www.elsevier.com/locate/clinph

Clinical Neurophysiology 117 (2006) 2308–2314

Kinesthetic but not visual imagery assists in normalizingthe CNV in Parkinson’s disease

Vanessa K. Lim a,*, Melody A. Polych b, Antje Hollander c, Winston D. Byblow b,Ian J. Kirk a, Jeff P. Hamm a

a Department of Psychology, Research Centre for Cognitive Neuroscience, The University of Auckland, Private Bag 92019, Auckland, New Zealandb Department of Sport and Exercise Science, The University of Auckland, Private Bag 92019, New Zealand

c Max-Planck-Institute for Cognition and Neuroscience, Stephanstraße 1a, 04103 Leipzig, Germany

Accepted 10 June 2006Available online 4 August 2006

Abstract

Objective: This study investigated whether kinesthetic and/or visual imagery could alter the contingent negative variation (CNV) forpatients with Parkinson’s disease (PD).Methods: The CNV was recorded in six patients with PD and seven controls before and after a 10 min block of imagery. There were twotypes of imagery employed: kinesthetic and visual, which were evaluated on separate days.Results: The global field power (GFP) of the late CNV did not change after the visual imagery for either group, nor was there a signif-icant difference between the groups. In contrast, kinesthetic imagery resulted in significant group differences pre-, versus post-imageryGFPs, which was not present prior to performing the kinesthetic imagery task. In patients with PD, the CNV amplitudes post-, relativeto pre-kinesthetic imagery, increased over the dorsolateral prefrontal regions and decreased in the ipsilateral parietal regions. There wereno such changes in controls.Conclusions: A 10-min session of kinesthetic imagery enhanced the GFP amplitude of the late CNV for patients but not for controls.Significance: While the study needs to be replicated with a greater number of participants, the results suggest that kinesthetic imagerymay be a promising tool for investigations into motor changes, and may potentially be employed therapeutically, in patients with Par-kinson’s disease.� 2006 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.

Keywords: Parkinson’s disease; Movement related potentials; Imagery; Visual; Kinesthetic; Contingent negative variation

1. Introduction

There is much debate about whether the primary visu-al (V1) and motor (M1) areas are involved during mentalimagery of visual or motor/kinesthetic action (for reviewssee Jeannerod and Frak, 1999; Kosslyn et al., 2001;Kosslyn and Thompson, 2003; Mellet et al., 1998). Nev-ertheless, there is clear evidence that similar neural sub-strates are activated by visual imagery and visualstimulation. Such substrates may include the occipitopa-

1388-2457/$32.00 � 2006 International Federation of Clinical Neurophysiolo

doi:10.1016/j.clinph.2006.06.713

* Corresponding author. Tel.: + 64 9 373 7599; fax: +64 9 373 7450.E-mail address: v.lim@auckland.ac.nz (V.K. Lim).

rietal and occipitotemporal cortices (Chen et al., 1998;Klein et al., 2004; Kosslyn et al., 1993). Similarly, aswith actual movement, motor/kinesthetic imageryinvolves the activation of the inferoparietal cortex, sup-plementary motor area (SMA), anterior cingulate cortex,premotor cortex, dorsolateral prefrontal cortex, and thecerebellum (Niyazov et al., 2005; Oullier et al., 2005;Leonardo et al., 1995; Stephan et al., 1995; Decetyet al., 1993; Deiber et al., 1998).

There has been little research that has investigatedvisual imagery for motor events. Functional magneticresonance imaging (fMRI) during visual imagery hasindicated that occipital activation is dominant compared

gy. Published by Elsevier Ireland Ltd. All rights reserved.

V.K. Lim et al. / Clinical Neurophysiology 117 (2006) 2308–2314 2309

to the primary motor and somatosensory areas (Solodkinet al., 2004). This suggests that the areas of corticalactivation depend upon the type of imagery beingperformed.

Mental imagery has been shown to assist in the perfor-mance of musicians and athletes (Cumming and Hall,2002; Jackson et al., 2001). Imagery has also beenemployed to improve motor functioning in patients whohave suffered a stroke (Stevens and Stoykov, 2003). Duringmotor imagery in patients with Parkinson’s disease (PD)there is a reduction, relative to controls, in the metabolicand electrophysiological activity during the imagery ofexternal motor tasks (Thobois et al., 2000; Cunningtonet al., 1997, 1999, 2001; Samuel et al., 1997). The networksinvolved in motor imagery may not differ between controlsand patients, at least when patients are on medication(Cunnington et al., 2001). Therefore, the network involvedin motor imagery may be able to promote actual movementin patients with PD, particularly whilst maintaining theirmedication.

