Basal ganglia and supplementary motor area subtend durationperception: an fMRI study

A.M. Ferrandez,a,* L. Hugueville,a S. Lehericy,b,c J.B. Poline,c C. Marsault,b and V. Pouthasa

a Unite de Neurosciences Cognitives et Imagerie Cerebrale, Hopital de la Salpetriere, Paris, Franceb Service de Neuroradiologie, Hopital de la Salpetriere, Paris, France

c Service Hospitalier Frederic Joliot, Orsay, France

Received 18 June 2002; revised 13 February 2003; accepted 3 March 2003

Abstract

Brain imaging studies on duration perception usually report the activation of a network that includes the frontal and mesiofrontal cortex(supplementary motor area, SMA), parietal cortex, and subcortical areas (basal ganglia, thalamus, and cerebellum). To address the questionof the specific involvement of these structures in temporal processing, we contrasted two visual discrimination tasks in which the relevantstimulus dimension was either its intensity or its duration. Eleven adults had to indicate (by pressing one of two keys) whether they thoughtthe duration or the intensity of a light (LED) was equal to (right hand) or different from (left hand) that of a previously presented standard.In a control task, subjects had to press one of the two keys at random. A similar broad network was observed in both the duration-minus-control and intensity-minus-control comparisons. The intensity-minus-duration comparison pointed out activation in areas known toparticipate in cognitive operations on visual stimuli: right occipital gyrus, fusiform gyri, hippocampus, precuneus, and intraparietal sulcus.In contrast, the duration-minus-intensity comparison indicated activation of a complex network that included the basal ganglia, SMA,ventrolateral prefrontal cortex, inferior parietal cortex, and temporal cortex. These structures form several subnetworks, each possibly incharge of specific time-coding operations in humans. The SMA and basal ganglia may be implicated in the time-keeping mechanism, andthe frontal-parietal areas may be involved in the attentional and mnemonic operations required for encoding and retrieving durationinformation.© 2003 Elsevier Science (USA). All rights reserved.

Introduction

Duration estimation is one of our daily life activities,although we are most often unaware of it. For example,when we drive, we know the duration of the traffic lights inour neighborhood and anticipate the right moment to startmoving again. Some authors (Church, 1984; Gibbon et al.,1984) have proposed a model of temporal information pro-cessing based on animal studies. Three stages are distin-guished; i.e., a clock mechanism measures duration (via anoscillatory pacemaker and an accumulator), a memory

mechanism compares the current duration with the durationstored in reference memory, and a decision mechanismselects the appropriate response. These different stages mayinvolve separate brain regions and be modulated by differ-ent neurotransmitters (Meck, 1996). The internal clock ap-pears to be linked to dopamine function in the basal ganglia,while temporal memory and attentional mechanisms appearto be linked to acetylcholine function in the frontal cortex.Structures that subtend these functions are connected to-gether by frontostriatal loops that perform the timing se-quence required for duration discrimination.

In humans, numerous studies have demonstrated the reg-ulatory role of dopamine in motor function. According toMarsden (1984), “the sequencing of motor action and thesequencing of thought could be a uniform function carriedout by the basal ganglia.” Harrington and her team (Har-rington, et al., 1998) found that Parkinson’s disease (PD)

* Corresponding author. Anne-Marie Ferrandez, Unite´ de Neuro-sciences Cognitives et Imagerie Ce´rebrale, CNRS UPR 640, LENA, Hoˆ-pital de la Salpeˆtriere, 75651, Paris Cedex 13, France. Fax:�33-1-45-86-25-37.

E-mail address: Anne-Marie.Ferrandez@chups.jussieu.fr (A.M. Fer-randez).

NeuroImage 19 (2003) 1532–1544 www.elsevier.com/locate/ynimg

1053-8119/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved.doi:10.1016/S1053-8119(03)00159-9

patients were impaired when they had to reproduce theinterresponse interval between two tones. Patients were alsoimpaired in their ability to time when they had to judge therelative duration of tones separated by either 300 or 600 ms.In the range of seconds, another study demonstrated that PDpatients were impaired in their ability to reproduce andestimate time intervals (Pastor et al., 1992). These studieson PD provide evidence that the basal ganglia are involvedin both the motor and the perceptual aspects of the timingfunction.

The basal ganglia are connected with frontal areas viaglutamatergic influences (Meck, 1996). In humans or pri-mates, the supplementary motor area (SMA) is one of themajor outputs of the putamen (Alexander et al., 1986). Italso sends projections into the basal ganglia (Saint-Cyr etal., 1995), and especially into the putamen (Takada et al.,1998). The main role of the SMA in duration processing hasbeen assessed in several studies (Macar et al., 2002). Basedon topographical analyses (surface Laplacians) of slow cor-tical potential changes in temporal production and judg-ments, Macar and collaborators (1999) proposed that theSMA subserves important functions in the timing of bothsensory and motor tasks. For these authors, “ the SMA couldeither provide the ‘pulse accumulation’ process, or receiveoutputs from this process through striatal efferent path-ways” (Macar et al., 1999; p. 271).

Numerous functional neuroimaging studies confirm theinvolvement of the basal ganglia and of the SMA in motoras well as in perceptual timing tasks (Macar et al., 2002).Most of these studies also report activation in subcorticalareas (thalamus and cerebellum) and in various corticalareas, including the premotor, right prefrontal, parietal, andtemporal regions. This broad corpus of literature allows usto extensively discuss our own data in the last part of thisreport.

To conclude that the perception of stimulus duration perse is responsible for the observed pattern of activation, itmust be demonstrated that this pattern is not elicited by theperception of another parameter of the stimulus. Two pre-vious studies from our group using position emission to-mography (PET) (Maquet et al., 1996) or PET combinedwith event-related potentials (ERPs) (Pouthas et al., 2000)investigated the specificity of brain regions involved in theperception of duration on intensity and duration visual-discrimination tasks. The PET data (Maquet et al., 1996)showed that the two tasks activated the same network,which included the right frontal, right parietal, and cingulatecortices, the left and right fusiform gyri, and the cerebellum.No significant differences were found between the twoconditions. In this PET study, the SMA was not included inthe scanned volume, so we were unable to provide evidenceof the involvement of the SMA in the duration processingtask. By combining the PET data from the Maquet andcollaborators’ experiment with ERPs (Maquet et al., 1996),Pouthas and collaborators (2000) were able to better de-scribe the time course of activation, and showed that two

different areas specifically subtended the processing of theintensity and duration of the same visual stimulus, that is,the cuneus for intensity, and the right frontal cortex forduration. In a recent PET study with auditory stimuli, Belinand collaborators (2002) investigated the neuroanatomicalsubstrate of sound duration discrimination. This study sup-plemented another one where the same experimental para-digm was used for sound intensity discrimination (Belin etal., 1998), which allowed the authors to assess how attend-ing to two different properties of auditory stimuli (intensityand duration) would affect the pattern of neuronal ability.The sound duration discrimination study (Belin et al., 2002)put in evidence two cerebral networks, i.e., a supramodalright frontoparietal cortical network responsible for alloca-tion of sensory attentional resources, and a network ofregions such as the basal ganglia, cerebellum, and rightprefrontal cortex more specifically involved in the temporalaspects of the discrimination task. In these two Belin andcollaborators’ studies, however, it is difficult to directlycompare the network responsible for sound duration dis-crimination (Belin et al., 2002) with the network responsiblefor sound intensity discrimination (Belin et al., 1998), be-cause these data, collected in two separate experiments, donot allow to compute specific contrasts, such as Duration-minus-Intensity comparisons.

