Jalinous, R., Freeston, I.L., 1985. Non-invasive magnetic stimulation ofthe human motor cortex. Lancet 1 (8437), 1106–1107.

Baumer, T., Bock, F., Koch, G., Lange, R., Rothwell, J.C., Siebner, H.R., Munchau, A.,2006. Magnetic stimulation of human premotor or motor cortex produces inter-hemispheric facilitation through distinct pathways. J. Physiol. 572, 857–868.

Beckers, G., Hömberg, V., 1992. Cerebral visual motion blindness: transitory akine-topsia induced by transcranial magnetic stimulation of human area V5. Proc.Biol. Sci. 249 (1325), 173–178.

Bestmann, S., Ruff, C.C., Blankenburg, F., Weiskopf, N., Driver, J., Rothwell, J.C., 2008a.Mapping causal interregional influences with concurrent TMS-fMRI. Exp. BrainRes. 191 (4), 383–402.

Bestmann, S., Swayne, O., Blankenburg, F., Ruff, C.C., Haggard, P., Weiskopf, N.,Josephs, O., Driver, J., Rothwell, J.C., Ward, N.S., 2008b. Dorsal premotor cortexexerts state-dependent causal influences on activity in contralateral primarymotor and dorsal premotor cortex. Cereb. Cortex 18 (6), 1281–1291.

Bohning, D.E., Pecheny, A.P., Epstein, C.M., Speer, A.M., Vincent, D.J., Dannels, W.,George, M.S., 1997. Mapping transcranial magnetic stimulation (TMS) fields invivo with MRI. Neuroreport 8, 2535–2538.

Bonato, C., Miniussi, C., Rossini, P.M., 2006. Transcranial magnetic stimulation andcortical evoked potentials: a TMS/EEG co-registration study. Clin. Neurophysiol.117 (8), 1699–1707.

Boorman, E.D., O’Shea, J., Sebastian, C., Rushworth, M.F., Johansen-Berg, H., 2007.Individual differences in white-matter microstructure reflect variation in func-tional connectivity during choice. Curr. Biol. 17 (16), 1426–1431.

Boroojerdi, B., Meister, I.G., Foltys, H., Sparing, R., Cohen, L.G., Töpper, R., 2002. Visualand motor cortex excitability: a transcranial magnetic stimulation study. Clin.Neurophysiol. 113, 1501–1504.

Boroojerdi, B., Phipps, M., Kopylev, L., Wharton, C.M., Cohen, L.G., Grafman, J., 2001.Enhancing analogic reasoning with rTMS over the left prefrontal cortex. Neurol-ogy 56 (4), 526–528.

Boroojerdi, B., Prager, A., Muellbacher, W., Cohen, L.G., 2000. Reduction of humanvisual cortex excitability using 1-Hz transcranial magnetic stimulation. Neurol-ogy 54, 1529–1531.

Brasil-Neto, J.P., Cohen, L.G., Panizza, M., Nilsson, J., Roth, B.J., Hallett, M., 1992a. Opti-mal focal transcranial magnetic activation of the human motor cortex: effects ofcoil orientation, shape of induced current pulse, and stimulus intensity. J. Clin.Neurophysiol. 9, 132–136.

Brasil-Neto, J.P., McShane, L.M., Fuhr, P., Hallet, M., Cohen, L.G., 1992b. Topographicmapping of the human motor cortex with magnetic stimulation: factors affect-ing accuracy and reproducibility. Electroencephalogr. Clin. Neurophysiol. 85,9–16.

Brighina, F., Piazza, A., Daniele, O., Fierro, B., 2002. Modulation of visual corticalexcitability in migraine with aura: effects of 1 Hz repetitive transcranial mag-netic stimulation. Exp. Brain Res. 145 (2), 177–181.

Buccino, G., Riggio, L., Melli, G., Binkofski, F., Gallese, V., Rizzolatti, G., 2005. Listen-ing to action-related sentences modulates the activity of the motor system: acombined TMS and behavioral study. Cogn. Brain Res. 24 (3), 355–363.

Buch, E.R., Mars, R.B., Boorman, E.D., Rushworth, M.F., 2010. A network centeredon ventral premotor cortex exerts both facilitatory and inhibitory control overprimary motor cortex during action reprogramming. J. Neurosci. 30 (4), 1395–1401.

Campana, G., Cowey, A., Walsh, V., 2002. Priming of motion direction and areaV5/MT: a test of perceptual memory. Cereb. Cortex 12 (6), 663–669.

Cappa, S.F., Sandrini, M., Rossini, P.M., Sosta, K., Miniussi, C., 2002. The role of theleft frontal lobe in action naming: rTMS evidence. Neurology 59, 720–723.

Cappelletti, M., Barth, H., Fregni, F., Spelke, E.S., Pascual-Leone, A., 2007. rTMS overthe intraparietal sulcus disrupts numerosity processing. Exp. Brain Res. 179 (4),631–642.

Cappelletti, M., Muggleton, N., Walsh, V., 2009. Quantity without numbers and num-bers without quantity in the parietal cortex. Neuroimage 46 (2), 522–529.

Castelli, F., Glaser, D.E., Butterworth, B., 2006. Discrete and analogue quantity pro-cessing in the parietal lobe: a functional MRI study. Proc. Natl. Acad. Sci. U.S.A.103, 4693–4698.

Catmur, C., Walsh, V., Heyes, C., 2009. Associative sequence learning: the role ofexperience in the development of imitation and the mirror system. Philos. Trans.R. Soc. Lond. B Biol. Sci. 364 (1528), 2369–2380.

Cattaneo, L., Sandrini, M., Schwarzbach, J., in press. State-dependent TMS revealsa hierarchical representation of observed acts in the temporal, parietal andpremotor cortices. Cereb. Cortex doi:10.1093/cercor/bhp291.

Cattaneo, Z., Devlin, J.T., Salvini, F., Vecchi, T., Silvanto, J., 2010. The causal role ofcategory-specific neuronal representations in the left ventral premotor cortex(PMv) in semantic processing. Neuroimage 49 (3), 2728–2734.

Cattaneo, Z., Devlin, J.T., Vecchi, T., Silvanto, J., 2009a. Dissociable neural represen-tations of grammatical gender in Broca’s area investigated by the combinationof satiation and TMS. Neuroimage 47 (2), 700–704.

Cattaneo, Z., Rota, F., Vecchi, T., Silvanto, J., 2008. Using state-dependency of tran-
scranial magnetic stimulation (TMS) to investigate letter selectivity in the leftposterior parietal cortex: a comparison of TMS-priming and TMS-adaptationparadigms. Eur. J. Neurosci. 28 (9), 1924–1929.
Cattaneo, Z., Rota, F., Walsh, V., Vecchi, T., Silvanto, J., 2009c. TMS-Adaptation revealsabstract letter selectivity in the left posterior parietal cortex. Cereb. Cortex 19(10), 2321–2325.

http://dx.doi.org/10.1093/cercor/bhp291

Biobeh

C

C

C

C

C

C

C

C

C

C

C

d

D

D

D

D

D

D

d

D

d

D

D

D

D

D

D

D

E


attaneo, Z., Silvanto, J., Battelli, L., Pascual-Leone, A., 2009d. The mental numberline modulates visual cortical excitability. Neurosci. Lett. 462 (3), 253–256.