The Contingent Negative Variation (CNV) is a slownegative movement- and sensory-related potential thatreaches its maximum amplitude at the vertex (Cz; Walteret al., 1964). The CNV occurs in response to two successivestimuli: the warning/preparation (S1); and the imperative(S2) signals. The external paced paradigm allows for theinvestigation of the CNV during imagery. The CNVcomprises two main components: the ‘‘early’’ and the‘‘late’’ CNV.

The early component is frontally distributed, involvesthe prefrontal area, SMA (Bender et al., 2004) and cingu-late motor areas (Gomez et al., 2001) and is linked toarousal and attention associated with the warning stimu-lus S1 (Brunia, 1999). The late component has a centraldistribution and can be recorded from the putamen (Baresand Rektor, 2001). The late CNV is generated by the bas-al-ganglia-thalamocortical loop (Ikeda et al., 1994) butalso involves the primary motor areas (Bender et al.,2004; Gomez et al., 2001), prefrontal cortex, middle fron-tal cortex, anterior cingulate cortex and SMA (Gomezet al., 2003). In addition, activations of posterior regionsof the brain including occipital cortex, posterior cingula-ted cortex, and temporal and parietal areas (Gomezet al., 2003) have been demonstrated and is thought tobe related to the perception and integration of the S1and S2 stimuli. Finally, the CNV can be generated in bothmovement and movement-imagery tasks (Cunningtonet al., 1996) and can be used to differentiate processesrelated to general arousal and actual movement genera-tion (Brunia, 1999).

PD is associated with basal-ganglia dysfunction, whichis part of the network associated with the generation ofthe late CNV (Ikeda et al., 1994). It has been shown thatthe late CNV is reduced in PD patients when off theirmedication compared to controls (Thobois et al., 2000;Cunnington et al., 2001). This reduction is not as readilyapparent when PD patients are on their medication

indicating that the CNV can be normalized to some extentwith medication (Oishi et al., 1995). Therefore, the CNV isan ideal event related potential with which to investigatepossible differences in PD and controls during movementand imagery.

As noted above, the CNV is generated both duringimagined and actual movements (Cunnington et al.,1996), and imagery has been shown to benefit actual per-formance in athletes and trained musicians (Cummingand Hall, 2002; Jackson et al., 2001). In addition, it hasbeen reported that movement imagery is impaired in PDwhile visual imagery is not (Amick et al., 2006). The aimof this study was to determine if imagery (kinestheticand/or visual) produces any change in the CNV producedby actual movements. As the CNV is measurable duringimagined movements, it provides an indication as towhether or not participants are performing the imagerytask. Admittedly, however, this does not indicate if theimagery is kinesthetic or visual. The CNV amplitude is typ-ically larger in PD patients when on medication (Thoboiset al., 2000; Cunnington et al., 2001). It might be expectedtherefore, if imagery has any similar potential benefit forPD patients, it would result in an increase in the CNV.However, because medication tends to normalize theCNV amplitude, it is possible that imagery will not resultin a change in CNV amplitude in PD patients while onmedication, or that the effect of imagery will be somewhat‘‘additive’’ with medication, and result in a larger than nor-mal CNV amplitude.

Therefore, the aims of the current study were to investi-gate whether imagery can alter the movement relatedpotentials for patients with PD by increasing the amplitudeof the late CNV after imagery compared to before imageryand investigate possible differences in imagery types. It wasexpected that motor or kinesthetic imagery would have agreater effect on the amplitudes of the CNV particularlyfor patients with PD than controls compared to visualimagery as it would activate more cortical regions involvedin actual movement.

2. Methods

2.1. Participants

There were six patients (five males) with Parkinson’s dis-ease (age 69 ± 9 years, Table 1). All patients were takingmedication, and all testing was completed in the morningwhen participants felt their best. The range on the ModifiedHoehn and Yahr staging was 1–2 (where 0 = no sign of dis-ease and 5 = wheelchair bound). There were seven controls(five males) drawn from a similar population (age 71 ± 7years). Using the Edinburgh Handedness Inventory (Old-field, 1971), there was one person in each group who wasambidextrous, with the remaining participants being righthanded. All procedures were approved by a local ethicscommittee and informed consent was given by allparticipants.

