www.sciencedirect.com

c o r t e x x x x ( 2 0 1 4 ) 1e9

Available online at

ScienceDirect

Journal homepage: www.elsevier.com/locate/cortex

Special issue: Research report

Cerebellar vermis plays a causal role in visualmotion discrimination

Zaira Cattaneo a,b,*, Chiara Renzi b, Stefano Casali b,c, Juha Silvanto d,e,Tomaso Vecchi b,c, Costanza Papagno a and Egidio D’Angelo b,c

aDepartment of Psychology, University of Milano-Bicocca, Milano, ItalybBrain Connectivity Center, National Neurological Institute C. Mondino, Pavia, ItalycDepartment of Brain and Behavioral Sciences, University of Pavia, Pavia, ItalydDepartment of Psychology, Faculty of Science and Technology, University of Westminster, UKeBrain Research Unit, O.V. Lounasmaa Laboratory, School of Science, Aalto University, Espoo, Finland

a r t i c l e i n f o

Article history:

Received 28 October 2013

Reviewed 18 December 2013

Revised 27 December 2013

Accepted 23 January 2014

Published online xxx

Keywords:

Cerebellum

Visual

Motion detection

Orientation discrimination

TMS

* Corresponding author. Department of PsycE-mail address: zaira.cattaneo@unimib.it

Please cite this article in press as: CattanCortex (2014), http://dx.doi.org/10.1016/j.

http://dx.doi.org/10.1016/j.cortex.2014.01.0120010-9452/ª 2014 Elsevier Ltd. All rights rese

a b s t r a c t

Cerebellar patients have been found to show deficits in visual motion discrimination,

suggesting that the cerebellum may play a role in visual sensory processing beyond

mediating motor control. Here we show that triple-pulse online transcranial magnetic

stimulation (TMS) over cerebellar vermis but not over the cerebellar hemispheres signifi-

cantly impaired motion discrimination. Critically, the interference caused by vermis TMS

on motion discrimination did not depend on an indirect effect of TMS over nearby visual

areas, as demonstrated by a control experiment in which TMS over V1 but not over cere-

bellar vermis significantly impaired orientation discrimination. These findings demon-

strate the causal role of the cerebellar vermis in visual motion processing in neurologically

normal participants.

ª 2014 Elsevier Ltd. All rights reserved.

1. Introduction

The cerebellum has been suggested to play a critical role in

sensory processing, beyond being involved in motor control

(Bower, 1997; D’Angelo & Casali, 2012; Manto et al., 2012).

Specifically, psychophysical studies with cerebellar patients

have demonstrated the functional significance of this region

in motion perception independent from actual movements

(Baumann & Mattingley, 2010; Jokisch, Troje, Koch, Schwarz,

hology, University of Mila(Z. Cattaneo).

eo, Z., et al., Cerebellarcortex.2014.01.012

rved.

& Daum, 2005; Nawrot & Rizzo, 1995, 1998; Thier, Haarmeier,

Treue, & Barash, 1999). Both in acute and chronic post-ictus

phases, cerebellar patients with lesions in midline structures

have deficits in motion discrimination, which was not the

case when lesions were restricted to the cerebellar hemi-

spheres (Nawrot & Rizzo, 1995, 1998). The performance defi-

cits, which occurred even though cortical processing was

intact, were neither due to aberrant eye movements nor an

indirect consequence of motor deficits (Ivry & Diener, 1991;

no-Bicocca, Milano, Italy.

vermis plays a causal role in visual motion discrimination,

c o r t e x x x x ( 2 0 1 4 ) 1e92

Nawrot & Rizzo, 1995, 1998; Thier et al., 1999). However, lateral

parts of the cerebellumhave also been associatedwith deficits

in motion discrimination (Jokisch et al., 2005); thus, the

localization of cerebellar areas controlling motion discrimi-

nation remains unclear.

Neuroimaging evidence on the role of the cerebellum in

motion discrimination is also somewhat conflicting. In posi-

tron emission tomography (PET) studies, Dupont, Orban, De

Bruyn, Verbruggen, and Mortelmans (1994) found vermal

activity in response to a moving dot pattern, and Barbur,

Watson, Frackowiak, and Zeki (1993) reported such activity

in a blindsight patient with a unilateral lesion in V1. How-

ever, using functional magnetic resonance imaging (fMRI)

Baumann and Mattingley (2010) observed a complex pattern

of activation in the cerebellar hemispheres, which was

correlated with both auditory and visual motion signal

strength.

Here we attempted to resolve this controversy by the use

of brain stimulation in neurologically healthy volunteers in

three different experiments. Brain stimulation can shed

light on the causal role of the targeted brain regions in

mediating specific perceptual functions, thus overcoming

the correlational nature of neuroimaging and the poor

focality of brain lesions. In the first two experiments online

transcranial magnetic stimulation (TMS) was applied either

over the cerebellar vermis (Experiment 1) or over the left

and right cerebellar hemispheres (Experiment 2), while

participants performed a visual motion discrimination task.

In order to rule out effects due to nonspecific effects of TMS

(e.g., tapping sensation, auditory distraction) or to spread of

stimulation from the cerebellar vermis to the adjacent vi-

sual cortex, in Experiment 3 TMS was applied during a vi-

sual orientation discrimination task, which is known to

causally involve the early visual cortex but not cerebellum

(e.g., Neary, Anand, & Hotson, 2005 for previous TMS evi-

dence using orientation discrimination).

Fig. 1 e The timeline of an experimental trial in the motion

discrimination task. In experiment 1, TMS was applied

over the cerebellar vermis and the primary visual area (V1).

In experiment 2, TMS was applied over the cerebellar left

and right hemisphere. In both experiments, sham TMS

(control condition) was also applied over the targeted sites

in addition to a No TMS condition.