In view of the above reported studies, several questionsremain to be elucidated. First, several areas were activatedin the study of Maquet and collaborators (1996), but thisstudy failed to detect areas more specifically associated withduration than intensity discrimination processes, becausethe Duration-minus-Intensity and Intensity-minus-Durationcomparisons did not yield significant activation. Second,several critical areas assumed to be involved in durationperception, such as the SMA and the basal ganglia, respec-tively, were not studied or did not show up.

In the present experiment, we used the same matching-to-sample task as in our previous PET study (Maquet et al.,1996). The neuroimaging technique we chose was func-tional magnetic resonance imaging (fMRI), which betterenables the detection of small and deep structures such asthe basal ganglia than the PET technique, and allows forsingle-subject analyses. We intended to better differentiatethe brain regions subtending time keeping and mnemonicmechanisms involved in duration perception as recentlydone by Rao and collaborators (2001), who contrasted timeperception to pitch perception. The main goal of the presentstudy was to assess the specific roles of the basal ganglia,SMA, and right frontal cortex in duration discrimination,compared to intensity discrimination for a visual stimulus.We hypothesized that the SMA, the basal ganglia, andprefrontal areas would be more specifically associated withduration than intensity estimation. The involvement of thesestructures in different processes involved in duration per-ception, that is, clock mechanisms, and attentional and mne-monic processes will be discussed in view of the results.

1533A.M. Ferrandez et al. / NeuroImage 19 (2003) 1532–1544

Methods

Subjects

The subjects were 11 healthy right-handed adults (5males and 6 females, mean age � SD � 24 � 3.5 years)with normal vision or vision corrected to normal by contactlenses. Subjects had no history of neurological or psychiat-ric disease. They gave their written informed consent andwere paid for their participation. The experimental protocolwas approved by the local ethics committee.

Experimental design

There were three different tasks: Duration, Intensity, andControl. Subjects underwent a total of 10 runs, i.e., five Dura-tion runs alternating with five Intensity runs (each Duration runwas followed by an Intensity run, and each Intensity run wasfollowed by a Duration run). Half of the subjects began with aDuration run, and the other half with an Intensity run. In eachrun, three 30-s activation periods (Duration task or Intensitytask) alternated with three 30-s control periods (Control task).Task switching was indicated to the subjects by a visual cue,“Duration” or “Control” (for a Duration run), and “Intensity”or “Control” (for an Intensity run). Four scans were discardedat the beginning of each run to achieve signal equilibrium.Each run lasted a total of 3 min 12 s. The total experimentlasted 60 to 90 min, including getting the subject set up andpresenting the instructions.

TasksThe Duration, Intensity, and Control tasks were the same

as those used in the Maquet and collaborators’ study (1996).In the Duration task, the subjects had to determine whetherthe length of time a green LED (five test stimuli: 490, 595,700, 805, and 910 ms) stayed lit was equal to (right buttonpress) or different from (left button press) that of a previ-ously presented standard duration (700 ms). The intensity ofeach stimulus was 15 cd/m2. The standard was presented sixtimes in succession before the beginning of the task. Theneach test stimulus (for each duration including the standardone), was presented six times in random order, making atotal of 30 stimulus presentations per run. In the Intensitytask, the subjects had to judge whether the intensity of theLED (five test stimuli: 3, 7, 15, 22, and 29 cd/m2) was equalto (right button press) or different from (left button press)that of a previously presented standard intensity (15 cd/m2).Each signal lasted 700 ms. As in the Duration task, thestandard was presented six times before the beginning of thetask, and then each test stimulus (for each intensity includ-ing the standard one) was presented six times in randomorder, making a total of 30 stimulus presentations per run. Inthe Control task, the subjects had to press one of the twokeys at random whenever the LED went off. The LED in thecontrol task lasted 700 ms and its luminance was 15 cd/m2.

In each task, the interstimulus interval (ISI) lasted between2200 and 3200 ms (on average, 2700 ms). Subjects wereinstructed not to count subvocally. At the end of the exper-iment, subjects were asked to state briefly how they pro-ceeded in performing the task.

fMRI image acquisition

The MR protocol was carried out by using a 1.5-Twhole-body system and Blood Oxygen Level Dependent(BOLD) fMRI. The subject’s head was immobilized usingfoam cushions and tape. The protocol included (1) onesagittal T1-weighted image to localize functional and ana-tomical axial slices, (2) 20 axial gradient echo-planar (EPI)images (5 mm no gap, TR 3000 ms, TE 60 ms, bandwidth125 kHz, � � 90°, FOV 240 � 240 mm2, matrix size 64 �64, in-plane resolution 3.75 mm � 3.75 mm), and (3) 110axial inversion recovery three-dimensional (3D) fast spoiledgradient (SPGR) images (1.5 mm thick, TI 400 ms, FOV240 � 240 mm2, matrix size 256 � 256) for anatomicallocalization. Images were acquired over 45 min.

fMRI image analysis

Data analysis was performed by using Statistical ParametricMapping, version 99 (SPM 99, Wellcome Department of Cog-nitive Neurology, London, UK). For each subject, anatomicalimages were transformed stereotactically to Talairach coordi-nates. The functional scans, corrected for subject motion, werethen normalized by using the same transformation, andsmoothed with a Gaussian spatial filter to a final smoothness of5 mm. The data were analyzed on an individual (subject bysubject) basis and across subjects (fixed-effect analysis). Forthe group analysis, the data from each run were approximatedby using the general linear model with separate delayed boxcarfunctions representing the hemodynamic responses of eachtask period. Overall signal differences between runs were alsomodeled. A temporal cutoff point of 240 s was applied to filtersubject-specific low-frequency drift related mostly to the sub-ject’s biological rhythms. The results were obtained by using asignificance level of P � 0.05, corrected for multiple compar-isons inside the volume of the brain. For the single-subjectanalyses, parametric maps were constructed by using a signif-icance level of P � 0.01, uncorrected for multiple compari-sons, and activated clusters were considered significant if theirspatial extent was 5 voxels, using the same contrasts as for thegroup analysis.