Table 1Patient demographics

Age Year diagnosed Hohen and Yahr scale Side affected Medication

P1 75 2000 1.5 Bilateral SinemetP2 57 2000 1 Right Medipar, Norflex, BrufenP3 63 1997 2 Right SinementP4 84 2002 2 Bilateral Sinemet, PrednisoneP5 68 1999 1 Right Sinemet 25/100P6 68 1995 2 Right Sinemet CR, Bromocripine, Tazma, Dozpine

2310 V.K. Lim et al. / Clinical Neurophysiology 117 (2006) 2308–2314

2.2. EEG

Electrical Geodesics Inc. 128-channel Ag/AgCl elec-trode nets were used. EEG was recorded continuously(1000 Hz sampling rate; 0.01–100 Hz analogue band pass),was acquired using a common vertex (Cz) reference, andre-referenced to average reference. Vertical and horizontalelectrooculogram (VEOG and HEOG) were recorded andsubsequently employed for artifact rejection.

2.3. EMG

Electromyography (EMG) was amplified (BioAmps,ADInstruments, Castle Hill, NSW), sampled at 1000 Hz,and filtered (0.3–1000 Hz band pass). Bipolar electrodes(3 cm diameter, 2–5 mm apart) were placed over the flexorcarpi radialis (FCR) and extensor carpi radialis (ECR) andextensor digitorum communis (EDC) muscles of bothforearms.

2.4. Procedure

During the entire experiment, participants looked at a fix-ation cross. Auditory evoked potentials (AEP) were collect-ed in a random order for three separate tones (middleC = 262 Hz; middle G = 392 Hz and one octave above mid-dle G – high G = 784 Hz). These tones were later employedin the main experiment (150 tones, 70 dB, 100 ms durationand a random interstimulus interval from 1750 to 2250 ms;mean 2000 ms). For the CNV tasks, participants rested theirright hand on top of a button box and held the response but-ton with their right index finger. A trial began with the pre-sentation of S1 (middle C) and S2 followed after 2000 ms(inter-trial interval varied between 6000 and 8000 ms, mean7000 ms). The tone of S2 indicated either a movement trial(middle G; a go trial) or a non-movement trial (high G; ano-go trial). On the ‘‘go trials’’, the participant released theirfinger off the button box, lifted their arm and hand to touchthe space bar of a keyboard which was approximately 30 cmaway and then returned back to the button box. There wereequal presentations of the two types of trials which were ran-domly presented (total of 90 trials).

Following a block of movement trials, participants com-pleted an identical block of imagery trials, with the onlydifference being that during ‘‘go trials’’, instead of actuallymoving, they were required to image the movementinstead. During kinesthetic imagery, they were instructed

to imagine ‘‘the feeling and sensations’’ that occurred asthey performed the sequence of movements. During visualimagery trials, participants were instructed to imagine ‘‘see-ing’’ their hand perform the movement. Following theimagery block (irrespective of imagery type), a secondblock of movements, identical to the first movement blockwas performed. The two imagery conditions were per-formed on sequential days, with the order counter-bal-anced across participants. Participants were unaware ofthe purpose of the two types of imagery.

At the end of the second day, participants were requiredto complete an imagery questionnaire based on the ‘‘Move-ment Imagery Questionnaire’’ (Hall and Martin, 1997) thatasked the participants to rate how ‘‘easy’’ or ‘‘hard’’ it wasto visualize or feel four different movements involving theupper-limb.

2.5. Data analyses

Only conditions which were free from eye blinks and eyemovements as measured by the VEOG and HEOG chan-nels were accepted for the final analyses. Custom writtensoftware was employed for the rejection of such artifacts.For the AEPs, the average P1 (140–160 ms), N1 (190–205 ms), and P2 (265–285 ms) components were examined.

The CNV is measured before the imperative stimulus(S2) is presented, and thus before a trial can be determinedto be a ‘‘go’’ or ‘‘no-go’’ trial, therefore both types of trialswere averaged together. The Go: No-Go paradigm wassimply employed to ensure that the participants remainedfocused upon the task. The early CNV was calculated1000–1100 ms after S1, and the late CNV was calculated100 ms before the S2 (1900–2000 ms).