2. Experiments 1 and 2: motiondiscrimination

2.1. Methods

2.1.1. ParticipantsTwelve healthy volunteers (10 F, mean age: 21.58 ys, SD: 1.0,

range: 20e23) took part in Experiment 1 and twelve healthy

volunteers (11 F, mean age: 22.17 ys, SD: 2.2, range: 20e27),

none of whom had participated in Experiment 1, took part in

Experiment 2. All participants were right-handed (Oldfield,

1971) and had normal or corrected to normal vision. Prior to

the experiment, each participant filled in a questionnaire

(translated and adapted from Rossi, Hallett, Rossini, &

Pascual-Leone, 2011) to evaluate compatibility with TMS.

None of the volunteers had a history of neurological disorders

or brain trauma or family history of epilepsy. Written

informed consent was obtained from all participants before

the experiment. The protocol was approved by the local

ethical committee and participants were treated in accor-

dance with the Declaration of Helsinki.

Please cite this article in press as: Cattaneo, Z., et al., CerebellarCortex (2014), http://dx.doi.org/10.1016/j.cortex.2014.01.012

2.1.2. Visual stimuliAll stimuli were presented centrally on a 1700 TFT-LCD com-

putermonitor (screen resolution: 1440*900 pixels; refresh rate:

75 Hz). Stimuli consisted of 100 white dots (one pixel each),

placed at random positions within an imaginary square that

subtended 4.3� � 4.3� of visual angle. The coherent dots were

moving either to the right or left within the virtual square on a

black background at a speed of 1 pixel per frame (at 2.15 deg/

sec); noise dots changed direction randomly at every screen

refresh. Each trial began with a fixation cross appearing in the

middle of the screen for an interval comprised between 2600

and 2800 msec, followed by a blank screen for 500 msec, after

which the stimulus appeared. Stimulus durationwas between

40 and 67 msec, depending on participant’s ability (see

thresholding procedure below). After response, the next trial

started. An example of experimental trial is shown in Fig. 1.

Participants were required to report whether the visual stim-

ulus moved to the left or right by left/right key presses using

their right index and middle finger. Response speed was

stressed in addition to accuracy.

2.1.3. ThresholdingPrior to the experiment, each participant underwent a

thresholding procedure to determine the ratio of coherent

moving dots and the number of frames (frame duration: 13

msec) necessary to obtain a stable performance around 75%

accuracy. This was achieved by running a block in which

motion coherence (i.e., the amount of dots moving coherently

either to the left or right) ranged from 40% to 90% in steps of

10% with six levels. In this block, the motion stimulus con-

tained five frames (with 20 trials per level with the method of

vermis plays a causal role in visual motion discrimination,

c o r t e x x x x ( 2 0 1 4 ) 1e9 3

constant stimuli, i.e., coherence levels were randomly inter-

leaved). Hence, another block was run using the three

consecutive levels of noise (20 trials each) that together

resulted in a mean level of accuracy around 75% (if partici-

pant’s accuracy was higher than 75% at the maximum levels

of noise in the first block, the block was repeated with a lower

number of displayed frames to make the discrimination more

difficult). From this block, we selected the level of noise whose

mean accuracy was closer to 75%. Overall, the percentage of

noise dots used in the experiment ranged from 40% to 85%, the

number of frames ranged from 3 (corresponding to a stimulus

duration of 40 msec) to 5 (corresponding to a stimulus dura-

tion of 67 msec).

2.1.4. TMSTMS was administered by means of a Magstim Rapid2 ma-

chine (Magstim Co Ltd, Whitland, UK) with a 70 mm butterfly

coil. At the beginning of each session the resting motor

threshold (rMT) was measured from the first dorsal inteross-

eous (FDI)muscle of the right hand. This intensity was defined

as the minimum stimulation intensity that produced motor

evoked potentials (MEPs) � 50 mV peak to peak in at least 5 out

of 10 trials (Rossini et al., 1994). For each participant, the in-

tensity of TMS stimulation was equal to 100% of the rMT

(mean intensity: 52.0%, SD: 6.5%).

In Experiment 1, there were two stimulation sites: V1 and

cerebellar vermis. The cerebellum was stimulated as in pre-

vious studies at a point 1 cm inferior to the inion on a line

joining the inion to the nasion (Lee et al., 2007; Schutter, De

Weijer, Meuwese, Morgan, & Van Honk, 2008; Theoret,

Haque, & Pascual-leone, 2001) with the handle pointing su-

periorly. V1 was localized as the point lying 1.5 cm superior to

the inion on the midline joining the inion to the nasion

Fig. 2 e Localization of V1 and vermis using neuronavigation. T

coordinates confirmed by neuronavigation. In the upper panel, t

inion, the putative location of V1/V2. The lower panel shows the

vermis. For both V1/V2 and vermis, there is a good correspond

Please cite this article in press as: Cattaneo, Z., et al., CerebellarCortex (2014), http://dx.doi.org/10.1016/j.cortex.2014.01.012

(Campana, Cowey, & Walsh, 2006; Heinen, Jolij, & Lamme,

2005; Pascual-Leone & Walsh, 2001). The accuracy of this

localization method was further verified by using a neuro-

navigation system (Nexstim, Helsinki, Finland) in three sub-

jects (that did not take part in the experiment) for whom

structural MRIswere available. On the structural MRIs of these

three individuals, the scalp coordinate 1 cm below the inion

indeed corresponded to the cerebellar vermis, and the scalp

coordinate 1.5 cm above the inion corresponded indeed to V1

(see Fig. 2). A No TMS “baseline” condition and two sham TMS

conditions were also included. In sham TMS conditions, TMS

was delivered either over V1 or vermis but the coil was flipped

90 degrees leftward or rightward (counterbalanced within

participants), in line with previous studies (Fried, Elkin-

Frankston, Rushmore, Hilgetag, & Valero-Cabre, 2011;

Lisanby, Gutman, Luber, Schroeder, & Sackeim, 2001). In the

main experiment, on each trial, we applied triple-pulse TMS

(20 Hz; i.e., pulse gap of 50 msec) at the onset of the motion

stimulus. Before starting the experiment, triple-pulse TMS

(20 Hz) at 100% of the rMTwas delivered over V1 ten times (ISI:

3000 msec) to assess for phosphene perception with eyes

open: none of the participants reported phosphenes

perception.