Results

Behavioral data

Reaction times were longer for duration than for inten-sity (741 ms and 614 ms, respectively; Student’s t test, P �0.01). Generalization gradients (“equal-to-standard” re-

1534 A.M. Ferrandez et al. / NeuroImage 19 (2003) 1532–1544

sponse for a given test stimulus) were comparable to pre-vious studies using the same two tasks (Maquet et al., 1996;Pouthas et al., 2000). The total percentage of equal-to-standard responses (right button press) was 41% for theDuration task and 32% for the Intensity task. The longerreaction times and flatter gradients observed in the Duration

task, as compared with the Intensity task (see Fig. 1), sug-gest that the Duration task was perceived as more difficult.

Data from fMRI

The group analysis revealed the activation of a verysimilar neural network in the Duration and Intensity tasks

Fig. 1. Average generalization gradients displaying the percentage of “standard” responses (ordinate) as a function of stimulus value (abscissa). Stimulusduration gradient (left). Stimulus intensity gradient (right).

Table 1Coordinates of significant cluster maxima in the group analysis for the Duration and Intensity tasks compared with the Control taska

Location (Brodmann area) Hemisphere Duration � Control t Intensity � Control t

FrontalBA6 L �48, 6, 48 12.92 �36, 3, 57 6.83BA8 R 9, 27, 48 25.59 9, 33, 45 21.22BA9 L �48, 12, 33 14.47 �42, 27, 27 9.13

R 51, 21, 27 25.61 45, 12, 33 23.64BA10/46 L �39, 45, 12 6.3

R 45, 57, 6 17.65BA44 L �39, 9, 30 15.40 �42, 9, 33 16.86

R 45, 9, 36 22.43Anterior insula L �33, 30, 3 18.68 �30, 33, 3 14.72

R 36, 33, 0 24.94 36, 33, 0 19.53Anterior cingulate L �9, 33, 27 7.29 �12, 36, 24 7.23Inferior parietal area (BA40) L �42, �36, 42 14.78 �42, �36, 42 14.55

R 39, �45, 42 18.19 36, �51, 48 19.8Superior temporal gyrus (BA21/22) L �51, �42, 9 7.49

R 54, �39, 6 12.34 51, �24, �3 5.41Superior occipital gyrus (BA19) R 39, �72, �9 6.87Anterior putamen L �15, 6, 6 10.65 �21, 3, 9 4.76

R 18, 3, 12 12.37 21, 3, 9 8.50Cerebellar vermis 3, �51, �18 8.91 0, �48, �18 8.92Mesencephalic nuclei R 6, �18, �9 7.23 6, �21, �9 6.79Caudate nucleus L �12, 9, 6 9.18

R 12, 6, 9 9.35 15, 9, 15 9.80Thalamus L �18, �9, 15 7.64

R 9, �3, 6 8.99 9, �12, 12 11.28

a Coordinates are in millimeters relative to the anterior commissure (as per the Talairach and Tournoux atlas). L � left; R � right; differences wereconsidered significant at P � 0.05, corrected for multiple comparison.

1535A.M. Ferrandez et al. / NeuroImage 19 (2003) 1532–1544

when contrasted to the Control task (see Table 1 and Fig. 2).The dorsolateral prefrontal area (BA9), anterior insula, in-ferior parietal area (BA40), and the anterior putamen (allbilaterally), and the cerebellar vermis, right mesenceph-alum, right caudate nucleus, right thalamus, left anteriorcingulate, left (BA6, BA44) and right (BA8) frontal areas,and the right superior temporal gyrus (BA21/22) were ac-tivated in both conditions. In addition, the Duration condi-tion yielded activation in right BA10/46 and in right BA44,as well as in the left superior temporal gyrus (BA22). TheIntensity condition yielded activation in left BA10/46, theleft caudate nucleus, left thalamus, and right superior oc-cipital gyrus (BA19).

The Duration-minus-Intensity comparison [(Duration �Control) � (Intensity � Control)] (see Table 2 and Fig. 3)yielded activation in the left premotor area (BA4/6), leftBA44, left putamen, right insula, bilateral BA45/47, SMA,inferior parietal lobes (BA40), and temporal gyri (BA21/22). Activation was examined at a more liberal threshold ofP � 0.001 (uncorrected for entire volume). No furtheractivation was found at this threshold.

The Intensity-minus-Duration comparison [(Intensity �Control) � (Duration � Control)] (see Table 3 and Fig. 3)yielded activation in the right superior occipital gyrus(BA19), bilateral fusiform gyri, left hippocampus, precu-neus, left posterior intraparietal sulcus (BA7/40), and leftprefrontal cortex (BA8, BA10). Activation was also exam-ined at a more liberal threshold of P � 0.001 (uncorrectedfor entire volume). Additional activation was found in thethalamus (right pulvinar, mediodorsal nucleus, and in thebilateral geniculate nuclei).

Activation in the Duration-minus-Intensity comparison,and in the Intensity-minus-Duration comparison was found

Fig. 2. Statistical parametric maps (SPMs) showing areas that were activatedmore during the Duration task than during the control task (upper), and moreduring the Intensity task than during the Control task (lower). SPM thresholdswere set at P � 0.05, corrected for multiple comparisons. Pixels are displayedon a gray scale (lower Z scores, light gray; higher Z scores, dark gray). SPMsare displayed on the Talairach space as a maximum intensity projection (allpixels activated on the cortical surface as well as in the deep structures arevisible as if in transparency through the brain) viewed from the right side(sagittal), back (coronal), and top (transversal) of the brain.

Table 2Coordinates of significant cluster maxima in the group analysis for theDuration task compared with the Intensity task [(Duration � Control) �(Intensity � Control)]a

Location (Brodmann area) Hemisphere Duration �Intensity

t

PrefrontalBA45/47 L �51, 15, 0 6.23

R 51, 30, 0 8.16BA44 L �54, 12, 18 6.99

SMA L �3, 9, 63 10.72R 6, 18, 57 8.43

Premotor cortex (BA4/6) L �48, 3, 48 5.97Insula R 36, 21, 6 6.49Superior temporal gyrus (BA21/22) L �54, �48, 9 10.59

R 57, �42, 12 8.82Inferior parietal cortex (BA40) L �60, �36, 27 10.02

R 60, �33, 30 7.62Putamen L �24, 3, 6 4.43

a Coordinates are in millimeters relative to the anterior commissure (asper the Talairach and Tournoux atlas). L � left; R � right; differenceswere considered significant at P � 0.05, corrected for multiple comparison.

1536 A.M. Ferrandez et al. / NeuroImage 19 (2003) 1532–1544

1537A.M. Ferrandez et al. / NeuroImage 19 (2003) 1532–1544

for the majority of the subjects (uncorrected P � 0.01), asillustrated in Fig. 4.