Global Field Power (GFP) was calculated for the earlyand late CNVs for all movement blocks (pre and postimagery) and imagery trials (visual and kinesthetic). Thisprocedure reduces the data from multiple electrodes intoa single time series and is a measure of field strength (Leh-mann and Skrandies, 1980). The formula for the averagereference used for the measure of GFP follows (Jeannerodand Frak, 1999):

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

1

n�Xn

i¼1

ui �

Pnj¼1

uj

n

0BB@

1CCA

2vuuuuut ; ð1Þ

AEP during movement task

CNVs Pre and Post Visual Imagery

0

1

2

3

4

5

6

7

8

-100 400 900 1400 1900 2400 2900

Time (ms)

GF

P

CNVs Pre and Post Kinesthetic Imagery

0

1

2

3

4

5

6

7

8

-100 400 900 1400 1900 2400 2900

Time (ms)

GF

P

a

b

c

ig. 1. (a) Patients (black) and controls (grey) AEP response recorded atz (Vertex) to S1 (Middle C) during the pre-visual condition. (b) Grandverage GFP CNV for controls and patients, pre (dotted lines) and postsolid lines) visual imagery and kinesthetic imagery (c). Note negativealues are plotted upwards and the baseline is 100 ms before theccurrence of S1, which occurs at time point zero.

V.K. Lim et al. / Clinical Neurophysiology 117 (2006) 2308–2314 2311

where n about equally distributed electrodes measure thepotentials, ei, i = 1, . . . ,n.

To investigate any possible effect of imagery on the fieldstrength a Mixed Analysis of Variance (MANOVA) wasperformed: Pre and Post imagery by Group (controls andpatients) for visual and kinesthetic imagery separately.The GFP analysis reduces data spatially therefore a furtheranalysis that employed all 129 electrodes was used to inves-tigate spatial changes of the amplitude. Independent t-testswere performed at each electrode comparing the twogroups. To filter out any comparisons that may simplyreflect statistical error, a v2 test was performed to deter-mine if the number of significant t-values found exceedsthat expected by chance (for complete description seeHamm et al., 2002, 2004).

3. Results

Comparisons of AEPs to the tones alone revealed nodifferences between groups. However, during the CNVtasks, the amplitude of the P2 component to the S1 stimulifor all conditions was greatly increased for the PD patients(t(11) = 2.2; p < 0.05; Fig. 1a). Despite the difference inamplitude of the P2 for the auditory stimuli S1, the baselineselected was 100 ms before S1.1

The first MANOVA employed involved the early CNVGFP (within groups factors: pre and post imagery – visualand kinesthetic separately and a between groups factor).There were no significant differences between groups forthe GFP of the early CNV for visual imagery(F(1,11) = 0.08; p > 0.05). Nor was there a difference forthe early CNV pre- and post-kinesthetic imagery(F(1,11) = 0.08; p > 0.05). Imagery was investigated as awithin factor and a between group factor was alsoemployed, there was also no significant effect of imagerytype (F(1,11) = 0.5; p > 0.05) nor was there an imagery bygroup interaction (F(1,11) = 0.1; p > 0.05).

The same analyses were repeated for the late CNV.There was a significant difference for the late pre- andpost-kinesthetic imagery (F(1,11) = 6.68; p < 0.05). Thisinteraction reflects the increase in GFP post-kinestheticimagery for the PD group, (3.65–4.48 lV; pre and post-kin-esthetic imagery, respectively) while controls showed nochange (4.22–4.07 lV for pre- and post-kinesthetic imag-ery). Furthermore, neither group demonstrated a changepost visual imagery (F(1,11) = 0.01; p > 0.05; 4.01 and4.64 lV for controls and 4.00–4.68 lV for patients, preand post visual imagery, respectively). The pre/post inter-action for the kinesthetic imagery was followed up by com-paring the amplitude at each electrode. As shown in Fig. 2,there were 27 significant electrodes for PD patients only(v2ð1Þ = 65.5; p < 0.05), which is more than expected by

statistical error (for complete description see Hamm

1 It should be noted that when the baseline was taken 700–800 ms afterS1 to eliminate any possible influence of the auditory evoked potential, theresults were similar and the interpretation of the paper remains the same.