In Experiment 2, the stimulation parameters were adapted

from Experiment 1. TMS was applied to the right cerebellar

hemisphere (right Cb) or the left cerebellar hemisphere (left

Cb); in addition, a No TMS “baseline” condition was conduct-

ed. In line with previous studies, the cerebellar hemispheres

were localized at a point 1 cm inferior to the inion on a line

joining the inion to the nasion and 3 cm laterally either to the

left or to the right (Arasanz, Staines, & Schweizer, 2012;

Theoret et al., 2001), with the handle pointing superiorly. As

for Experiment 1, the precision of this localization method

he accuracy of localization method based on scalp

he pointer is indicating the scalp position 1.5 cm above the

point localized 1 cm below the inion, the putative cerebellar

ence between the scalp coordinate and the anatomy.

vermis plays a causal role in visual motion discrimination,

Fig. 3 e The effect of TMS on cerebellar vermis in the

motion discrimination task. (A) Mean percentage accuracy

and (B) mean reaction times for correct responses as a

function of TMS condition (No TMS, Sham TMS, V1, vermis)

in Experiment 1. The asterisks indicate that TMS over both

V1 and vermis selectively impaired participants’ accuracy,

whereas it had no effect on reaction times. Error bars

represent ±1 SEM.

c o r t e x x x x ( 2 0 1 4 ) 1e94

was confirmed by using neuronavigation on the structural

MRI of three individuals (that did not take part in the experi-

ment). A sham TMS condition was also included: in half of the

sham trials (one block) the coil was held over the right Cb, in

the other half over the left Cb. For sham stimulation, the coil

was flipped 90 degrees leftward or rightward (counter-

balanced within participants). Mean intensity of stimulation

was 51.9% (SD: 5.8%), corresponding to the participants’ mean

100% rMT. None of the participants reported phosphenes

perception during the experiment.

2.1.5. ProcedureParticipants sat comfortably at a distance of 57 cm from the

screen with their head resting on an adjustable chin-rest. At

the beginning of the first session, we localized each partici-

pant’s V1 ad vermis (Experiment 1) and cerebellar hemi-

spheres (Experiment 2). In both experiments, initially,

participants were given practice blocks with the motion

discrimination task. Theywere then thresholded to determine

a level of accuracy around 75% (see above). Once the visual

motion parameters were decided, each participant underwent

two blocks of stimulation for each TMS condition, for a total of

ten blocks in Experiment 1 and eight blocks in Experiment 2.

The first half of the experiment contained one block of each of

the TMS conditions; the order of these was randomly

assigned. The order of the second half of the experiment was

the reverse of the first half. Each block consisted of 60 trials (30

with leftward and 30 with rightward motion) presented in

random order. The software E-prime 2.0 (Psychology Software

Tools, Pittsburgh, PA) was used for stimuli presentation, data

collection and TMS triggering. Each experiment took approx-

imately 2 h.

2.2. Results

Mean accuracy and mean reaction times (RT) for correct re-

sponses were computed for each TMS condition for each

participant. Trials inwhich individual response latencieswere

beyond 3 standard deviations with respect to each partici-

pant’s mean performance in each experimental block were

excluded from the analyses. Data were submitted to repeated-

measures ANOVAs (significant threshold set at p ¼ .05,

Bonferroni-Holm correction applied to post-hoc comparisons)

with TMS condition as within-subjects variable.

2.2.1. The effect of cerebellar vermis and V1 stimulationIn a first group of 12 subjects (Experiment 1), TMS was

delivered to the posterior cerebellum. A total of .93% of the

trials were excluded as outliers (see above for exclusion

criteria). Planned pairwise comparisons indicated that the

two sham conditions were not significantly different from

each other neither in terms of accuracy, t(11) ¼ 1.24, p ¼ .24,

nor in terms of response latencies, t(11) ¼ .19, p ¼ .85. In the

following analyses, the mean of the two sham conditions

was therefore considered as a unique control sham TMS

condition.

Fig. 3a shows mean participants’ accuracy in each of the

different TMS conditions. A repeated-measures ANOVA with

TMS as within-subjects variable (No TMS, Sham, V1, and

vermis) on mean accuracy scores revealed a significant effect

Please cite this article in press as: Cattaneo, Z., et al., CerebellarCortex (2014), http://dx.doi.org/10.1016/j.cortex.2014.01.012

of TMS, F(3,33) ¼ 9.38, p < .001, ƞp2 ¼ .46. Pairwise comparisons

(BonferronieHolm correction applied) indicated that TMS over

V1 significantly reduced accuracy compared to both the Sham

TMS condition, t(11) ¼ 4.64, p ¼ .005, and the baseline No TMS

condition, t(11) ¼ 4.18, p ¼ .008. Similarly, TMS over vermis

significantly reduced accuracy compared to both the Sham

TMS condition, t(11) ¼ 2.99, p ¼ .036, and the No TMS condi-

tion, t(11) ¼ 2.72, p ¼ .04. The Sham TMS condition and the No

TMS conditions did not significantly differ from each other,

t(11) ¼ 1.04, p ¼ .32.

Fig. 3b showsmean RT for correct responses in the no TMS,

Sham, V1 and vermis stimulation conditions. The ANOVA

with TMS condition (No TMS, Sham, V1, vermis) as within-

subjects variable showed no significant effect of TMS,

F(3,33) ¼ .580, p ¼ .81, ƞp2 ¼ .01.

Overall, these results indicate a significant role of the

cerebellar vermis in controlling motion detection.