Discussion

The present study was aimed at detecting the brain re-gions specialized in duration discrimination compared tointensity discrimination for visual stimuli in a matching-to-sample task. By using fMRI, we were able to strengthen andclarify previous results obtained with the same paradigmusing PET (Maquet et al., 1996), or by combining PET andERP (Pouthas et al., 2000).

Common regions activated in Duration and Intensity tasks

As in our previous PET experiment (Maquet et al.,1996), a broad pattern of activation was observed in theDuration-minus-Control and Intensity-minus-Control com-parisons. Here, the pattern was nearly the same in the twotasks, and included cortical areas known to play a major rolein attentional and mnemonic processes. Thus, bilateral ac-tivation in prefrontal, temporal, and parietal regions (includ-ing BA9, cingulate, BA21, BA22, BA40) may subserveworking memory processes, activation in areas BA9, BA44,and right BA10 may subserve encoding and retrieval pro-cesses, and activation in bilateral BA8 may subserve prob-lem-solving processes (for a review, see Cabeza and Ny-berg, 2000). Furthermore, the left precentral gyrus (BA6),

bilateral insula, and bilateral superior temporal gyrus(BA22) have been shown to be among the regions activatedduring attention-orienting tasks (Coull et al., 2000). All ofthese cognitive processes are useful or even necessary forperforming matching-to-sample tasks. In the present study,subcortical areas were also activated, including the puta-men, caudate nuclei, thalamus, cerebellar vermis, and mes-encephalic nuclei. According to Cabeza and Nyberg (2000),tasks involving sequential decisions consistently engage thebasal ganglia, thalamic, and cerebellar regions. Studies onpatients or healthy humans suggest that the cerebellumparticipates in temporal processing, on perceptual (Ivry andKeele, 1989; Rao et al., 2001; Wittmann, 1999) and motortasks (Gaser et al., 2001; Ivry and Keele, 1989; Rao et al.,2001). In our PET study using the same tasks as in thepresent study (Maquet et al., 1996), activation was found inthe cerebellar vermis. This activation in the Duration-mi-nus-Control comparison, but also in the Intensity-minus-Control comparison, was interpreted as being possibly re-lated to the time locking of the motor response to thestimulus or to the more or less periodic distribution of thestimuli (every 1500 to 2300 ms). In the present experiment,we also found activation of the cerebellar vermis in bothtasks compared to the Control task, but not in the Duration-minus-Intensity comparison. The cerebellum may act as anintegrator for extracting sensory information (duration orintensity), and as a motor system for producing a timelyresponse (Gibbon et al., 1997; Penhune et al., 1998). Fi-nally, activation of the mesencephalic nuclei may reflect theinvolvement of serotoninergic or dopaminergic nuclei—such as the substantia nigra—which participate in loops thatlink the above subcortical structures together. However,given the spatial resolution of the technique, exactly whichmesencephalic structures were activated (substantia nigra orred nucleus) cannot be ascertained.

In sum, the brain regions found to be activated in theDuration-minus-Control comparison and in the Intensity-minus-Control comparison largely overlapped, thereby re-vealing a broad network that subserves common cognitiveprocesses. This network is probably necessary for solvingmatching-to-sample tasks no matter what property (intensityor duration) has to be discriminated.

Regions activated in the Duration task compared to theIntensity task

For the Duration-minus-Intensity comparison ([Duration� Control] � [Intensity � Control]), the regions activatedincluded those we expected to be specifically involved inthe duration condition, that is, prefrontal regions and, mostimportantly, the basal ganglia and SMA, that we failed toput in evidence in the PET study (Maquet et al., 1996).

Fig. 3. Brain regions activated in the Duration-minus-Intensity comparison [(Duration � Control) � (Intensity � Control)] in red, and in the Intensity-minus-Duration comparison [(Intensity � Control) � (Duration � Control)] in green (P � 0.05, corrected for multiple comparisons inside the volume of the brain).

Table 3Coordinates of significant cluster maxima in the group analysis for theIntensity task compared with the Duration task [(Intensity � Control) �(Duration � Control)]a

Location (Brodmann area) Hemisphere Intensity � Duration t

Parietal cortexBA7/40 L �27, �66, 36 5.63BA31/7 3, �66, 33 6.28

Fusiform gyri (BA37/19) L �30, �42, �18 8.20R 45, �54, �12 5.56

Superior occipital gyrus (BA19) R 42, �69, 30 6.47Prefrontal (BA8) L �18, 45, 45 6.21Prefrontal (BA10) L �6, 57, 6 4.53Hippocampus L �27, �21, �9 4.91Thalamusb

Pulvinar R 18, �27, 6 3.63Mediodorsal nucleus 0, �12, 12 3.87Lateral geniculate bodies L �24, �27, �3 3.82

R 27, �27, �6 3.30

a Coordinates are in millimeters relative to the anterior commissure (asper the Talairach and Tournoux atlas). L � left; R � right.

b Differences were considered significant at P � 0.05, corrected formultiple comparison, or (b) at P � 0.001, uncorrected and if their spatialextent was �5 voxels.

1538 A.M. Ferrandez et al. / NeuroImage 19 (2003) 1532–1544

Regions activated were the left putamen, SMA, ventrolat-eral prefrontal cortex, left premotor cortex, inferior parietalcortex, and temporal cortex. Some of these cortical struc-tures had been activated in the PET study (Maquet et al.,1996), but without showing their specific involvement in theDuration task. First, we will focus on the structures that maymore specifically subtend internal time-keeping mecha-nisms (basal ganglia and SMA). Second, we will examinethe cortical networks that form the systems generally asso-ciated with attention and memory functions.

Basal ganglia

In our study, the left putamen was activated in the Du-ration-minus-Intensity comparison, but not in the Intensity-minus-Duration comparison ([Intensity � Control] � [Du-ration � Control]), suggesting that activation of the leftputamen is specific to duration processing. Several fMRI orPET studies have reported activation in the putamen duringtiming tasks involving a strong motor component (Lejeuneet al., 1997; Penhune et al., 1998; Rao et al., 1997). This isconsistent with the well-known involvement of the putamenin motor tasks (Bhatia and Marsden, 1994; Lehericy et al.,1998), but recent studies have challenged the view that thebasal ganglia are solely concerned with motor control. Theyhave shown how neuronal activity in a network linking thebasal ganglia to the motor cortex is correlated with move-ment parameters, while neuronal activity linking the basalganglia to areas of the prefrontal cortex is more closelyrelated to aspects of cognitive function (for a review, seeMiddleton and Strick, 2000). The event-related fMRI studyof Rao and collaborators (2001) showed activation of theputamen during encoding processes, as well as in decisionmaking and response selection on a time-interval discrimi-nation task, i.e., in a time perception task. Furthermore,enhanced activation of the left putamen has been foundwhen the intervals to be estimated were longer (Coull et al.,2000), suggesting that activation of the putamen may bemodulated by the length of to-be-estimated time intervals.Time-keeping mechanisms are assumed to be under theinfluence of dopaminergic function, especially throughoutthe basal ganglia (Gibbon et al., 1997; Meck, 1996). Ourpresent fMRI study contributes to demonstrating the spe-cific involvement of the basal ganglia, and particularly theputamen, in the fine discrimination of duration.