attaneo, Z., Silvanto, J., Pascual-Leone, A., Battelli, L., 2009b. The role of the angulargyrus in the modulation of visuospatial attention by the mental number line.Neuroimage 44 (2), 563–568.

elnik, P., Stefan, K., Hummel, F., Duque, J., Classen, J., Cohen, L.G., 2006. Encodinga motor memory in the older adult by action observation. Neuroimage 29 (2),677–684.

hambers, C.D., Payne, J.M., Stokes, M.G., Mattingley, J.B., 2004. Fast and slow parietalpathways mediate spatial attention. Nat. Neurosci. 7 (3), 217–218.

heeran, B., Koch, G., Stagg, C.J., Baig, F., Teo, J., 2010. Transcranial magnetic stimula-tion: from neurophysiology to pharmacology, molecular biology and genomics.The Neuroscientist 16 (3), 210–221.

hen, R., Classen, J., Gerloff, C., Celnik, P., Wassermann, E.M., Hallett, M., Cohen,L.G., 1997. Depression of motor cortex excitability by low-frequency transcranialmagnetic stimulation. Neurology 48, 1398–1403.

ho, S.S., Pellecchia, G., van Eimeren, T., Cilia, R., Strafella, A.P., in press. Continuoustheta burst stimulation of right dorsolateral prefrontal cortex induces changesin impulsivity level. Brain Stimul. doi:10.1016/j.brs.2009.10.002.

ho, S.S., Strafella, A.P., 2009. rTMS of the left dorsolateral prefrontal cortexmodulates dopamine release in the ipsilateral anterior cingulate cortex andorbitofrontal cortex. PLoS One 4 (8), e6725.

ohen Kadosh, R., Cohen Kadosh, K., Linden, D.E.J., Gevers, W., Berger, A., Henik, A.,2007b. The brain locus of interaction between number and size: a combinedfunctional magnetic resonance imaging and event-related potential study. J.Cogn. Neurosci. 19, 957–970.

ohen Kadosh, R., Cohen Kadosh, K., Schuhmann, T., Kaas, A., Goebel, R., Henik,A., Sack, A.T., 2007a. Virtual dyscalculia induced by parietal-lobe TMS impairsautomatic magnitude processing. Curr. Biol. 17 (8), 689–693.

ohen Kadosh, R., Muggleton, N., Silvanto, J., Walsh, V., in press. Double dissociationof format-dependent and number-specific neurons in human parietal cortex.Cereb. Cortex doi:10.1093/cercor/bhp273.

’Alfonso, A.A., van Honk, J., Hermans, E., Postma, A., de Haan, E.H., 2000. Lateral-ity effects in selective attention to threat after repetitive transcranial magneticstimulation at the prefrontal cortex in female subjects. Neurosci. Lett. 280,195–198.

avare, M., Lemon, R., Olivier, E., 2008. Selective modulation of interactions betweenventral premotor cortex and primary motor cortex during precision grasping inhumans. J. Physiol. 586, 2735–2742.

avare, M., Montague, K., Olivier, E., Rothwell, J.C., Lemon, R.N., 2009. Ventral premo-tor to primary motor cortical interactions during object-driven grasp in humans.Cortex 45 (9), 1050–1057.

avare, M., Rothwell, J.C., Lemon, R.N., 2010. Causal connectivity between the humananterior intraparietal area and premotor cortex during grasp. Curr. Biol. 20 (2),176–181.

eblieck, C., Thompson, B., Iacoboni, M., Wu, A.D., 2008. Correlation between motorand phosphene thresholds: a transcranial magnetic stimulation study. Hum.Brain Mapp. 29 (6), 662–670.

e Gennaro, L., Bertini, M., Pauri, F., Cristiani, R., Curcio, G., Ferrara, M., Rossini, P.M.,2004. Callosal effects of transcranial magnetic stimulation (TMS): the influenceof gender and stimulus parameters. Neurosci. Res. 48 (2), 129–137.

e Gennaro, L., Marzano, C., Veniero, D., Moroni, F., Fratello, F., Curcio, G., Fer-rara, M., Ferlazzo, F., Novella, L., Pellicciari, M.C., Bertini, M., Rossini, P.M., 2007.Neurophysiological correlates of sleepiness: a combined TMS and EEG study.Neuroimage 36 (4), 1277–1287.

e Graaf, T.A., Jacobs, C., Roebroeck, A., Sack, A.T., 2009. FMRI effective connectivityand TMS chronometry: complementary accounts of causality in the visuospatialjudgment network. PLoS One 4 (12), e8307.

ehaene, S., 1996. The organization of brain activations in number comparison:event-related potentials and the additive-factors methods. J. Cogn. Neurosci. 8,47–68.

e Labra, C., Rivadulla, C., Grieve, K., Marino, J., Espinosa, N., Cudeiro, J., 2007. Changesin visual responses in the feline dLGN: selective thalamic suppression inducedby transcranial magnetic stimulation of V1. Cereb. Cortex 17, 1376–1385.

i Lazzaro, V., Oliviero, A., Profice, P., Meglio, M., Cioni, B., Tonali, P., Rothwell, J.C.,2001. Descending spinal cord volleys evoked by transcranial magnetic and elec-tric stimulation of the motor cortex leg area in conscious humans. J. Physiol. 537,1047–1058.

i Lazzaro, V., Ziemann, U., Lemon, R.N., 2008. State of the art: physiology of tran-scranial motor cortex stimulation. Brain Stimul. 1, 345–362.

imyan, M.A., Cohen, L.G., 2010. Contribution of transcranial magnetic stimulationto the understanding of functional recovery mechanisms after stroke. Neurore-habil. Neural Repair 24 (2), 125–135.

ormal, V., Andres, M., Pesenti, M., 2008. Dissociation of numerosity and durationprocessing in the left intraparietal sulcus: a transcranial magnetic stimulationstudy. Cortex 44 (4), 462–469.

räger, B., Breitenstein, C., Helmke, U., Kamping, S., Knecht, S., 2004. Specific andnonspecific effects of transcranial magnetic stimulation on picture-word verifi-cation. Eur. J. Neurosci. 20 (6), 1681–1687.

river, J., Blankenburg, F., Bestmann, S., Vanduffel, W., Ruff, C.C., 2009. Concurrent
brain-stimulation and neuroimaging for studies of cognition. Trends Cogn. Sci.13 (7), 319–327.
uncan, J., 2001. An adaptive coding model of neural function in prefrontal cortex.Nat. Rev. Neurosci. 2, 820–829.

isenegger, C., Treyer, V., Fehr, E., Knoch, D., 2008. Time-course of “off-line” pre-frontal rTMS effects—a PET study. Neuroimage 42 (1), 379–384.

avioral Reviews 35 (2011) 516–536 533

Ellison, A., Lane, A.R., Schenk, T., 2007. The interaction of brain regions during visualsearch processing as revealed by transcranial magnetic stimulation. Cereb. Cor-tex 17 (11), 2579–2584.

Fadiga, L., Craighero, L., Buccino, G., Rizzolatti, G., 2002. Speech listening specificallymodulates the excitability of tongue muscles: a TMS study. Eur. J. Neurosci. 15(2), 399–402.

Fadini, T., Matthäus, L., Rothkegel, H., Sommer, M., Tergau, F., Schweikard, A.,Paulus, W., Nitsche, M.A., 2009. H-coil: induced electric field properties andinput/output curves on healthy volunteers, comparison with a standard figure-of-eight coil. Clin. Neurophysiol. 120 (6), 1174–1182.