2.2.2. The effect of cerebellar hemispheres stimulationIn a second group of 12 subjects (Experiment 2), TMS was

delivered to the lateral cerebellum. A total of 1.4% of the trials

were overall excluded as outliers. Fig. 4a depicts mean par-

ticipants’ accuracy in each of the different TMS conditions. A

repeated-measures ANOVA with TMS as within-subjects

variable (No TMS, Sham, left Cb, right Cb) on mean accuracy

vermis plays a causal role in visual motion discrimination,

Fig. 4 e The effect of TMS on cerebellar hemispheres in the

motion discrimination task. (A) Mean percentage accuracy

and (B) mean reaction times for correct responses as a

function of TMS condition (No TMS, Sham TMS, left Cb,

right Cb) in Experiment 2. TMS had no effect on either

accuracy or RT. Error bars represent ±1 SEM.

c o r t e x x x x ( 2 0 1 4 ) 1e9 5

scores showed no significant effect of TMS, F(3,33) ¼ .95,

p ¼ .95, ƞp2 ¼ .01.

Fig. 4b depicts mean RT for correct responses in the

different experimental conditions. The ANOVA with TMS

condition (No TMS, Sham, left Cb, right Cb) as within-subjects

variable revealed no significant effect of TMS, F(3,33) ¼ 1.07,

p ¼ .38, ƞp2 ¼ .09.

Overall, results of Experiment 2 indicate that the cerebellar

hemispheres are not causally involved in processing visual

motion detection.

3. Experiment 3

Findings of Experiment 1 point to a causal role of the cere-

bellar vermis in mediating visual motion discrimination.

However, the detrimental effects of vermis TMS on perfor-

mance observed in Experiment 1 may have depended on

nonspecific effects of vermis TMS (e.g., tapping sensation,

auditory distraction). Moreover, one may argue that the ef-

fects we reported in Experiment 1 depended on spread of

stimulation effects from the cerebellar vermis to the adjacent

visual cortex. In order to rule out these possibilities, in

Experiment 3 TMS was applied during a visual orientation

discrimination task which is known to causally involve the

early visual cortex but not cerebellum (e.g., Neary et al., 2005).

In this experiment, participants had to decide about the

orientation of a series of Gabor patches (typically used to

Please cite this article in press as: Cattaneo, Z., et al., CerebellarCortex (2014), http://dx.doi.org/10.1016/j.cortex.2014.01.012

assess orientation sensitivity in V1, see Ringach, 2004), while

TMS was applied as in Experiment 1.

3.1. Method

3.1.1. ParticipantsSixteenright-handed (Oldfield, 1971) studentsof theUniversity

of Pavia (4 males, mean age: 22.9 years, SD: 3.5, range: 20e35

years) took part in the experiment. All had normal or corrected

tonormal vision.Written informed consentwasobtained from

all participants before the experiment (same inclusion criteria

were used as in Experiments 1 and 2). The protocol was

approved by the local ethical committee and participantswere

treated in accordance with the Declaration of Helsinki.

3.1.2. Visual stimuli and thresholdingThe stimuli were sinusoidal luminance-modulated gratings

(with a diameter of 5� of visual angle; generatedwithMATLAB,

The MathWorks, Natick, MA), presented foveally on a gray

background from a viewing distance of 57 cm and oriented 1,

2, 3 or 4 degrees to the left or to the right from the vertical

meridian. The spatial frequency of the gratings was 1.44 cy-

cles/� and Michelson contrast of .1. Each trial started with a

central fixation cross with a duration changing randomly be-

tween 500 and 600msec, followed by the Gabor patch that was

presented for 13 msec. The Gabor patch was immediately

replaced by a stimulus mask consisting of a black circle (same

diameter as the Gabor patch) that was presented for 100msec.

Participants’ response was followed by the presentation of a

2500 msec grey screen background. An example of experi-

mental trial is shown in Fig. 5.

3.1.3. TMSThestimulationparameterswere the sameas inExperiments1

and 2. TMS was applied over the same cortical sites as in

Experiment 1, for both real andshamTMS.ThreepulsesofTMS

(20 Hz) were delivered at the 100% of the rMT (see Methods of

Experiments 1 and 2 for rMT assessment; mean intensity:

54.0%, SD: 6.1%) over the primary visual area (V1) or over the

cerebellar vermis at the offset of the Gabor patch. Before

starting the experiment, triple-pulse TMS (20Hz) at 100%of the

rMTwas delivered overV1 ten times (ISI:minimum3000msec)

to assess forphospheneperceptionwitheyesopen:noneof the

participants reported phosphenes perception.

3.1.4. ProcedureParticipants sat comfortably at a distance of 57 cm from a 1700

TFT-LCD computer monitor (screen resolution: 800*600 pixels;

refresh rate: 75 Hz) with their heads resting on an adjustable

chin-rest. Participants were asked to judge on each trial

whether the grating was oriented rightward or leftward

compared to the vertical meridian, by rightward/leftward key

presses using their right middle and index finger. Response

speed was stressed in addition to accuracy. Initially, partici-

pants were given practice blocks with the orientation

discrimination task. Hence, each participant underwent eight

blocks of stimulation (two for each TMS condition). The first

half of the experiment (i.e., the first four blocks) contained one

block of each of the four TMS conditions; the order of these

was randomly assigned. The order of the second half of the

vermis plays a causal role in visual motion discrimination,

Fig. 6 e The effect of TMS on V1 in the orientation

discrimination task. (A) Mean percentage accuracy and (B)

mean reaction times for correct responses as a function of

TMS condition (Sham TMS, V1, vermis) in Experiment 3.

The asterisks indicate that TMS over V1, but not over the

vermis, impaired participants’ accuracy, whereas it had no

effect on reaction times. Error bars represent ±1 SEM.

Fig. 5 e The timeline of an experimental trial in the orientation discrimination task. In Experiment 3 TMS was applied over

the cerebellar vermis and the primary visual area (V1). Sham TMS (control condition) was also applied over the targeted

sites.

c o r t e x x x x ( 2 0 1 4 ) 1e96

Please cite this article in press as: Cattaneo, Z., et al., CerebellarCortex (2014), http://dx.doi.org/10.1016/j.cortex.2014.01.012

experiment was the reverse of the first half. Each block con-

sisted of 40 trials (5 leftward and 5 rightward orientated Gabor

for each of the four possible orientation degrees) presented in

random order, for a total of 80 trials for each stimulation

condition. The software E-prime 2.0 (Psychology Software

Tools, Pittsburgh, PA) was used for stimuli presentation, data

collection and TMS triggering. The whole experiment took

approximately 90 min.