Supplementary motor area

Interestingly, the SMA was found to be activated specif-ically in the Duration task (Duration-minus-Intensity com-parison). There is a body of convergent data suggesting thecritical role of a medial premotor pathway (SMA and basalganglia) in the internal representation of time. In one fMRIstudy, activation of a network that included the bilateralSMA, bilateral thalamus, and left putamen was observed ona self-paced tapping task but not on a synchronization task

(Rao et al., 1997). The Rao and collaborators’ event-relatedfMRI study (2001) contrasting time perception to pitchperception showed SMA activation during both time andpitch perception, although only duration perception yieldedactivation in the basal ganglia. The processing of the inter-val or ordinal properties of time has been found to yieldactivation in the SMA bilaterally, but the activation wasgreater during the processing of interval properties com-pared to ordinal properties, and was associated with activa-tion of the basal ganglia (Schubotz and von Cramon, 2001).This same network (bilateral SMA, bilateral thalamus, andleft putamen) was shown to be activated when subjects wereattending to a target after a long delay compared to attend-ing after a short delay (Coull et al., 2000). Our presentfindings are in line with these studies, and confirm not onlythat the SMA is involved in temporal processes (Macar etal., 1999, 2002; Pouthas et al., 2001), but also that the SMAtogether with the basal ganglia may play a crucial role intime-keeping mechanisms, probably by means of striataldopaminergic neurotransmission.

Cortical networks

The Duration-minus-Intensity comparison yielded acti-vation in the ventrolateral prefrontal cortex (BA45/47 bilat-erally), right insula, left premotor cortex (BA4/6), Broca’sarea, bilateral inferior parietal lobe (BA40), and bilateraltemporal gyrus (BA21/22).

Activation in the ventrolateral prefrontal cortex is con-sistent with our previous combined PET-ERP study usingthe same paradigm (Pouthas et al., 2000), which reportedthat the time course of activity in this area depended on theduration to be estimated. Several experiments studying tem-poral processing (Brunia et al., 2000; Coull et al., 2000;Kawashima et al., 2000; Lejeune et al., 1997; Maquet et al.,1996; Rao et al., 2001) have reported activation in theventrolateral prefrontal cortex (BA45/47). Activation in thisarea has been related to a number of different functions.This area could be involved in motor control when motoractions are based upon internal cues, and it could alsosupport the comparison of alternative solutions in workingmemory where the temporal contingencies of past and fu-ture movements are represented (Brunia et al., 2000). Dur-ing the temporal orientation task used by Coull and collab-orators (2000), it may have served to sustain attention for ashort period as well as hold and update representations inworking memory. The ventrolateral prefrontal cortex(VLFPC), which subtends such memory functions, shouldsupport a variety of computations, and this may account forwhy it has not been found to be specifically activated intiming tasks (Maquet et al., 1996; Rao et al., 2001).

Working memory function has been assumed to be sub-tended by both dorsolateral and ventrolateral (including theanterior insula) frontal regions, because they showed sus-tained activity during retention delays in visual memorytasks (Courtney et al., 1997), whereas other brain regions

1539A.M. Ferrandez et al. / NeuroImage 19 (2003) 1532–1544

showed transient activity and were therefore related to stim-ulus presentation (Haxby et al., 2000). Smith and Jonides(1999) assumed that VLPFC and dorsolateral prefrontalcortex (DLPFC) are involved in different working memoryfunction. They proposed that ventrolateral regions (BA 45and 47) mediate operations needed to sustain storage anddorsolateral regions (BA 46 and 9) implement the activemanipulation of information held in storage. In the presentstudy, VLPFC and not DLPFC was activated in the Dura-tion-minus-Intensity comparison whereas DLPFC was acti-vated in both the Duration-minus-Control and the Intensity-minus-Control comparisons, suggesting the involvement ofthis region in manipulating stored information, whatever the

attribute of the stimulus (duration or intensity). In contrast,specific activation of VLPFC in a task where the stimulusinformation has to be held in memory all along its durationsuggests that this area supports maintenance processes oninformation held in working memory, in accordance withSmith and Jonides (1999).

In addition, in most of the studies on temporal processingthat report activation in the ventrolateral prefrontal cortex,the premotor cortex was also activated (Coull and Nobre,1998; Gruber et al., 2000; Kawashima et al., 2000; Rao etal., 2001; Schubotz and von Cramon, 2001). The tasks usedin these studies share cognitive processes with the task usedin our own experiment, i.e., perceptual analysis of the du-

Fig. 4. Brain region maxima activated in single-subject analyses (P � 0.01 uncorrected and spatial extent � 5 voxels) for the Duration task compared withthe Intensity task [(Duration � Control) � (Intensity � Control)], on the upper half of the figure, and for the Intensity task compared with the Duration task[(Intensity � Control) � (Duration � Control)], on the lower half of the figure.

1540 A.M. Ferrandez et al. / NeuroImage 19 (2003) 1532–1544

ration of a stimulus and comparison with a previously en-coded duration. A meta-analysis of neuroimaging studiesclearly implicates the premotor cortex in temporary main-tenance of information (Smith and Jonides, 1999).

In sum, we assume that activation in left premotor re-gions, together with activation in ventrolateral prefrontalcortices obtained here, is mostly related to sustaining atten-tion for maintaining representations of temporal informationin working memory.

Another prefrontal region found to be activated in theDuration-minus-Intensity comparison was the left BA44(Broca’s area). The functional status most commonly attrib-uted to this area is that it is involved in the subvocalrehearsal system (Paulesu et al., 1993). Two recent studiesfound that Broca’s area was specifically involved in theencoding and maintenance of a subvocal rhythm and couldbe associated with chronometric counting (Gruber et al.,

2000; Kawashima et al., 2000). Interestingly, the latter stud-ies used visual stimuli, suggesting that Broca’s area can beactivated independently of the modality used to perceive theoriginal stimulus. Subjects, in fact, followed the instructionnot to count subvocally. However, most of them reportedthat, in the Duration task, they internally sang a musicalphrase, or imagined a sound during the lighting up of theLED. This is corroborated by the activation we obtained inthe superior temporal gyrus (BA21/22). Auditory imageryand perception seem to share neural systems within theauditory cortex (Zatorre et al., 1996). We can thereforeconclude that the combined activation of Broca’s area andthe temporal gyrus may reflect the use of a subvocal strategyand/or the representation of an internal auditory schemeduring the presentation of visual stimuli. The combinedactivation of Broca’s area and the temporal gyrus was prob-ably found in the Duration-minus-Intensity comparison, be-

Fig. 4 (continued)

1541A.M. Ferrandez et al. / NeuroImage 19 (2003) 1532–1544

cause in the Intensity discrimination task processing mostlikely took place relatively soon after stimulus presentation.In the Intensity task, subjects may have already taken theirdecision (is the stimulus equal to standard or not, and whatkey to press) before the LED switch-off. This is impossible,of course, in the Duration task, where subjects have to waitfor the LED switch-off for making their decision. So tem-porary maintenance of information in memory probably wasnot needed in the Intensity task as it was in the Durationdiscrimination task.