Ferbert, A., Priori, A., Rothwell, J.C., Day, B.L., Colebatch, J.G., Marsden, C.D., 1992.Interhemispheric inhibition of the human motor cortex. J. Physiol. 453, 525–546.

Feredoes, E., Tononi, G., Postle, B.R., 2007. The neural bases of the short-term storageof verbal information are anatomically variable across individuals. J. Neurosci.27 (41), 11003–11008.

Ferrari, P., Cattaneo, L., Schwarzbach, J.V., Sandrini, M., Jovicich, J., 2010. Technicaland Safety Aspects in Concurrent TMS-fMRI. International Society of MagneticResonance in Medicine, Stockholm, Sweden.

Fias, W., Fischer, M.H., 2005. Spatial representation of numbers. In: Campbell, J.I.D.(Ed.), Handbook of Mathematical Cognition. Psychology Press, Hove, pp. 43–54.

Fitzgerald, P.B., Fountain, S., Daskalakis, Z., 2006. A comprehensive review of theeffects of rTMS on motor cortex excitability and inhibition. Clin. Neurophysiol.117, 2584–2596.

Fox, P., Ingham, R., George, M.S., Mayberg, H., Ingham, J., Roby, J., Martin, C., Jer-abek, P., 1997. Imaging human intra-cerebral connectivity by PET during TMS.Neuroreport 8 (12), 2787–2791.

Friston, K.J., Harrison, L., Penny, W., 2003. Dynamic causal modelling. Neuroimage19, 1273–1302.

Fuggetta, G., Pavone, E.F., Walsh, V., Kiss, M., Eimer, M., 2006. Cortico-cortical inter-actions in spatial attention: a combined ERP/TMS study. J. Neurophysiol. 95 (5),3277–3280.

Fuggetta, G., Rizzo, S., Pobric, G., Lavidor, M., Walsh, V., 2009. Functional repre-sentation of living and nonliving domains across the cerebral hemispheres: acombined event-related potential/transcranial magnetic stimulation study. J.Cogn. Neurosci. 21 (2), 403–414.

Galea, J.M., Albert, N.B., Ditye, T., Miall, R.C., 2010. Disruption of the dorsolateral pre-frontal cortex facilitates the consolidation of procedural skills. J. Cogn. Neurosci.22 (6), 1158–1164.

Gerwig, M., Kastrup, O., Meyer, B.U., Niehaus, L., 2003. Evaluation of corticalexcitability by motor and phosphene thresholds in transcranial magnetic stim-ulation. J. Neurol. Sci. 215, 75–78.

Göbel, S.M., Calabria, M., Farnè, A., Rossetti, Y., 2006. Parietal rTMS distorts themental number line: simulating ‘spatial’ neglect in healthy subjects. Neuropsy-chologia 44 (6), 860–868.

Göbel, S., Walsh, V., Rushworth, M.F., 2001. The mental number line and the humanangular gyrus. Neuroimage 14 (6), 1278–1289.

Grosbras, M.H., Paus, T., 2003. Transcranial magnetic stimulation of the humanfrontal eye field facilitates visual awareness. Eur. J. Neurosci. 18 (11), 3121–3126.

Hamidi, M., Tononi, G., Postle, B.R., 2008. Evaluating frontal and parietal con-tributions to spatial working memory with repetitive transcranial magneticstimulation. Brain Res. 1230, 202–210.

Harris, J.A., Clifford, C.W., Miniussi, C., 2008. The functional effect of transcranialmagnetic stimulation: signal suppression or neural noise generation? J. Cogn.Neurosci. 20 (4), 734–740.

Hershenson, M., 1962. Reaction time as a measure of intersensory facilitation. J. Exp.Psychol. 63, 289–293.

Herwig, U., Satrapi, P., Schönfeldt-Lecuona, C., 2003. Using the international 10–20EEG system for positioning of transcranial magnetic stimulation. Brain Topogr.16 (2), 95–99.

Hilgetag, C.C., Théoret, H., Pascual-Leone, A., 2001. Enhanced visual spatial atten-tion ipsilateral to rTMS-induced ‘virtual lesions’ f human parietal cortex. Nat.Neurosci. 4 (9), 953–957.

Huang, Y.Z., Edwards, M.J., Rounis, E., Bhatia, K.P., Rothwell, J.C., 2005. Theta burststimulation of the human motor cortex. Neuron 45 (2), 201–206.

Hubbard, E.M., Piazza, M., Pinel, P., Dehaene, S., 2005. Interactions between numberand space in the parietal cortex. Nat. Rev. Neurosci. 6, 435–448.

Huber, R., Esser, S.K., Ferrarelli, F., Massimini, M., Peterson, M.J., Tononi, G., 2007.TMS-induced cortical potentiation during wakefulness locally increases slowwave activity during sleep. PLoS One 2 (3), e276.

Ilmoniemi, R.J., Kicic, D., 2010. Methodology for combined TMS and EEG. BrainTopogr. 22 (4), 233–248.

Ilmoniemi, R.J., Virtanen, J., Ruohonen, J., Karhu, J., Aronen, H.J., Naatanen, R., Katila,T., 1997. Neuronal responses to magnetic stimulation reveal cortical reactivityand connectivity. NeuroReport 8, 3537–3540.

Jasper, H.H., 1958. The ten–twenty electrode system of the International Federation.Electroencephalogr. Clin. Neurophysiol. 10, 370–375.

Jin, Y., Hilgetag, C.C., 2008. Perturbation of visuospatial attention by high-frequencyoffline rTMS. Exp. Brain Res. 189 (1), 121–128.

Kähkönen, S., Komssi, S., Wilenius, J., Ilmoniemi, R.J., 2005. Prefrontal transcranialmagnetic stimulation produces intensity-dependent EEG responses in humans.Neuroimage 24 (4), 959–960.

Kalla, R., Muggleton, N.G., Cowey, A., Walsh, V., 2009. Human dorsolateral prefrontalcortex is involved in visual search for conjunctions but not features: a theta TMSstudy. Cortex 45 (9), 1085–1090.

http://dx.doi.org/10.1016/j.brs.2009.10.002


5 Biobeh

K

K

K

K

K

K

K

K

K

K

K

K

K

K

K

K

K

K

K

L

L

L

L

L

L

M

M


ammer, T., Beck, S., Erb, M., Grodd, W., 2001b. The influence of current directionon phosphene thresholds evoked by transcranial magnetic stimulation. Clin.Neurophysiol. 112 (11), 2015–2021.

ammer, T., Beck, S., Thielscher, A., Laubis-Herrmann, U., Topka, H., 2001a. Motorthresholds in humans: a transcranial magnetic stimulation study comparingdifferent pulsewave forms, current directions and stimulator types. Clin. Neu-rophysiol. 112 (2), 250–258.

ansaku, K., Carver, B., Johnson, A., Matsuda, K., Sadato, N., Hallett, M., 2007. Therole of the human ventral premotor cortex in counting successive stimuli. Exp.Brain Res. 178 (3), 339–350.

apur, N., 1996. Paradoxical functional facilitation in brain-behavior research: acritical review. Brain 119, 1775–1790.

im, Y.H., Min, S.J., Ko, M.H., Park, J.W., Jang, S.H., Lee, P.K., 2005. Facilitating visu-ospatial attention for the contralateral hemifield by repetitive TMS on theposterior parietal cortex. Neurosci. Lett. 382 (3), 280–285.