FCa(vo

et al., 2002, 2004). No other significant pre/post compari-sons were found. The grand average CNVs from a selectionof electrodes (frontal electrode 16, Fz: 11, C3: 37, C4:105,and Cz:129) of is also shown in Fig. 2. The CNVs for thecontrols (top) and patients (bottom) are shown pre- (black)and post- (grey) kinesthetic imagery. The CNV electrodesfor the controls show very little change pre- and post-kin-esthetic imagery. In contrast, the patients CNV pre- andpost-kinesthetic imagery follow the statistical analyses withan increase in CNV amplitude in the frontal electrodes(e.g., Jackson et al., 2001) and a decrease in the right sideelectrode (e.g., C4).

Fig. 2. The grand averaged CNV waves are presented for electrodes (11 = Fz, 16 = frontal, 37 = C3, 105 = C4, and 129 = Cz). The top represents theaveraged CNV for the controls pre- (black) and post- (grey) kinesthetic imagery. The bottom represents the averaged CNV for the patients pre (black) andpost (grey). The middle panel shows the topographic maps of the difference waves pre- and post-kinesthetic imagery (left column) for the controls (toppanel) and PD patients (bottom panel) for the late CNV (1900–2000 ms). The far right column shows the significant electrodes pre- and post-kinestheticimagery for controls (top) and PD patients (bottom).

2312 V.K. Lim et al. / Clinical Neurophysiology 117 (2006) 2308–2314

Analysis of EMG (root mean square – RMS, baselinewas 1000 ms before movement and activity was 1000 msduring movement) indicated no movement for either mus-cles or groups during either imagery tasks (F(1,11) = 0.5;p > 0.05). PD patients rated the visual imagery as beingeasier than controls (t(11) = 1.9; p < 0.05) but their ratingof the kinesthetic imagery did not differ from controls(t(11) = 1.3; p > 0.05).

4. Discussion

A 10-min session of visual imagery did not influence theGFP of the late CNV for either group (Fig. 1b). Kinesthet-ic imagery, in contrast, increased the GFP of the late CNVfor patients but not for controls (Fig. 1c). PD patients offmedication typically show a reduction, relative to controls,in brain activity during imagery of external motor tasks(Thobois et al., 2000; Cunnington et al., 1997, 1999,2001; Samuel et al., 1997). On medication, however,patients with PD may not show differences in the CNVamplitude (Cunnington et al., 2001). In the current study,the patients with PD maintained their normal levels ofmedication and there was nevertheless an overall increasein the amplitude of the CNV after 10 min of kinestheticimagery. This was not observed after visual imagery, how-ever, suggesting perhaps that only the effect of kinestheticimagery on the late CNV amplitude was additive with thenormalizing effect of medication.

When the topography of the changes was investigated,increases were found distributed over the left frontal elec-

trodes, and decreases over the right parietal regions(Fig. 2). It should be noted that the large left frontal chang-es in contour after kinesthetic imagery corresponded pri-marily to changes at electrode 33. While the significantchange in CNV amplitude at electrode 33 is also evidentin the surrounding electrodes, the peak in the topographyis effectively driven by the amplitude at electrode 33. Thechanges in the topography post-kinesthetic imagery areconsistent with the idea that kinesthetic imagery alteredthe activity of the generators associated with the lateCNV in PD patients, but not that of the controls. The areasthat were modified after kinesthetic imagery includeregions which previous studies have shown to be affectedin PD. In particular, in patients with PD, neural activationin frontal regions during movement tasks have shown to bereduced relative to controls, while parietal areas have morebilateral activation (Cunnington et al., 2001; Samuel et al.,1997). The direction of change was in the direction thatwould be consistent with improvement, although becausethe participants were not removed from their medication,the net result appears to manifests as a larger than normalCNV amplitude.

There were no differences in topography or amplitudewhen the two imagery types were examined (irrespectiveof group). It is possible that the two imagery types weresufficiently similar because the kinesthetic imagery taskdoes not exclude visual imagery, such that previouslyreported differences were masked (Kosslyn et al., 2001;Mellet et al., 1998; Solodkin et al., 2004; Porro et al.,2000; Rao et al., 1993). Alternatively, EEG may not have

V.K. Lim et al. / Clinical Neurophysiology 117 (2006) 2308–2314 2313

a sufficient spatial resolution to detect differences intopography between imagery types. While this may havebeen the case, there was a clear effect on imagery type onthe post-movement related potential. Here, visual imagerymay not have been as effective as kinesthetic imagerybecause the areas involved in visual imagery are primarilyoccipital rather than primary motor and somatosensoryareas (Solodkin et al., 2004). Kinesthetic imagery, in con-trast, involves a more extensive motor network, thatincludes the inferoparietal cortex, SMA, anterior cingulatecortex, premotor cortex, dorsolateral prefrontal cortex,and the cerebellum (Niyazov et al., 2005; Oullier et al.,2005; Leonardo et al., 1995; Stephan et al., 1995; Decetyet al., 1993; Deiber et al., 1998). Moreover, using transcra-nial magnetic stimulation, only kinesthetic imagery modu-lated the corticomotor excitability in humans and notvisual imagery (Stinear et al., 2005). Therefore, only kines-thetic imagery enhanced the CNV of the PD patients due tothe enhanced motor networks involved in this type ofimagery compared to visual imagery.