4. Results

Mean accuracy and mean RT for correct responses were

computed for each TMS condition for each participant. Trials

in which individual response latencies were beyond 3 stan-

dard deviations with respect to each participant’s mean per-

formance in each experimental block were excluded from the

analyses. A total of 1.4% of the trials were excluded as outliers

(see above for exclusion criteria).

Planned pairwise comparisons indicated that the two

sham conditions were not significantly different from each

other neither in terms of accuracy, t(15) ¼ .32, p ¼ .76, nor in

terms of response latencies (mean RT ¼ 416 msec and

402 msec for sham V1 and for sham vermis, respectively),

t(15) ¼ 1.74, p ¼ .10. In the following analyses, the mean of the

two sham conditions was therefore considered as a unique

control sham TMS condition.

Fig. 6a shows mean participants’ accuracy in each of the

different TMS conditions. A repeated-measures ANOVA with

TMS as within-subjects variable (Sham, V1, and vermis) on

mean accuracy scores revealed a significant effect of TMS,

F(2,30) ¼ 5.95, p ¼ .007, ƞp2 ¼ .28.

Pairwise comparisons (BonferronieHolm correction

applied) indicated that TMS over V1 significantly reduced

vermis plays a causal role in visual motion discrimination,

c o r t e x x x x ( 2 0 1 4 ) 1e9 7

accuracy compared to both the Sham TMS condition,

t(15) ¼ 2.90 p ¼ .033, and the vermis condition, t(15) ¼ 2.56,

p ¼ .044. Accuracy did not differ when TMS was applied over

vermis compared to the ShamTMS condition, t(15)< 1, p¼ .93.

Fig. 6b shows mean RT for correct responses in the Sham,

V1 and vermis stimulation conditions. The ANOVA with TMS

condition (Sham, V1, vermis) as within-subjects variable

showed no significant effect of TMS, F(2,30) ¼ .25, p ¼ .78,

ƞp2 ¼ .02 on participants’ response latencies.

Overall, results of Experiment 3 indicate that the cerebellar

vermis is not involved in orientation discrimination.

5. Discussion

Online TMS applied over the cerebellar vermis but not over the

cerebellar hemispheres significantly impaired participants’

performance in a visual motion discrimination task (Experi-

ments 1 and 2). The detrimental effects of vermis TMS on

motion discrimination were similar to those observed when

V1 was stimulated. In Experiment 3, TMS over vermis did not

affect participants’ accuracy in an orientation discrimination

task, whereas TMS over V1 did induce an impairment. The

results of Experiment 3 thus demonstrate that the effect of

vermis TMS on motion discrimination cannot be explained in

terms of nonspecific effects of TMS (such as discomfort), as no

effects were observed on the detection of another type of vi-

sual stimuli (gratings). Furthermore, they also rule out the

possibility that the effect of vermis TMS in Experiment 1 was

due to the spread of the TMS effect to V1, rather than due to a

role of vermis in motion processing. Our data shed light on

previous neuroimaging and neurophysiological evidence that

have reported cerebellar activations during visual motion

processing but have been inconsistent regardingwhether only

midline cerebellar structures or also cerebellar hemispheres

were involved (Barbur et al., 1993; Baumann & Mattingley,

2010; Dupont et al., 1994; Ivry & Diener, 1991; Jokisch et al.,

2005; Nawrot & Rizzo, 1995, 1998; Thier et al., 1999). Previous

patient and nonhuman primate evidence have highlighted the

role of midline cerebellar structures in mediating motion

discrimination (Ignashchenkova et al., 2009; Nawrot & Rizzo,

1995, 1998): our data extend this evidence by showing that

direct stimulation of the cerebellar vermis region affects an

ongoing visual process in healthy humans.

TMS also affected motion discrimination when delivered

overV1 at theonsetof the visual stimulus, according toprevious

evidence (Laycock, Crewther, Fitzgerald, & Crewther, 2007;

Nawrot & Rizzo, 1995, 1998; Silvanto, Cowey, Lavie, & Walsh,

2005). These findings are in line with the tenet that motion

processing ismediatedbya feedforward/feedback loopbetween

striate and extrastriate cortex (Bullier, Hupe, James, & Girard,

2001; Galletti & Fattori, 2003). In our study, TMS applied over

vermis at the onset of the motion stimulus also disrupted mo-

tion discrimination. How does the cerebellum take part to the

network subtendingmotiondiscrimination?UsingMEG,Handel

et al. (Handel, Thier, & Haarmeier, 2009) recorded cortical re-

sponses in a group of patients with cerebellar lesions while

viewing motion stimuli of varied coherence. Interestingly, the

impairment in global motion discrimination shown by the pa-

tients was paralleled by qualitative differences in responses

Please cite this article in press as: Cattaneo, Z., et al., CerebellarCortex (2014), http://dx.doi.org/10.1016/j.cortex.2014.01.012

recorded fromparieto-temporal cortex, in the formof a reduced

responsiveness to coherent visual motion and a loss of bilateral

representations of motion coherence. These data suggest that

visualmotionprocessing incerebral cortex criticallydependson

an intact cerebellum (Handel et al., 2009). In line with this, a

number of recent fMRI studies have reported functional con-

nectivity between the cerebellar vermis and the visual network

(Kellerman et al., 2012; O’Reilly, Beckmann, Tomassini,

Ramnani, & Johansen-Berg, 2010; Sang et al., 2012). For

instance, Kellermann et al. (2012) showed modulation of cere-

bellar inputs over a large portion of the dorsal visual stream,

including V1, V5 and the posterior parietal cortex (PPC), in an

attention-demanding task requiring participants to attend to

slight changes in the velocity of vertical bars. Psychophysio-

logical analyses revealed enhanced connectivity between cere-

bellum and V5 during the task, whereas connections between

V5 and PPC were suppressed. Kellermann et al. (2012) argued

that their findings suggest a modulatory role of cerebellum in

predictive coding (Kellermann et al., 2012; Schlerf, Ivry, &

Diedrichsen, 2012). In this context, the cerebellum activation

would reflect top-down predictions about motion characteris-

tics, whereas the PPC suppression would reflect diminished

sensitivity to bottom up information.