irschen, M.P., Davis-Ratner, M.S., Jerde, T.E., Schraedley-Desmond, P., Desmond, J.E.,2006. Enhancement of phonological memory following transcranial magneticstimulation (TMS). Behav. Neurol. 17 (3–4), 187–194.

noch, D., Brugger, P., Regard, M., 2005. Suppressing versus releasing a habit:frequency-dependent effects of prefrontal transcranial magnetic stimulation.Cereb. Cortex 15 (7), 885–887.

noch, D., Pascual-Leone, A., Meyer, K., Treyer, V., Fehr, E., 2006. Diminishing recip-rocal fairness by disrupting the right prefrontal cortex. Science 314 (5800),829–832.

nops, A., Nuerk, H.C., Sparing, R., Foltys, H., Willmes, K., 2006. On the functionalrole of human parietal cortex in number processing: how gender mediatesthe impact of a ‘virtual lesion’ induced by rTMS. Neuropsychologia 44 (12),2270–2283.

och, G., Fernandez Del Olmo, M., Cheeran, B., Ruge, D., Schippling, S., Caltagirone, C.,Rothwell, J.C., 2007b. Focal stimulation of the posterior parietal cortex increasesthe excitability of the ipsilateral motor cortex. J. Neurosci. 27 (25), 6815–6822.

och, G., Fernandez Del Olmo, M., Cheeran, B., Schippling, S., Caltagirone, C., Driver,J., Rothwell, J.C., 2008. Functional interplay between posterior parietal and ipsi-lateral motor cortex revealed by twin-coil transcranial magnetic stimulationduring reach planning toward contralateral space. J. Neurosci. 28, 5944–5953.

och, G., Franca, M., Fernandez Del Olmo, M., Cheeran, B., Milton, R., Alvarez Sauco,M., Rothwell, J.C., 2006. Time course of functional connectivity between dorsalpremotor and contralateral motor cortex during movement selection. J. Neu-rosci. 26, 7452–7459.

och, G., Franca, M., Mochizuki, H., Marconi, B., Caltagirone, C., Rothwell, J.C., 2007a.Interactions between pairs of transcranial magnetic stimuli over the human leftdorsal premotor cortex differ from those seen in primary motor cortex. J. Physiol.578, 551–562.

och, G., Rothwell, J.C., 2009. TMS investigations into the task-dependent functionalinterplay between human posterior parietal and motor cortex. Behav. Brain Res.202 (2), 147–152.

och, G., Ruge, D., Cheeran, B., Fernandez Del Olmo, M., Pecchioli, C., Marconi, B., Ver-sace, V., Lo Gerfo, E., Torriero, S., Oliveri, M., Caltagirone, C., Rothwell, J.C., 2009.TMS activation of interhemispheric pathways between the posterior parietalcortex and the contralateral motor cortex. J. Physiol. 587, 4281–4292.

öhler, S., Paus, T., Buckner, R.L., Milner, B., 2004. Effects of left inferior prefrontalstimulation on episodic memory formation: a two-stage fMRI-rTMS study. J.Cogn. Neurosci. 16 (2), 178–188.

omssi, S., Kähkönen, S., 2006. The novelty value of the combined use of elec-troencephalography and transcranial magnetic stimulation for neuroscienceresearch. Brain Res. Rev. 52 (1), 183–192.

omssi, S., Kähkönen, S., Ilmoniemi, R.J., 2004. The effect of stimulus intensity onbrain responses evoked by transcranial magnetic stimulation. Hum. Brain Mapp.21 (3), 154–164.

ujirai, T., Caramia, M.D., Rothwell, J.C., Day, B.L., Thompson, P.D., Ferbert, A., Wroe,S., Asselman, P., Marsden, C.D., 1993. Corticocortical inhibition in human motorcortex. J. Physiol. 471, 501–519.

ambon Ralph, M.A., Pobric, G., Jefferies, E., 2009. Conceptual knowledge is under-pinned by the temporal pole bilaterally: convergent evidence from rTMS. Cereb.Cortex 19 (4), 832–838.

ang, N., Siebner, H.R., Ernst, D., Nitsche, M.A., Paulus, W., Lemon, R.N., Rothwell, J.C.,2004. Preconditioning with transcranial direct current stimulation sensitizes themotor cortex to rapid-rate transcranial magnetic stimulation and controls thedirection of after-effects. Biol. Psychiatry 56 (9), 634–639.

evkovitz, Y., Roth, Y., Harel, E.V., Braw, Y., Sheer, A., Zangen, A., 2007. A randomizedcontrolled feasibility and safety study of deep transcranial magnetic stimulation.Clin. Neurophysiol. 118 (12), 2730–2744.

ewald, J., Foltys, H., Töpper, R., 2002. Role of the posterior parietal cortex in spatialhearing. J. Neurosci. 22 (3), RC207.

ogie, R.H., 1995. Visuo-spatial Working Memory, vol. 14. Erlbaum, Hillsdale, NJ, pp.161.

uber, B., Kinnunen, L.H., Rakitin, B.C., Ellsasser, R., Stern, Y., Lisanby, S.H., 2007.Facilitation of performance in a working memory task with rTMS stimulationof the precuneus: frequency- and time-dependent effects. Brain Res. 1128 (1),120–129.

antovani, M., Van Velthoven, V., Fuellgraf, H., Feuerstein, T.J., Moser, A., 2006. Neu-ronal electrical high frequency stimulation enhances GABA outflow from humanneocortical slices. Neurochem. Int. 49 (4), 347–350.

ars, R.B., Klein, M.C., Neubert, F.X., Olivier, E., Buch, E.R., Boorman, E.D., Rushworth,M.F., 2009. Short-latency influence of medial frontal cortex on primary motorcortex during action selection under conflict. J. Neurosci. 29 (21), 6926–6931.


Marzi, C.A., Miniussi, C., Maravita, A., Bertolasi, L., Zanette, G., Rothwell, J.C., Sanes,J.N., 1998. Transcranial magnetic stimulation selectively impairs interhemi-spheric transfer of visuo-motor information in humans. Exp. Brain Res. 118 (3),435–438.

Massimini, M., Ferrarelli, F., Huber, R., Esser, S.K., Singh, H., Tononi, G., 2005.Breakdown of cortical effective connectivity during sleep. Science 309 (5744),2228–2232.

Mills, K.R., Nithi, K.A., 1997. Corticomotor threshold to magnetic stimulation: normalvalues and repeatability. Muscle Nerve 20 (5), 570–576.

Miniussi, C., Cappa, S.F., Cohen, L., Floel, A., Fregni, F., Nitsche, M., Oliveri, M., Pascual-Leone, A., Paulus, W., Priori, A., Walsh, V., 2008. Efficacy of rTMS/tDCS in cognitiveneurorehabilitation. Brain Stimul. 1, 326–336.

Miniussi, C., Ruzzoli, M., Walsh, V., 2010. The mechanism of transcranial magneticstimulation in cognition. Cortex 46 (1), 128–130.

Miniussi, C., Thut, G., 2010. Combining TMS and EEG offers new prospects in cogni-tive neuroscience. Brain Topogr. 22 (4), 49–56.

Moliadze, V., Zhao, Y., Eysel, U., Funke, K., 2003. Effect of transcranial magnetic stim-ulation on single-unit activity in the cat primary visual cortex. J. Physiol. 553,665–679.