The changes in the amplitude of the CNV post-kinesthet-ic imagery do not appear to be due to a change in generalarousal or attentional levels because the early CNV, whichis thought to reflect these processes, was not influenced. Inaddition, the increase in CNV activity only occurred afterkinesthetic imagery and not visual imagery. Finally, thisincrease only occurred for the PD patients. The lack ofany changes in the control participants’ CNV post-kines-thetic imagery may reflect a ceiling effect, insufficient imag-ery duration for controls, or may simply reflect the factthat the controls do not have impaired CNV generators.

The current results would suggest that kinesthetic imag-ery may be useful in altering the activity of the brain duringsimple movement tasks. It remains to be seen how long thiseffect lasts. Explicit instructions on the speed to respond tothe cues may help investigate any potential functionalchanges as a result of kinesthetic imagery. In this study,accuracy of go/no-go performance (range: 96–100%) butnot speed was encouraged. This study is limited by not test-ing the effect of medication on imagery and the limitedsample size. It seems unlikely that medication for PDwould interact with one type of imagery (kinesthetic) andnot the other form (visual). The increases in the CNVamplitude occurred after 10 min of kinesthetic imagerywith normal levels of medication. It is advantageous thatpatients do not necessarily have to be withdrawn frommedication for future studies to investigate the nature ofkinesthetic imagery. Likewise, future studies should alsoinvestigate the effect of withdrawing medication on kines-thetic imagery on larger sample sizes, and perhaps alsoon a wider range of impairment levels (greater range onthe Modified Hoehn and Yahr) group with greater diversi-ty of symptoms. These limitations notwithstanding, thecurrent findings are suggestive that imagery, and in partic-ular kinesthetic imagery, may have at least a temporaryand positive effect on the circuits of the basal ganglia thatare impaired in Parkinson’s disease. Certainly these

findings suggest that further investigations of kinestheticimagery and PD are warranted.

In summary, visual and kinesthetic imagery did not haveany effect on subsequent movement related CNV ampli-tudes for controls. In patients with PD, visual imagerydid not influence post imagery CNV variables. In contrast,kinesthetic imagery appeared to influence CNV ampli-tudes. These results suggest that kinesthetic imagery maybe a promising method to investigate for motor rehabilita-tion in stroke patients for example, and for patients withPD, kinesthetic imagery may assist in the initiation or com-pletion of a movement task by imaging the sequence beforemovement itself.

Acknowledgements

V.K.L. was supported by a New Zealand NeurologicalFoundation grant (Philip Wrightson Postdoctoral Fellow-ship) and a Staff grant from the University of Auckland.

References

Amick MM, Schendan HE, Ganis G, GCronin-Golomb A. Frontostriatalcircuits are necessary for visuomotor transformation: mental rotationin Parkinson’s disease. Neuropsychologia 2006;44:339–49.

Bares M, Rektor I. Basal ganglia involvement in sensory and cognitiveprocessing. A depth electrode CNV study in human subjects. ClinNeurophysiol 2001;112(11):2022–30.

Bender S, Resch F, Weisbrod M, Oelkers-Ax R. Specific task anticipationversus unspecific orienting reaction during early contingent negativevariation. Clin Neurophysiol 2004;115(8):1836–45.

Brunia CHM. Neural aspects of anticipatory behavior. Acta Psychol1999;101:213–42.

Chen W, Kato T, Zhu XH, Ogawa S, Tank DW, Ugurbil K. Humanprimary visual cortex and lateral geniculate nucleus activation duringvisual imagery. Neuroreport 1998;9(16):3669–74.

Cumming J, Hall C. Deliberate imagery practice: the development ofimagery skills in competitive athletes. J Sports Sci 2002;20:137–45.

Cunnington R, Iansek R, Bradshaw JL, Phillips JG. Movement-relatedpotentials associated with movement preparation and motor imagery.Exp Brain Res 1996;111(3):429–36.