In line with this, we may speculate that the impairment in

motion detection following TMS over the vermis depended on

TMS interfering with those sensory predictions mechanisms

handled by this cerebellar region that are needed for oculo-

motor control. The role of cerebellar vermis in oculo-motor

control is well established. Indeed, TMS over the posterior

cerebellum delivered in close relationship with movement

interferes with smooth pursuit (Ohtsuka & Enoki, 1998),

visually guided saccades (Hashimoto & Ohtsuka, 1995) and

with the synkinesis of eye and head coordination (Nagel &

Zangemeister, 2003). Moreover, theta-burst stimulation (TBS)

over the posterior cerebellum has long-term effects on

saccadic adaptation (Jenkinson & Miall, 2010), internal repre-

sentations of motor information related to eye movements

(Colnaghi et al., 2011), and classical eye-blink conditioning

(Hoffland et al., 2012). Following from this, one may argue

though that the detrimental effects of vermis TMS we re-

ported depended on vermis TMS directly interfering with eye

movements control during motion discrimination (e.g.,

Colnaghi et al., 2011; Ohtsuka & Enoki, 1998). However, it is

important to consider that the average latency of a saccade

has been reported to be around 200 msec (Carpenter, 1988),

beyond the duration of our motion stimuli (that lasted be-

tween 40 and 67 msec). Although microsaccades may have

indeed been affected by cerebellar TMS, the main function of

microsaccades is to restoring faded vision during fixation, for

both foveal and peripheral targets (see Martinez-Conde,

Otero-Millan, & Macknik, 2013, for a review). Indeed, fading

only occurs with longer lasting stimuli compared to ours, and

microsaccades usually occur at a frequency of around 1 Hz

(Martinez-Conde et al., 2013), whereas our motion stimuli last

less than 1 sec. In light of the above, it is unlikely that the

effects we reported merely reflected TMS effects on eye

movements control. Moreover, neuroimaging, primate

neurophysiology and TMS evidence converges in indicating

that the cerebellar hemispheres are also involved in oculo-

motor control (e.g., Haarmeier & Kammer, 2010; Liem, Frens,

vermis plays a causal role in visual motion discrimination,

c o r t e x x x x ( 2 0 1 4 ) 1e98

Smits, & van der Geest, 2013; Nitschke, Arp, Stavrou,

Erdmann, & Heide, 2005; Ohki et al., 2008; Panouilleres et al.,

2012). In light of this evidence, if the decrease in perfor-

mance we observed depended on vermis TMS affecting eye

movements, similar detrimental effects should have been

observed following TMS over the cerebellar hemispheres,

whereas this was not the case.

Several findings point to a role of the cerebellar hemi-

spheres in visual discrimination (Claeys et al., 2003; Orban,

Dupont, Vogels, Bormans, & Mortelmans, 1997) and in sen-

sory prediction mechanisms related to arm and eyes’ motion

(Cerminara, Apps, &Marple-Horvat, 2009; Liu et al., 2000). Also,

neuroimaging data suggest that the lateral cerebellummay be

more driven by the requirements for (tactile) sensory pro-

cessing than bymotor coordination per se (Gao et al., 1996; Liu

et al., 2000). In this perspective, the lack of effect of TMS over

the cerebellar hemispheres is not completely clear. Our data

suggest that the vermis plays a more prominent role in visual

sensoryprocessing (beyondsensorypredictionmechanismsor

visual attention) compared to the cerebellar hemispheres,

although further investigation is needed to clarify this issue.

It might be argued that vermis TMS impaired motion

perception not because this region contributes to motion

discrimination, but because online vermis TMS modulated

activity in striate cortex due to spread of activation (see also

Renzi, Vecchi, D’Angelo, Silvanto, & Cattaneo, 2014). However,

this explanation is inconsistent with the results of Experiment

3, in which stimulation of the cerebellar vermis did not impair

orientation discrimination whereas TMS over V1 did.

Although the impact of vermis TMS may spread to anatomi-

cally connected regions (this likely depending on intensity and

frequency of stimulation), such a spread of activation was not

sufficient to modulate behavior in Experiment 3, supporting

the view that the results in Experiment 1 do reflect the role of

vermis in motion encoding. Moreover, the TMS effects over

vermis found in Experiment 1 are unlikely to result from an

inaccurate localization procedure, since its accuracy was

confirmed by neuronavigation in individuals for whom

structural MRI was available. Finally, any discomfort induced

by TMS is unlikely to have driven these effects, since cere-

bellar vermis and cerebellar hemispheres stimulation pro-

duced similar level of discomfort and because the effects of

discomfort should have also been present in Experiment 3.

In conclusion, the effect of TMS on visual motion reported

here points to a causal role of the cerebellar vermis in visual

motion discrimination. In a broader perspective, the involve-

ment of the cerebellar vermis in visual motion discrimination

could be for motor sake and be related to the complex ar-

rangements that the cerebellum imposes to various brain

structures in order to plan, regulate and coordinate learned

movements (Bower, 1997; Manto et al., 2012).

r e f e r e n c e s

Arasanz, C. P., Staines, W. R., & Schweizer, T. (2012). Isolating acerebellar contribution to rapid visual attention usingtranscranial magnetic stimulation. Frontiers in BehavioralNeuroscience, 6, 55.

Please cite this article in press as: Cattaneo, Z., et al., CerebellarCortex (2014), http://dx.doi.org/10.1016/j.cortex.2014.01.012

Barbur, J. L., Watson, J. D., Frackowiak, R. S., & Zeki, S. (1993).Conscious visual perception without V1. Brain, 116, 1293e1302.