Mottaghy, F.M., Gangitano, M., Sparing, R., Krause, B.J., Pascual-Leone, A., 2002. Seg-regation of areas related to visual working memory in the prefrontal cortexrevealed by rTMS. Cereb. Cortex 12 (4), 369–375.

Mottaghy, F.M., Hungs, M., Brügmann, M., Sparing, R., Boroojerdi, B., Foltys, H., Huber,W., Töpper, R., 1999. Facilitation of picture naming after repetitive transcranialmagnetic stimulation. Neurology 53 (8), 1806–1812.

Nieder, A., Miller, E.K., 2004. A parieto-frontal network for visual numerical infor-mation in the monkey. Proc. Natl. Acad. Sci. U.S.A. 101, 7457–7462.

Nyffeler, T., Cazzoli, D., Wurtz, P., Lüthi, M., von Wartburg, R., Chaves, S., Déruaz,A., Hess, C.W., Müri, R.M., 2008. Neglect-like visual exploration behaviour aftertheta burst transcranial magnetic stimulation of the right posterior parietalcortex. Eur. J. Neurosci. 27 (7), 1809–1813.

Nyffeler, T., Wurtz, P., Lüscher, H.R., Hess, C.W., Senn, W., Pflugshaupt, T., von Wart-burg, R., Lüthi, M., Müri, R.M., 2006. Repetitive TMS over the human oculomotorcortex: comparison of 1-Hz and theta burst stimulation. Neurosci. Lett. 409 (1),57–60.

Okamoto, M., Dan, H., Sakamoto, K., Takeo, K., Shimizu, K., Kohno, S., Oda, I., Isobe, S.,Suzuki, T., Kohyama, K., Dan, I., 2004. Three-dimensional probabilistic anatom-ical cranio-cerebral correlation via the international 10–20 system oriented fortranscranial functional brain mapping. Neuroimage 21 (1), 99–111.

Oliver, R., Bjoertomt, O., Driver, J., Greenwood, R., Rothwell, J., 2009. Novel ‘hunting’method using transcranial magnetic stimulation over parietal cortex disruptsvisuospatial sensitivity in relation to motor thresholds. Neuropsychologia 47(14), 3152–3161.

Oliveri, M., Caltagirone, C., Filippi, M.M., Traversa, R., Cicinelli, P., Pasqualetti, P.,Rossini, P.M., 2000. Paired transcranial magnetic stimulation protocols reveala pattern of inhibition and facilitation in the human parietal cortex. J. Physiol.529, 461–468.

Oliveri, M., Finocchiaro, C., Shapiro, K., Gangitano, M., Caramazza, A., Pascual-Leone,A., 2004a. All talk and no action: a transcranial magnetic stimulation study ofmotor cortex activation during action word production. J. Cogn. Neurosci. 16 (3),374–381.

Oliveri, M., Rausei, V., Koch, G., Torriero, S., Turriziani, P., Caltagirone, C., 2004b.Overestimation of numerical distances in the left side of space. Neurology 63(11), 2139–2141.

Oliviero, A., Profice, P., Tonali, P.A., Pilato, F., Saturno, E., Dileone, M., Ranieri, F., DiLazzaro, V., 2006. Effects of aging on motor cortex excitability. Neurosci. Res. 55,74–77.

O’Shea, J., Johansen-Berg, H., Trief, D., Göbel, S., Rushworth, M.F., 2007b. Functionallyspecific reorganization in human premotor cortex. Neuron 54 (3), 479–490.

O’Shea, J., Sebastian, C., Boorman, E.D., Johansen-Berg, H., Rushworth, M.F., 2007c.Functional specificity of human premotor–motor cortical interactions duringaction selection. Eur. J. Neurosci. 26 (7), 2085–2095.

O’Shea, J., Taylor, P.C., Rushworth, M.F., 2008. Imaging causal interactions duringsensorimotor processing. Cortex 44 (5), 598–608.

O’Shea, J., Walsh, V., 2007a. Transcranial magnetic stimulation. Curr. Biol. 17 (6),R196–199.

Pascual-Leone, A., Rubio, B., Pallardó, F., Catalá, M.D., 1996. Rapid-rate transcra-nial magnetic stimulation of left dorsolateral prefrontal cortex in drug-resistantdepression. Lancet 348 (9022), 233–237.

Pascual-Leone, A., Walsh, V., 2001. Fast backprojections from the motion to theprimary visual area necessary for visual awareness. Science 92 (5516), 510–512.

Pascual-Leone, A., Walsh, V., Rothwell, J., 2000. Transcranial magnetic stimulationin cognitive neuroscience-virtual lesion, chronometry, and functional connec-tivity. Curr. Opin. Neurobiol. 10, 232–237.

Pasley, B.N., Allen, E., Freeman, R.D., 2009. State-dependent variability of neuronalresponses to transcranial magnetic stimulation of the visual cortex. Neuron 62,291–303.

Paus, T., 1999. Imaging the brain before, during, and after transcranial magneticstimulation. Neuropsychologia 37 (2), 219–224.

Paus, T., 2005. Inferring causality in brain images: a perturbation approach. Philos.Trans. R. Soc. Lond. B Biol. Sci. 360 (1457), 1109–1114.

Paus, T., Jech, R., Thompson, C.J., Comeau, R., Peters, T., Evans, A.C., 1997. Tran-scranial magnetic stimulation during positron emission tomography: a newmethod for studying connectivity of the human cerebral cortex. J. Neurosci. 17(9), 3178–3184.

Biobeh

P

P

P

P

P

P

P

P

P

R

R

R

R

R

R

R

R

R

R

R

R

R

R

R

R

R

R

R

R


einemann, A., Lebner, C., Conrad, B., Siebner, H.R., 2001. Age-related decrease inpaired-pulse intracortical inhibition in the human primary motor cortex. Neu-rosci. Lett. 313, 33–36.

erez, M.A., Tanaka, S., Wise, S.P., Sadato, N., Tanabe, H.C., Willingham, D.T., Cohen,L.G., 2007. Neural substrates of intermanual transfer of a newly acquired motorskill. Curr. Biol. 17 (21), 1896–1902.

iazza, M., Pinel, P., Le Bihan, D., Dehaene, S., 2007. A magnitude code common tonumerosities and number symbols in human intraparietal cortex. Neuron 53(2), 293–305.

inel, P., Piazza, M., Le Bihan, D., Dehaene, S., 2004. Distributed and overlappingcerebral representations of number, size, and luminance during comparativejudgments. Neuron 41, 983–993.

obric, G., Jefferies, E., Ralph, M.A., 2010. Amodal semantic representations dependon both anterior temporal lobes: Evidence from repetitive transcranial magneticstimulation. Neuropsychologia 48 (5), 1336–1342.

obric, G., Lambon Ralph, M.A., Jefferies, E., 2009. The role of the anterior temporallobes in the comprehension of concrete and abstract words: rTMS evidence.Cortex 45 (9), 1104–1110.

ostle, B.R., Feredoes, E., 2010. Stronger inference with direct manipulation of brainfunction. Cortex 46 (1), 121–123.

ridmore, S., Fernandes Filho, J.A., Nahas, Z., Liberatos, C., George, M.S., 1998. Motorthreshold in transcranial magnetic stimulation: a comparison of a neurophysi-ological method and a visualization of movement method. J. ECT. 14 (1), 25–27.

ulvermüller, F., Hauk, O., Nikulin, V.V., Ilmoniemi, R.J., 2005. Functional linksbetween motor and language systems. Eur. J. Neurosci. 21 (3), 793–797.