The parietal cortex was found to be activated both in theDuration-minus-Intensity comparison and in the Intensityminus Duration comparison, although not in the same re-gion (bilateral inferior parietal lobe BA40, and left posteriorintraparietal sulcus BA7/40, respectively). Time informa-tion processing has been shown to require attention (Blocket al., 1998), and the parietal cortex to be implicated insustained attention (Pardo, Fox and Raichle, 1991). Asstressed by Brunia and collaborators (2000), activation ofthe inferior parietal lobe has been shown to be associatedwith activation in different prefrontal regions in most stud-ies on temporal processing, in both perceptual tasks (Ka-washima et al., 2000; Rao et al., 2001) and motor tasks(Jancke et al., 2000; Macar et al., 2002). Thus, frontopari-etal networks could form systems for allocating attentionalresources (Posner et al., 1984), and for keeping informationavailable on line (Cohen et al., 1997), although they may beindependent of stimulus properties, as shown by Maquetand collaborators (Maquet et al., 1996; Cohen et al., 1997).

In summary, a complex network including several brainregions was found to be activated in the Duration-minus-Intensity comparison. In our study, cognitive functions thatare necessary for executing the matching-to-sample task ofDuration discrimination appear to be subtended by distinctsub-networks, consistently with the Rao and collaborators’study (2001). Activation of the basal ganglia (especially theleft putamen) and the bilateral SMA would reflect clockmechanisms. The prefrontal cortex (right ventrolateral pre-frontal and left premotor areas), together with the parietalcortex, would subtend attentional processes for maintainingtemporal information in working memory. Broca’s area,possibly in conjunction with the temporal cortex, appears tosubserve internal representations of duration.

Regions activated in the Intensity task compared to theDuration task

The intensity-minus-Duration comparison ([Intensity �Control] � [Duration � Control]) pointed out activation inthe right superior occipital gyrus (BA19), bilateral fusiformgyri, left hippocampus, parietal cortex (precuneus BA31/7,left posterior intraparietal sulcus BA7/40), and left prefron-tal cortex (BA8, BA10). Additional activation was found inthe thalamus (right pulvinar, mediodorsal nucleus, and bi-lateral geniculate nuclei) at a more liberal threshold. Similarnetworks have been observed in visual memory tasks (Rom-

bouts et al., 2001). Our matching-to-sample task requiredworking memory processes, and sustained attention in boththe Duration and the Intensity conditions, but the Intensitycondition required fine analysis of the visual characteristicsof the stimulus. Let us look first at the structures thatsubtend attentional and memory processes, and then at thenetwork known to be involved in visual processes.

The parietal cortex is part of a frontal-parietal networkthat includes the dorsolateral prefrontal cortex and interactswith extrastriate regions during visual processing (Corbetta,1998). Its activation could reflect the engagement of theattentional control network that orchestrates the facilitationof attended inputs in the extrastriate visual cortex (Martinezet al., 1999). The left intraparietal sulcus is specificallyinvolved in the voluntary orientation and maintenance ofattention to a target location (Corbetta et al., 2000). Coor-dinating the activity in the multiple extrastriate, parietal, andprefrontal cortical areas is assumed to reflect internalchanges in perception and to contribute to conscious vision(Lumer and Rees, 1999). In our intensity task, this networkmay have been implicated in the working memory andattentional processes required for matching-to-sample tasksin the visual modality. The hippocampus is known to playan important role in memory for visual stimuli (Milner etal., 1998) and in working memory (Curtis et al., 2000;Ranganath and D’Esposito, 2001). Frontal-temporal inter-actions, involving the prefrontal cortex and the medial tem-poral lobe, have been put in evidence in working memory(Kirchhoff et al., 2000; Stern et al., 2001). In the presentstudy, then, the hippocampus, together with the prefrontalcortex, may have been involved in the memory and com-parison processes needed to match the current luminance tothat of the previously encoded standard stimulus.

The visual pathway is known to involve the thalamicnuclei, namely, the lateral geniculate bodies projecting tothe striate cortex (Fujita et al., 2001) and the pulvinar, whichis connected to both the extrastriate cortex and the parietalcortex (Yeterian and Pandya, 1997). Pulvinar involvementin high-order visual transduction (Grieve et al., 2000) andmediodorsal nucleus involvement in object recognitionmemory (Parker and Gaffan, 1998) may be some of thefunctional roles played by these structures in the intensitydiscrimination task used in this study. In visual workingmemory tasks (Courtney et al., 1997), extrastriate cortexregions (inferior occipital and midfusiform gyri) were givento participate in early stages of visual processing. The fusi-form gyri and the superior occipital gyrus, which is part ofthe extrastriate cortex, certainly subtended the analysis ofthe characteristics of the visual stimulus, that is, its lumi-nance.

In sum, the Intensity-minus-Duration comparison indi-cated activation of a network subserving the attentional andmnemonic processing of a visual stimulus. Activation of avisual network in the Intensity-minus-Duration comparisonstresses that, on the contrary, processes linked to the visual

1542 A.M. Ferrandez et al. / NeuroImage 19 (2003) 1532–1544

nature of the stimulus were not at play in the Duration-minus-Intensity comparison.

Conclusion

The present study addressed the issue of the brain cor-relates of timing mechanisms in a delayed “same-different”paradigm. It complements and extends a previous studyusing the same paradigm with PET (Maquet et al., 1996) inwhich the activation pattern obtained in duration discrimi-nation could not be clearly differentiated from that obtainedin luminance discrimination. We confirmed here the patternalready observed in the PET study, showing activation of abroad network when comparing the Intensity or the Dura-tion task to the Control. This supramodal network could beresponsible for allocation of attentional resources and work-ing memory processes, whatever the attribute of the stimu-lus to perceive. We went further in differentiating patternsof activation subtending duration processing from thosesubtending intensity processing. Intensity processingyielded activation in a network classically known to beinvolved in operations applied to visual stimuli. In contrast,duration processing yielded activation in a complex networkthat included several subnetworks, each subserving specificoperations. The SMA and basal ganglia may subserve theclock mechanism, and the frontal-parietal areas may beinvolved in the attentional and mnemonic operations re-quired for comparing a current duration to a previouslyencoded standard. As in a previous study (Pouthas et al.,2000), combining functional neuroimaging data with ERPdata should provide a means of clearly describing the acti-vation time course of these subnetworks during the process-ing of duration information.