Baumann, O., & Mattingley, J. B. (2010). Scaling of neuralresponses to visual and auditory motion in the humancerebellum. Journal of Neuroscience, 30, 4489e4495.

Bower, J. M. (1997). Is the cerebellum sensory for motor’s sake, ormotor for sensory’s sake: the view from the whiskers of a rat?Progress in Brain Research, 114, 463e496.

Bullier, J., Hupe, J., James, A., & Girard, P. (2001). The role offeedback connections in shaping the responses of visualcortical neurons. Progress in Brain Research, 134, 193e204.

Campana, G., Cowey, A., & Walsh, V. (2006). Visual area V5/MTremembers “what” but not “where”. Cerebral Cortex, 16,1766e1770.

Carpenter, R. H. S. (1988). Movements of the eyes (2nd ed.). London:Pion.

Cerminara, N. L., Apps, R., & Marple-Horvat, D. E. (2009). Aninternal model of a moving visual target in the lateralcerebellum. Journal of Physiology, 587, 429e442.

Claeys, K. G., Orban, G. A., Dupont, P., Sunaert, S., Hecke, P. Van, &Schutter, E. De (2003). Involvement of multiple functionallydistinct cerebellar regions in visual discrimination: a humanfunctional imaging study. NeuroImage, 20, 840e854.

Colnaghi, S., Ramat, S., D’Angelo, E., Cortese, A., Beltrami, G.,Moglia, A., et al. (2011). Theta-Burst stimulation of thecerebellum interferes with internal representations ofsensory-motor information related to eye movements inhumans. Cerebellum, 10, 711e719.

Dupont, N., Orban, G. A., De Bruyn, B., Verbruggen, A., &Mortelmans, L. (1994). Many areas in the human brain respondto visual motion. Journal of Neurophysiology, 72, 1420e1424.

D’Angelo, E., & Casali, S. (2012). Seeking a unified framework forcerebellar function and dysfunction: from circuit operations tocognition. Frontiers in Neural Circuits, 6, 116.

Fried, P. J., Elkin-Frankston, S., Rushmore, R. J., Hilgetag, C. C., &Valero-Cabre, A. (2011). Characterization of visual perceptsevoked by noninvasive stimulation of the human posteriorparietal cortex. PloS One, 6, e27204.

Galletti, C., & Fattori, P. (2003). Neuronal mechanisms fordetection of motion in the field of view. Neuropsychologia, 41,1717e1727.

Gao, J., Parsons, L., Bower, J., X, J., Li, J., & Fox, P. (1996). Cerebellumimplicated in sensory acquisition and discrimination ratherthan motor control. Science, 272, 545e547.

Haarmeier, T., & Kammer, T. (2010). Effect of TMS on oculomotorbehavior but not perceptual stability during smooth pursuiteye movements. Cerebral Cortex, 20, 2234e2243.

Handel, B., Thier, P., & Haarmeier, T. (2009). Visual motionperception deficits due to cerebellar lesions are paralleled byspecific changes in cerebro-cortical activity. Journal ofNeuroscience, 29, 15126e15133.

Hashimoto, M., & Ohtsuka, K. (1995). Transcranial magneticstimulation over the posterior cerebellum during visuallyguided saccades in man. Brain, 118, 1185e1193.

Heinen, K. C., Jolij, J., & Lamme, V. A. F. (2005). Figure-groundsegregation requires two distinct periods of activity in V1: atranscranial magnetic stimulation study. NeuroReport, 16,1483e1487.

Hoffland, B. S., Bologna, M., Kassavetis, P., Teo, J. T. H.,Rothwell, J. C., Yeo, C. H., et al. (2012). Cerebellar theta burststimulation impairs eyeblink classical conditioning. Journal ofPhysiology, 590, 887e897.

Ignashchenkova, a, Dash, S., Dicke, P. W., Haarmeier, T.,Glickstein, M., & Thier, P. (2009). Normal spatial attention butimpaired saccades and visual motion perception after lesions ofthe monkey cerebellum. Journal of Neurophysiology, 102,3156e3168.

vermis plays a causal role in visual motion discrimination,

c o r t e x x x x ( 2 0 1 4 ) 1e9 9

Ivry, R. B., & Diener, H. C. (1991). Impaired velocity perception inpatients with lesions of the cerebellum. Journal of CognitiveNeuroscience, 3, 355e366.

Jenkinson, N., & Miall, R. C. (2010). Disruption of saccadicadaptation with repetitive transcranial magnetic stimulationof the posterior cerebellum in humans. Cerebellum, 9, 548e555.

Jokisch, D., Troje, N. F., Koch, B., Schwarz, M., & Daum, I. (2005).Differential involvement of the cerebellum in biological andcoherent motion perception. European Journal of Neuroscience,21, 3439e3446.

Kellermann, T., Regenbogen, C., De Vos, M., Moßnang, C.,Finkelmeyer, A., & Habel, U. (2012). Effective connectivity ofthe human cerebellum during visual attention. Journal ofNeuroscience, 32, 11453e11460.

Laycock, R., Crewther, D. P., Fitzgerald, P. B., & Crewther, S. G.(2007). Evidence for fast signals and later processing in humanV1/V2 and V5/MTþ: a TMS study of motion perception. Journalof Neurophysiology, 98, 1253e1262.

Lee, K.-H., Egleston, P. N., Brown, W. H., Gregory, A. N.,Barker, A. T., & Woodruff, P. W. R. (2007). The role of thecerebellum in subsecond time perception: evidence fromrepetitive transcranial magnetic stimulation. Journal ofCognitive Neuroscience, 19, 147e157.

Liem, E. I., Frens, M. A., Smits, M., & van der Geest, J. N. (2013).Cerebellar activation related to saccadic inaccuracies.Cerebellum, 12, 224e235.