eis, J., Swayne, O.B., Vandermeeren, Y., Camus, M., Dimyan, M.A., Harris-Love, M.,Perez, M.A., Ragert, P., Rothwell, J.C., Cohen, L.G., 2008. Contribution of transcra-nial magnetic stimulation to the understanding of cortical mechanisms involvedin motor control. J. Physiol. 586 (2), 325–351.

idding, M.C., Rothwell, J.C., 2007. Is there a future for therapeutic use of transcranialmagnetic stimulation? Nat. Rev. Neurosci. 8 (7), 559–567.

o, T., Cheifet, S., Ingle, H., Shoup, R., Rafal, R., 1999. Localization of the humanfrontal eye fields and motor hand area with transcranial magnetic stimulationand magnetic resonance imaging. Neuropsychologia 37 (2), 225–231.

obertson, E.M., Theoret, H., Pascual-Leone, A., 2003. Studies in cognition: the prob-lems solved and created by transcranial magnetic stimulation. J. Cogn. Neurosci.15, 948–960.

oebroeck, A., Formisano, E., Goebel, R., 2005. Mapping directed influence over thebrain using Granger causality and fMRI. Neuroimage 25, 230–242.

osenthal, C.R., Roche-Kelly, E., Husain, M., Kennard, C., 2009. Response-dependentcontributions of primary motor cortex and angular gyrus to manual and percep-tual sequence learning. J. Neurosci. 29 (48), 15115–15125.

ossi, S., Cappa, S.F., Babiloni, C., Pasqualetti, P., Miniussi, C., Carducci, F., Babiloni,F., Rossini, P.M., 2001. Prefrontal cortex in long-term memory: an “interference”approach using magnetic stimulation. Nat. Neurosci. 4 (9), 948–952.

ossi, S., Ferro, M., Cincotta, M., Ulivelli, M., Bartalini, S., Miniussi, C., Giovannelli,F., Passero, S., 2007. A real electro-magnetic placebo (REMP) device for shamtranscranial magnetic stimulation (TMS). Clin. Neurophysiol. 118 (3), 709–716.

ossi, S., Hallett, M., Rossini, P.M., Pascual-Leone, A., et al., 2009. The safety of TMSconsensus group. Safety, ethical considerations, and application guidelines forthe use of transcranial magnetic stimulation in clinical practice and research.Clin. Nurophysiol. 120 (12), 2008–2039.

ossi, S., Innocenti, I., Polizzotto, N.R., Feurra, M., De Capua, A., Ulivelli, M., Bartalini,S., Cappa, S.F., in press. Temporal dynamics of memory trace formation in thehuman prefrontal cortex. Cereb. Cortex doi:10.1093/cercor/bhq103.

ossi, S., Miniussi, C., Pasqualetti, P., Babiloni, C., Rossini, P.M., Cappa, S.F., 2004.Age-related functional changes of prefrontal cortex in long-term memory:a repetitive transcranial magnetic stimulation study. J. Neurosci. 24 (36),7939–7944.

ossi, S., Rossini, P.M., 2004. TMS in cognitive plasticity and the potential for reha-bilitation. Trends Cogn. Sci. 8 (6), 273–279.

ossini, P.M., Barker, A.T., Berardelli, A., Caramia, M.D., Caruso, G., Cracco, R.Q., Dim-itrijevic, M.R., Hallett, M., Katayama, Y., Lücking, C.H., et al., 1994. Non-invasiveelectrical and magnetic stimulation of the brain, spinal cord and roots: basicprinciples and procedures for routine clinical application. Report of an IFCNcommittee. Electroencephalogr. Clin. Neurophysiol. 91 (2), 79–92.

ossini, P.M., Desiato, M.T., Caramia, M.D., 1992. Age-related changes of motorevoked potentials in healthy humans: non-invasive evaluation of central andperipheral motor tract excitability and conductivity. Brain Res. 593, 14–19.

oth, Y., Amir, A., Levkovitz, Y., Zangen, A., 2007. Three-dimensional distribution ofthe electric field induced in the brain by transcranial magnetic stimulation usingfigure-8 and deep H-coils. J. Clin. Neurophysiol. 24 (1), 31–38.

othkegel, H., Sommer, M., Paulus, W., 2010. Breaks during 5 Hz rTMS are essentialfor facilitatory after effects. Clin. Neurophysiol. 121 (3), 426–430.

othwell, J.C., Hallett, M., Berardelli, A., Eisen, A., Rossini, P.M., Paulus, W., 1999.Magnetic stimulation: motor evoked potentials. The International Federation ofClinical Neurophysiology. Electroencephalogr. Clin. Neurophysiol. (Suppl. 52),97–103.

ubinsten, O., Henik, A., 2005. Automatic activation of internal magnitudes: a studyof developmental dyscalculia. Neuropsychology 19, 641–648.

ubinsten, O., Henik, A., 2006. Double dissociation of functions in developmentaldyslexia and dyscalculia. J. Educ. Psychol. 98, 854–867.

uff, C.C., Bestmann, S., Blankenburg, F., Bjoertomt, O., Josephs, O., Weiskopf, N.,Deichmann, R., Driver, J., 2008. Distinct causal influences of parietal versusfrontal areas on human visual cortex: evidence from concurrent TMS-fMRI.Cereb. Cortex 18 (4), 817–827.

avioral Reviews 35 (2011) 516–536 535

Ruff, C.C., Blankenburg, F., Bjoertomt, O., Bestmann, S., Freeman, E., Haynes, J.D., Rees,G., Josephs, O., Deichmann, R., Driver, J., 2006. Concurrent TMS-fMRI and psy-chophysics reveal frontal influences on human retinotopic visual cortex. Curr.Biol. 16 (15), 1479–1488.

Ruff, C.C., Blankenburg, F., Bjoertomt, O., Bestmann, S., Weiskopf, N., Driver, J., 2009b.Hemispheric differences in frontal and parietal influences on human occipitalcortex: direct confirmation with concurrent TMS-fMRI. J. Cogn. Neurosci. 21 (6),1146–1161.

Ruff, C.C., Driver, J., Bestmann, S., 2009a. Combining TMS and fMRI: from ‘virtuallesions’ to functional-network accounts of cognition. Cortex 45 (9), 1043–1049.

Rusconi, E., Bueti, D., Riello, M., Walsh, V., Butterworth, B., 2008. Orienting attentionto numbers: involvement of frontal lobes. In: OHBM Annual Meeting, Mel-bourne.

Rusconi, E., Bueti, D., Walsh, V., Butterworth, B., in press. Contribu-tion of frontal cortex to the spatial representation of number. Cortexdoi:10.1016/j.cortex.2009.08.005.

Rusconi, E., Turatto, M., Umiltà, C., 2007. Two orienting mechanisms in posteriorparietal lobule: an rTMS study of the Simon and SNARC effects. Cogn. Neuropsy-chol. 24 (4), 373–392.

Rusconi, E., Umiltà, C., 2008. Mathematics and TMS. In: Wassermann, E., Epstein,C.M., Ziemann, U., Walsh, V., Paus, T., Lisanby, L.S. (Eds.), The Oxford Handbookof Transcranial Stimulation. OUP, Oxford.

Rusconi, E., Walsh, V., Butterworth, B., 2005. Dexterity with numbers: rTMS over leftangular gyrus disrupts finger gnosis and number processing. Neuropsychologia43 (11), 1609–1624.