Acknowledgments

This study was supported by the Centre National de laRecherche Scientifique. The authors thank Vivian E. Waltzfor her assistance with the English version of the manu-script.

References

Alexander, G.E., DeLong, M.R., Strick, P.L., 1986. Parallel organization offunctionally segregated circuits linking basal ganglia and cortex. Annu.Rev. Neurosci. 9, 357–358.

Belin, P., McAdams, S., Smith, B., Savel, S., Thivard, L., Samson, S.,Samson, Y., 1998. The functional anatomy of sound intensity discrim-ination. J. Neurosci. 18, 6388–6394.

Belin, P., McAdams, S., Thivard, L., Smith, B., Savel, S., Zilbovicius, M.,Samson, S., Samson, Y., 2002. The neuroanatomical substrate of soundduration discrimination. Neuropsychologia 40, 1956–1964.

Bhatia, K.P., Marsden, C.D., 1994. The behavioural and motor conse-quences of focal lesions of the basal ganglia in man. Brain 117 (Pt 4),859–876.

Block, R.A., Zakay, D., Hancock, P.A., 1998. Human aging and durationjudgments: a meta-analytic review. Psychol. Aging 13, 584–596.

Brunia, C.H.M., deJong, B.M., vandenBergLenssen, M.M.C., Paans,A.M.J., 2000. Visual feedback about time estimation is related to aright hemisphere activation measured by PET. Exp. Brain Res. 130,328–337.

Cabeza, R., Nyberg, L., 2000. Imaging cognition. II: An empirical reviewof 275 PET and fMRI studies. J. Cogn. Neurosci. 12, 1–47.

Church, R.M., 1984. Properties of the internal clock. In: Gibbon, G., Allan,L. (Eds.), Timing and Time Perception. Annals of the New YorkAcademy of Sciences, New York, pp. 566–582.

Cohen, J.D., Perlstein, W.M., Braver, T.S., Nystrom, L.E., Noll, D.C.,Jonides, J., Smith, E.E., 1997. Temporal dynamics of brain activationduring a working memory task. Nature 386, 604–608.

Corbetta, M., 1998. Frontoparietal cortical networks for directing attentionand the eye to visual locations: identical, independent, or overlappingneural systems? Proc. Natl. Acad. Sci. USA 95, 831–838.

Corbetta, M., Kincade, J.M., Ollinger, J.M., McAvoy, M.P., Shulman,G.L., 2000. Voluntary orienting is dissociated from target detection inhuman posterior parietal cortex. Nat. Neurosci. 3, 292–297.

Coull, J.T., Frith, C.D., Buchel, C., Nobre, A.C., 2000. Orienting attentionin time: behavioural and neuroanatomical distinction between exoge-nous and endogenous shifts. Neuropsychologia 38, 808–819.

Coull, J.T., Nobre, A.C., 1998. Where and when to pay attention: the neuralsystems for directing attention to spatial locations and to time intervalsas revealed by both PET and fMRI. J. Neurosci. 18, 7426–7435.

Courtney, S.M., Ungerlelder, L.G., Kell, K., Haxby, J.V., 1997. Transientand sustained activity in a distributed neural system for human workingmemory. Nature 386, 608–611.

Curtis, C.E., Zald, D.H., Lee, J.T., Pardo, J.V., 2000. Object and spatialalternation tasks with minimal delays activate the right anterior hip-pocampus proper in humans. Neuroreport 11, 2203–2207.

Fujita, N., Tanaka, H., Takanashi, M., Hirabuki, N., Abe, K., Yoshimura,H., Nakamura, H., 2001. Lateral geniculate nucleus: anatomic andfunctional identification by use of MR imaging. AJNR Am. J. Neuro-radiol. 22, 1719–1726.

Gaser, C., Nenadic, I., Buchsbaum, B.R., Hazlett, E.A., Buchsbaum, M.S.,2001. Deformation-based morphometry and its relation to conventionalvolumetry of brain lateral ventricles in MRI. Neuroimage 13, 1140–1145.

Gibbon, J., Church, R.M., Meck, W.H., 1984. Scalar timing in memory. In:Gibbon, G., Allan, L. (Eds.), Timing and Time Perception. Annals ofthe New York Academy of Sciences, New York, pp. 52–77.

Gibbon, J., Malapani, C., Dale, C.L., Gallistel, C., 1997. Toward a neuro-biology of temporal cognition: advances and challenges. Curr. Opin.Neurobiol. 7, 170–184.

Grieve, K.L., Acuna, C., Cudeiro, J., 2000. The primate pulvinar nuclei:vision and action. Trends Neurosci. 23, 35–39.

Gruber, O., Kleinschmidt, A., Binkofski, F., Steinmetz, H., von Cramon,D.Y., 2000. Cerebral correlates of working memory for temporal in-formation. Neuroreport 11, 1689–1693.

Harrington, D.L., Haaland, K.Y., Hermanowicz, N., 1998. Temporal pro-cessing in the basal ganglia. Neuropsychology 12, 3–12.

Haxby, J.V., Petit, L., Ungerleider, L.G., Courtney, S.M., 2000. Distin-guishing the functional roles of multiple regions in distributed neuralsystems for visual working memory. Neuroimage 11, 145–156.

Ivry, R.B., Keele, S.W., 1989. Timing functions of the cerebellum. J. Cogn.Neurosci. 1, 136–152.

Jancke, L., Loose, R., Lutz, K., Specht, K., Shah, N.J., 2000. Corticalactivations during paced finger-tapping applying visual and auditorypacing stimuli. Brain Res. Cogn. Brain Res. 10, 51–66.

Kawashima, R., Okuda, J., Umetsu, A., Sugiura, M., Inoue, K., Suzuki, K.,Tabuchi, M., Tsukiura, T., Narayan, S.L., Nagasaka, T., Yanagawa, I.,Fujii, T., Takahashi, S., Fukuda, H., Yamadori, A., 2000. Humancerebellum plays an important role in memory-timed finger movement:an fMRI study. J. Neurophysiol. 83, 1079–1087.

1543A.M. Ferrandez et al. / NeuroImage 19 (2003) 1532–1544

Kirchhoff, B.A., Wagner, A.D., Maril, A., Stern, C.E., 2000. Prefrontal-temporal circuitry for episodic encoding and subsequent memory.J. Neurosci. 20, 6173–6180.

Lehericy, S., van de Moortele, P.F., Lobel, E., Paradis, A.L., Vidailhet, M.,Frouin, V., Neveu, P., Agid, Y., Marsault, C., LeBihan, D., 1998.Somatotopical organization of striatal activation during finger and toemovement: a 3-T functional magnetic resonance imaging study. Ann.Neurol. 44, 398–404.