Lisanby, S. H., Gutman, D., Luber, B., Schroeder, C., &Sackeim, H. A. (2001). Sham TMS: intracerebral measurementof the induced electrical field and the induction of motor-evoked potentials. Biological Psychiatry, 49, 460e463.

Liu, Y., Pu, Y., Gao, J. H., Parsons, L. M., Xiong, J., Liotti, M., et al.(2000). The human red nucleus and lateral cerebellum insupporting roles for sensory information processing. HumanBrain Mapping, 10, 147e159.

Manto, M., Bower, J. M., Conforto, A. B., Delgado-Garcıa, J. M., DaGuarda, S. N. F., Gerwig, M., et al. (2012). Consensus paper:roles of the cerebellum in motor controlethe diversity of ideason cerebellar involvement in movement. Cerebellum, 11,457e487.

Martinez-Conde, S., Otero-Millan, J., & Macknik, S. L. (2013). Theimpact of microsaccades on vision: towards a unified theory ofsaccadic function. Nature Reviews Neuroscience, 14, 83e96.

Nagel, M., & Zangemeister, W. (2003). The effect of transcranialmagnetic stimulation over the cerebellum on the synkinesis ofcoordinated eye and head movements. Journal of NeurologicalScience, 213, 35e45.

Nawrot, M., & Rizzo, M. (1995). Motion perception deficits frommidline cerebellar lesions in human. Vision Research, 35,723e731.

Nawrot, M., & Rizzo, M. (1998). Chronic motion perception deficitsfrom midline cerebellar lesions in human. Vision Research, 38,2219e2224.

Neary, K., Anand, S., & Hotson, J. R. (2005). Perceptual learning ofline orientation modifies the effects of transcranial magneticstimulation of visual cortex. Experimental Brain Research, 162,23e34.

Nitschke, M. F., Arp, T., Stavrou, G., Erdmann, C., & Heide, W.(2005). The cerebellum in the cerebro-cerebellar network forthe control of eye and hand movements d an fMRI study.Progress in Brain Research, 148, 151e164.

O’Reilly, J. X., Beckmann, C. F., Tomassini, V., Ramnani, N., &Johansen-Berg, H. (2010). Distinct and overlapping functional

Please cite this article in press as: Cattaneo, Z., et al., CerebellarCortex (2014), http://dx.doi.org/10.1016/j.cortex.2014.01.012

zones in the cerebellum defined by resting state functionalconnectivity. Cerebral Cortex, 20, 953e965.

Ohki, M., Kitazawa, H., Hiramatsu, T., Kaga, K., Kitamura, T.,Yamada, J., et al. (2008). Role of primate cerebellar hemispherein voluntary eye movement control revealed by lesion effects.Journal of Neurophysiology, 101, 934e947.

Ohtsuka, K., & Enoki, T. (1998). Transcranial magnetic stimulationover the posterior cerebellum during smooth pursuit eyemovements in man. Brain, 121, 429e435.

Oldfield, R. C. (1971). The assessment and analysis of handedness:the Edinburgh inventory. Neuropsychologia, 1, 97e113.

Orban, G. A., Dupont, P., Vogels, R., Bormans, G., & Mortelmans, L.(1997). Human brain activity related to orientationdiscrimination tasks. European Journal of Neuroscience, 9,246e259.

Panouilleres, M., Neggers, S. F., Gutteling, T. P., Salemme, R., vander Stigchel, S., van der Geest, J. N., et al. (2012). Transcranialmagnetic stimulation and motor plasticity in human lateralcerebellum: dual effect on saccadic adaptation. Human BrainMapping, 33, 1512e1525.

Pascual-Leone, A., & Walsh, V. (2001). Fast backprojections fromthe motion to the primary visual area necessary for visualawareness. Science, 292, 510e512.

Renzi, C., Vecchi, T., D’Angelo, E., Silvanto, J., & Cattaneo, Z.(2014). Phosphene induction by cerebellar transcranialmagnetic stimulation. Clinical Neurophysiology.

Ringach, D. L. (2004). Mapping receptive fields in primary visualcortex. Journal of Physiology, 558, 717e728.

Rossi, S., Hallett, M., Rossini, P. M., & Pascual-Leone, A. (2011).Screening questionnaire before TMS: an update. ClinicalNeurophysiology, 122, 1686.

Rossini, P., Barker, A., Berardelli, A., Caramia, M., Caruso, G.,Cracco, R., et al. (1994). Non-invasive electrical and magneticstimulation of the brain, spinal cord and roots: basic principlesand procedures for routine clinical application. Report of anIFCN committee. Electroencephalography and ClinicalNeurophysiology, 91, 79e92.

Sang, L., Qin, W., Liu, Y., Han, W., Zhang, Y., Jiang, T., et al. (2012).Resting-state functional connectivity of the vermal andhemispheric subregions of the cerebellum with both thecerebral cortical networks and subcortical structures.NeuroImage, 61, 1213e1225.

Schlerf, J., Ivry, R. B., & Diedrichsen, J. (2012). Encoding of sensoryprediction errors in the human cerebellum. Journal ofNeuroscience, 32, 4913e4922.

Schutter, D. J. L. G., De Weijer, A. D., Meuwese, J. D. I., Morgan, B.,& Van Honk, J. (2008). Interrelations between motivationalstance, cortical excitability, and the frontalelectroencephalogram asymmetry of emotion: a transcranialmagnetic stimulation study. Human Brain Mapping, 29,574e580.

Silvanto, J., Cowey, A., Lavie, N., & Walsh, V. (2005). Striate cortex(V1) activity gates awareness of motion. Nature Neuroscience, 8,143e144.

Theoret, H., Haque, J., & Pascual-leone, A. (2001). Increasedvariability of paced finger tapping accuracy followingrepetitive magnetic stimulation of the cerebellum in humans.Neuroscience Letters, 306, 304e307.

Thier, P., Haarmeier, T., Treue, S., & Barash, S. (1999). Absence of acommon functional denominator of visual disturbances incerebellar disease. Brain, 122, 2133e2146.

vermis plays a causal role in visual motion discrimination,