Rushworth, M.F., Ellison, A., Walsh, V., 2001. Complementary localization and later-alization of orienting and motor attention. Nat. Neurosci. 4 (6), 656–661.

Ruzzoli, M., Marzi, C.A., Miniussi, C., 2010. The neural mechanisms of the effectsof transcranial magnetic stimulation on perception. J. Neurophysiol. 103 (6),2982–2989.

Ryan, S., Bonilha, L., Jackson, S.R., 2006. Individual variation in the location of theparietal eye fields: a TMS study. Exp. Brain Res. 173 (3), 389–394.

Sack, A.T., 2010. Does TMS need functional imaging? Cortex 46 (1), 131–133.Sack, A.T., Cohen Kadosh, R., Schuhmann, T., Moerel, M., Walsh, V., Goebel, R., 2009.

Optimizing functional accuracy of TMS in cognitive studies: a comparison ofmethods. J. Cogn. Neurosci. 21 (2), 207–221.

Sack, A.T., Kohler, A., Bestmann, S., Linden, D.E., Dechent, P., Goebel, R., Baudewig,J., 2007. Imaging the brain activity changes underlying impaired visuospatialjudgments: simultaneous FMRI, TMS, and behavioral studies. Cereb. Cortex 17(12), 2841–2852.

Sack, A.T., Linden, D.E.J., 2003. Combining transcranial magnetic stimulation andfunctional imaging in cognitive brain research: possibilities and limitations.Brain Res. Rev. 43, 41–56.

Salillas Pérez, E.S., Basso, D., Baldi, M., Semenza, C., Vecchi, T., 2009. Motion onnumbers: transcranial magnetic stimulation on the ventral intraparietal sulcusalters both numerical and motion processes. J. Cogn. Neurosci. 21 (11), 2129–2138.

Sandrini, M., Cappa, S.F., Rossi, S., Rossini, P.M., Miniussi, C., 2003. The role of pre-frontal cortex in verbal episodic memory: rTMS evidence. J. Cogn. Neurosci. 15(6), 855–861.

Sandrini, M., Rossini, P.M., Miniussi, C., 2004. The differential involvement of inferiorparietal lobule in number comparison: an rTMS study. Neuropsychologia 42 (14),1902–1909.

Sandrini, M., Rusconi, E., 2009. A brain for numbers. Cortex 45 (7), 796–803.Sato, M., Cattaneo, L., Rizzolatti, G., Gallese, V., 2007. Numbers within our hands:

modulation of corticospinal excitability of hand muscles during numerical judg-ment. J. Cogn. Neurosci. 19 (4), 684–693.

Sawaki, L., Yaseen, Z., Kopylev, L., Cohen, L.G., 2003. Age-dependent changes in theability to encode a novel elementary motor memory. Ann. Neurol. 53, 521–524.

Schäfer, M., Biesecker, J.C., Schulze-Bonhage, A., Ferbert, A., 1997. Transcranial mag-netic double stimulation: influence of the intensity of the conditioning stimulus.Electroencephalogr. Clin. Neurophysiol. 105, 462–469.

Schuhmann, T., Schiller, N.O., Goebel, R., Sack, A.T., 2009. The temporal characteris-tics of functional activation in Broca’s area during overt picture naming. Cortex45 (9), 1111–1116.

Siebner, H.R., Bergmann, T.O., Bestmann, S., Massimini, M., et al., 2009b. Consensuspaper: combining transcranial magnetic stimulation with neuroimaging. BrainStimul. 2, 58–80.

Siebner, H.R., Hartwigsen, G., Kassuba, T., Rothwell, J.C., 2009a. How does transcra-nial magnetic stimulation modify neuronal activity in the brain? Implicationsfor studies of cognition. Cortex 45 (9), 1035–1042.

Siebner, H.R., Lang, N., Rizzo, V., Nitsche, M.A., Paulus, W., Lemon, R.N., Rothwell, J.C.,2004. Preconditioning of low-frequency repetitive transcranial magnetic stim-ulation with transcranial direct current stimulation: evidence for homeostaticplasticity in the human motor cortex. J. Neurosci. 24 (13), 3379–3385.

Siebner, H.R., Rothwell, J., 2003. Transcranial magnetic stimulation: new insightsinto representational cortical plasticity. Exp. Brain Res. 148 (1), 1–16.

Siebner, H.R., Willoch, F., Peller, M., Auer, C., Boecker, H., Conrad, B., Bartenstein, P.,1998. Imaging brain activation induced by long trains of repetitive transcranialmagnetic stimulation. Neuroreport 9 (5), 943–948.

Silvanto, J., Lavie, N., Walsh, V., 2006. Stimulation of the human frontal eye fieldsmodulates sensitivity of extrastriate visual cortex. J. Neurophysiol. 96 (2),941–945.

Silvanto, J., Muggleton, N.G., Cowey, A., Walsh, V., 2007. Neural activation state deter-mines behavioral susceptibility to modified theta burst transcranial magneticstimulation. Eur. J. Neurosci. 26 (2), 523–528.

http://dx.doi.org/10.1093/cercor/bhq103

http://dx.doi.org/10.1016/j.cortex.2009.08.005

5 Biobeh

S

S

S

S

S

S

S

S

S

S

S

S

S

S

S

S

T

T

T


ilvanto, J., Muggleton, N., Lavie, N., Walsh, V., 2009. The perceptual and functionalconsequences of parietal top-down modulation on the visual cortex. Cereb. Cor-tex 19 (2), 327–330.

ilvanto, J., Muggleton, N., Walsh, V., 2008. State-dependency in brain stimulationstudies of perception and cognition. Trends Cogn. Sci. 12 (12), 447–454.

ilvanto, J., Muggleton, N.G., 2008. Testing the validity of the TMS state-dependencyapproach: targeting functionally distinct motion-selective neural populationsin visual areas V1/V2 and V5/MT+. Neuroimage 40 (4), 1841–1848.

ilvanto, J., Pascual-Leone, A., 2008. State-dependency of transcranial magneticstimulation. Brain Topogr. 21 (1), 1–10.

nyder, A., 2009. Explaining and inducing savant skills: privileged access to lowerlevel, less-processed information. Phil. Trans. R. Soc. B 364, 1399–1405.

olé-Padullés, C., Bartrés-Faz, D., Junqué, C., Clemente, I.C., Molinuevo, J.L., Bargalló,N., Sánchez-Aldeguer, J., Bosch, B., Falcón, C., Valls-Solé, J., 2006. Repetitive tran-scranial magnetic stimulation effects on brain function and cognition amongelders with memory dysfunction. A randomized sham-controlled study. Cereb.Cortex 16 (10), 1487–1493.

ommer, M., Alfaro, A., Rummel, M., Speck, S., Lang, N., Tings, T., Paulus, W., 2006.Half sine, monophasic and biphasic transcranial magnetic stimulation of thehuman motor cortex. Clin. Neurophysiol. 117 (4), 838–844.

paring, R., Buelte, D., Meister, I.G., Paus, T., Fink, G.R., 2008. Transcranial magneticstimulation and the challenge of coil placement: a comparison of conven-tional and stereotaxic neuronavigational strategies. Hum. Brain Mapp. 29 (1),82–96.

paring, R., Hesse, M.D., Fink, G.R., 2010. Neuronavigation for transcranial mag-netic stimulation (TMS): where we are and where we are going. Cortex 46 (1),118–120.