Lejeune, H., Maquet, P., Bonnet, M., Casini, L., Ferrara, A., Macar, F.,Pouthas, V., TimsitBerthier, M., Vidal, F., 1997. The basic pattern ofactivation in motor and sensory temporal tasks: positron emissiontomography data. Neurosci. Lett. 235 (1-2), 21–24.

Lumer, E.D., Rees, G., 1999. Covariation of activity in visual and prefron-tal cortex associated with subjective visual perception. Proc. Natl.Acad. Sci. USA 96, 1669–1673.

Macar, F., Lejeune, H.F., Ferrara, M., Pouthas, V., Vidal, F., Maquet, P.,2002. Activation of the supplementary motor area and of attentionalnetworks in temporal processing. Exp. Brain Res. 142, 475–485.

Macar, F., Vidal, F., Casini, L., 1999. The supplementary motor area inmotor and sensory timing: evidence from slow brain potential changes.Exp. Brain Res. 125, 271–280.

Maquet, P., Lejeune, H., Pouthas, V., Bonnet, M., Casini, L., Macar, F.,Timsit-Berthier, M., Vidal, F., Ferrara, A., Degueldre, C., Quaglia, L.,Delfiore, G., Luxen, A., Woods, R., Mazziotta, J.C., Comar, D., 1996.Brain activation induced by estimation of duration: a PET study.Neuroimage 3, 119–126.

Marsden, C.D., 1984. Which motor disorder in Parkinson’s disease indi-cates true motor function of the basal ganglia? In: Evered, D.,O’Connor, M. (Eds.), Functions of the Basal Ganglia. Ciba FundationSymposium. Pitman, London, pp. 225–241.

Martinez, A., Anllo-Vento, L., Sereno, M.I., Frank, L.R., Buxton, R.B.,Dubowitz, D.J., Wong, E.C., Hinrichs, H., Heinze, H.J., Hillyard, S.A.,1999. Involvement of striate and extrastriate visual cortical areas inspatial attention. Nat. Neurosci. 2, 364–369.

Meck, W.H., 1996. Neuropharmacology of timing and time perception.Cogn. Brain Res. 3, 227–242.

Middleton, F.A., Strick, P.L., 2000. Basal ganglia and cerebellar loops:motor and cognitive circuits. Brain Res. Rev. 31 (2-3), 236–250.

Milner, B., Squire, L.R., Kandel, E.R., 1998. Cognitive neuroscience andthe study of memory. Neuron 20, 445–468.

Pardo, J.V., Fox, P.T., Raichle, M.E., 1991. Localization of a humansystem for sustained attention by positron emission tomography. Na-ture 349, 61–64.

Parker, A., Gaffan, D., 1998. Interaction of frontal and perirhinal corticesin visual object recognition memory in monkeys. Eur. J. Neurosci. 10,3044–3057.

Pastor, M.A., Artieda, J., Jahanshahi, M., Obeso, J.A., 1992. Time estima-tion and reproduction is abnormal in Parkinson’s disease. Brain 115 (Pt1), 211–225.

Paulesu, E., Frith, C.D., Frackowiak, R.S., 1993. The neural correlates ofthe verbal component of working memory. Nature 362, 342–345.

Penhune, V.B., Zattore, R.J., Evans, A.C., 1998. Cerebellar contributionsto motor timing: a PET study of auditory and visual rhythm reproduc-tion. J. Cogn. Neurosci. 10, 752–765.

Posner, M.I., Walker, J.A., Friedrich, F.J., Rafal, R.D., 1984. Effects ofparietal injury on covert orienting of attention. J. Neurosci. 4, 1863–1874.

Pouthas, V., Garnero, L., Ferrandez, A.M., Renault, B., 2000. ERPs andPET analysis of time perception: spatial and temporal brain mappingduring visual discrimination tasks. Hum. Brain Mapp. 10, 49–60.

Pouthas, V., George, N., Poline, J.B., Van de Moorteele, P.F., Hugueville,L., Pfeuty, M., Renault, B., LeBihan, D., Renault, B., 2001. Modulationof mesial frontocentral cortex activity by duration to be estimated. J.Cogn. Neurosci. Suppl. 120, 142 [abstract].

Ranganath, C., D’Esposito, M., 2001. Medial temporal lobe activity asso-ciated with active maintenance of novel information. Neuron 31, 865–873.

Rao, S.M., Harrington, D.L., Haaland, K.Y., Bobholz, J.A., Cox, R.W.,Binder, J.R., 1997. Distributed neural systems underlying the timing ofmovements. J. Neurosci. 17, 5528–5535.

Rao, S.M., Mayer, A.R., Harrington, D.L., 2001. The evolution of brainactivation during temporal processing. Nat. Neurosci. 4, 317–323.

Rombouts, S.A., Barkhof, F., Witter, M.P., Machielsen, W.C., Scheltens,P., 2001. Anterior medial temporal lobe activation during attemptedretrieval of encoded visuospatial scenes: an event-related fMRI study.Neuroimage 14, 67–76.

Saint-Cyr, J.A., Taylor, A.E., Nicholson, K., 1995. Behavior and the basalganglia. Adv. Neurol. 65, 1–28.

Schubotz, R.I., von Cramon, D.Y., 2001. Interval and ordinal properties ofsequences are associated with distinct premotor areas. Cereb. Cortex11, 210–222.

Smith, E.E., Jonides, J., 1999. Storage and executive processes in thefrontal lobes. Science 283, 1657–1661.

Stern, C.E., Sherman, S.J., Kirchhoff, B.A., Hasselmo, M.E., 2001. Medialtemporal and prefrontal contributions to working memory tasks withnovel and familiar stimuli. Hippocampus 11, 337–346.

Takada, M., Tokuno, H., Nambu, A., Inase, M., 1998. Corticostriatalprojections from the somatic motor areas of the frontal cortex in themacaque monkey: segregation versus overlap of input zones from theprimary motor cortex, the supplementary motor area, and the premotorcortex. Exp. Brain Res. 120, 114–128.

Wittmann, M., 1999. Time perception and temporal processing levels ofthe brain. Chronobiol. Int. 16, 17–32.

Yeterian, E.H., Pandya, D.N., 1997. Corticothalamic connections of extra-striate visual areas in rhesus monkeys. J. Comp. Neurol. 378, 562–585.

Zatorre, R.J., Halpern, A.R., Perry, D.W., Meyer, E., Evans, A.C., 1996.Hearing in the mind’s ear: a PET investigation of musical imagery andperception. J. Cogn. Neurosci. 8, 29–46.

1544 A.M. Ferrandez et al. / NeuroImage 19 (2003) 1532–1544