paring, R., Mottaghy, F.M., Hungs, M., Brügmann, M., Foltys, H., Huber, W., Töp-per, R., 2001. Repetitive transcranial magnetic stimulation effects on languagefunction depend on the stimulation parameters. J. Clin. Neurophysiol. 18 (4),326–330.

tewart, L.M., Walsh, V., Rothwell, J.C., 2001. Motor and phosphene thresholds:a transcranial magnetic stimulation correlation study. Neuropsychologia 39,415–419.

toeckel, C., Gough, P.M., Watkins, K.E., Devlin, J.T., 2009. Supramarginal gyrusinvolvement in visual word recognition. Cortex 45 (9), 1091–1096.

tokes, M.G., Chambers, C.D., Gould, I.C., English, T., McNaught, E., McDonald, O., Mat-tingley, J.B., 2007. Distance-adjusted motor threshold for transcranial magneticstimulation. Clin. Neurophysiol. 118 (7), 1617–1625.

tokes, M.G., Chambers, C.D., Gould, I.C., Henderson, T.R., Janko, N.E., Allen, N.B.,Mattingley, J.B., 2005. Simple metric for scaling motor threshold based on scalp-cortex distance: application to studies using transcranial magnetic stimulation.J. Neurophysiol. 94 (6), 4520–4527.

trafella, A.P., Paus, T., Barrett, J., Dagher, A., 2001. Repetitive transcranial magneticstimulation of the human prefrontal cortex induces dopamine release in thecaudate nucleus. J. Neurosci. 21, RC157.

trafella, A.P., Paus, T., Fraraccio, M., Dagher, A., 2003. Striatal dopamine releaseinduced by repetitive transcranial magnetic stimulation of the human motorcortex. Brain 126, 2609–2615.

anaka, S., Honda, M., Hanakawa, T., Cohen, L.G., in press. Differential contri-bution of the supplementary motor area to stabilization of a procedural
motor skill acquired through different practice schedules. Cereb. Cortexdoi:10.1093/cercor/bhp276.
aylor, P.C., Nobre, A.C., Rushworth, M.F., 2007. FEF TMS affects visual cortical activ-ity. Cereb. Cortex 17 (2), 391–399.

aylor, P.C., Walsh, V., Eimer, M., 2008. Combining TMS and EEG to study cognitivefunction and cortico-cortico interactions. Behav. Brain Res. 191 (2), 141–147.


Terao, Y., Ugawa, Y., Sakai, K., Miyauchi, S., Fukuda, H., Sasaki, Y., Takino, R., Hanajima,R., Furubayashi, T., Pütz, B., Kanazawa, I., 1998. Localizing the site of magneticbrain stimulation by functional MRI. Exp. Brain Res. 121 (2), 145–152.

Théoret, H., Haque, J., Pascual-Leone, A., 2001. Increased variability of paced fingertapping accuracy following repetitive magnetic stimulation of the cerebellumin humans. Neurosci. Lett. 306 (1–2), 29–32.

Théoret, H., Kobayashi, M., Valero-Cabré, A., Pascual-Leone, A., 2003. Exploring para-doxical functional facilitation with TMS. Suppl. Clin. Neurophysiol. 56, 211–219.

Thut, G., Miniussi, C., 2009. New insights into rhythmic brain activity from TMS–EEGstudies. Trends Cogn. Sci. 13 (4), 182–189.

Thut, G., Nietzel, A., Pascual-Leone, A., 2005. Dorsal posterior parietal rTMS affectsvoluntary orienting of visuospatial attention. Cereb. Cortex 15 (5), 628–638.

Töpper, R., Mottaghy, F.M., Brügmann, M., Noth, J., Huber, W., 1998. Facilitation ofpicture naming by focal transcranial magnetic stimulation of Wernicke’s area.Exp. Brain Res. 121 (4), 371–378.

Torriero, S., Oliveri, M., Koch, G., Caltagirone, C., Petrosini, L., 2004. Interference ofleft and right cerebellar rTMS with procedural learning. J. Cogn. Neurosci. 16 (9),1605–1611.

Ugawa, Y., Uesaka, Y., Terao, Y., Hanajima, R., Kanazawa, I., 1995. Magnetic stimula-tion over the cerebellum in humans. Ann. Neurol. 37, 703–713.

Umiltà, C., Priftis, K., Zorzi, M., 2009. The spatial representation of numbers: evidencefrom neglect and pseudoneglect. Exp. Brain Res. 192 (3), 561–569.

Valero-Cabre, A., Payne, B.R., Rushmore, J., Lomber, S.G., Pascual-Leone, A., 2005.Impact of repetitive transcranial magnetic stimulation of the parietal cortex onmetabolic brain activity: a 14C–2DG tracing study in the cat. Exp. Brain Res. 163,1–12.

Vallesi, A., Shallice, T., Walsh, V., 2007. Role of the prefrontal cortex in the forepe-riod effect: TMS evidence for dual mechanisms in temporal preparation. Cereb.Cortex 17 (2), 466–474.

van Donkelaar, P., Müri, R., 2002. Craniotopic updating of visual space across sac-cades in the human posterior parietal cortex. Proc. Biol. Sci. 269 (1492), 735–739.

Wagner, T., Rushmore, J., Eden, U., Valero-Cabre, A., 2009. Biophysical foundationsunderlying TMS: setting the stage for an effective use of neurostimulation in thecognitive neurosciences. Cortex 45 (9), 1025–1034.

Walsh, V., Cowey, A., 2000. Transcranial magnetic stimulation and cognitive neuro-science. Nat. Rev. Neurosci. 1 (1), 73–79.

Walsh, V., Ellison, A., Battelli, L., Cowey, A., 1998. Task-specific impairments andenhancements induced by magnetic stimulation of human visual area V5. Proc.Biol. Sci. 265 (1395), 537–543.

Ward, N.S., Cohen, L.G., 2004. Mechanisms underlying recovery of motor functionafter stroke. Arch. Neurol. 61 (12), 1844–1848.

Wassermann, E., Epstein, C., Ziemann, U., Walsh, V., Paus, T., Lisanby, S. (Eds.), 2008.The Oxford Handbook of Transcranial Stimulation. OUP, Oxford.

Wassermann, E.M., 1998. Risk and safety of repetitive transcranial magnetic stim-ulation: report and suggested guidelines from the International Workshop onthe Safety of Repetitive Transcranial Magnetic Stimulation, June 5–7, 1996. Elec-troencephalogr. Clin. Neurophysiol. 108 (1), 1–16.

Wassermann, E.M., 2002. Variation in the response to transcranial magnetic brainstimulation in the general population. Clin. Neurophysiol. 113, 1165–1571.

Wilkinson, L., Teo, J.T., Obeso, I., Rothwell, J.C., Jahanshahi, M., 2009. The contri-bution of primary motor cortex is essential for probabilistic implicit sequence
learning: evidence from theta burst magnetic stimulation. J. Cogn. Neurosci. 22(3), 427–436.
Ziemann, U., 2010. TMS in cognitive neuroscience: virtual lesion and beyond. Cortex46 (1), 124–127.

Ziemann, U., Rothwell, J.C., Ridding, M.C., 1996. Interaction between intracorticalinhibition and facilitation in human motor cortex. J. Physiol. 496, 873–881.


the use of transcranial magnetic stimulation in cognitive ... · tms has nowadays become a standard...

Documents