Cunnington R, Iansek R, Johnson KA, Bradshaw JL. Movement-relatedpotentials in Parkinson’s disease. Motor imagery and movementpreparation. Brain 1997;120(8):1339–53.

Cunnington R, Iansek R, Bradshaw JL. Movement-related potentials inParkinson’s disease: external cues and attentional strategies. MovDisord 1999;14(1):63–8.

Cunnington R, Egan GF, O’Sullivan JD, Hughes AJ, Bradshaw JL,Colebatch JG. Motor imagery in Parkinson’s disease: a PET study.Mov Disord 2001;16(5):849–57.

Decety J, Jeannerod M, Durozard D, Baverel G. Central activation ofautonomic effectors during mental simulation of motor actions in man.J Physiol Lond 1993;461:549–63.

Deiber MP, Ibanez V, Honda M, Sadato N, Raman R, Hallett M.Cerebral processes related to visuomotor imagery and generation ofsimple finger movements studied with positron emission tomography.Neuroimage 1998;7(2):73–85.

Gomez CM, Delinte A, Vaquero E, Cardoso MJ, Vazquez M, Cromme-linck M, et al. Current source density analysis of CNV duringtemporal Gap paradigm. Brain Topogr 2001;13(3):149–59.

Gomez CM, Marco J, Grau C. Preparatory visuo-motor cortical networkof the contingent negative variation estimated by current density.Neuroimage 2003;20(1):216–24.

2314 V.K. Lim et al. / Clinical Neurophysiology 117 (2006) 2308–2314

Hall CR, Martin KA. Measuring movement imagery abilities: a revisionof the movement imagery questionnaire. J Ment Imagery1997;21:143–54.

Hamm JP, Johnson BW, Kirk IJ. Comparison of the N300 and N400ERPs to picture stimuli in congruent and incongruent contexts. ClinNeurophysiol 2002;113(8):1339–50.

Hamm JP, Johnson BW, Corballis MC. One good turn deserves another:an event-related brain potential study of rotated mirror-normal letterdiscriminations. Neuropsychologia 2004;42(6):810–20.

Ikeda A, Shibasaki H, Nagamine T, Terada K, Kaji R, Fukuyama H,et al. Dissociation between contingent negative variation and Ber-eitschafts potential in a patient with cerebellar efferent lesion.Electroencephalogr Clin Neurophysiol 1994;90(5):359–64.

Jackson PL, Lafleur AF, Malouin F, Richards C, Doyon J. Potential roleof mental practice using motor imagery in neurologic rehabilitation.Arch Phys Med Rehabil 2001;82(8):1133–41.

Jeannerod M, Frak V. Mental imaging of motor activity in humans. CurrOpin Neurobiol 1999;9(6):735–9.

Klein I, Dubois J, Mangin JF, Kherif F, Flandin G, Poline JB, et al.Retinotopic organization of visual mental images as revealed byfunctional magnetic resonance imaging. Cogn Brain Res2004;22(1):26–31.

Kosslyn SM, Alpert NM, Thompson WL, Maljkovic V, Weise SB,Chabris CF, et al. Visual mental-imagery activates topographicallyorganized visual-cortex – pet investigations. J Cogn Neurosci1993;5(3):263–87.

Kosslyn SM, Ganis G, Thompson WL. Neural foundations of imagery.Nat Rev Neurosci 2001;2(9):635–42.

Kosslyn SM, Thompson WL. When is early visual cortex activated duringvisual mental imagery? Psychol Bull 2003;129(5):723–46.

Lehmann D, Skrandies W. Reference-free identification of components ofcheckerboard-evoked multichannel potential fields. Electroencepha-logr Clin Neurophysiol 1980;48:609–21.

Leonardo M, Fieldman J, Sadato N, Campbell G, Ibanez V, Cohen L,et al. A functional magnetic resonance imaging study of corticalregions associated with motor task execution and motor ideation inhumans. Hum Brain Mapp 1995;3(2):83–92.

Mellet E, Petit L, Mazoyer B, Denis M, Tzourio N. Reopening the mentalimagery debate: lessons from functional anatomy. Neuroimage1998;8(2):129–39.

Niyazov DM, Butler AJ, Kadah YM, Epstein CM, Hu XP. Functionalmagnetic resonance imaging and transcranial magnetic stimulation:effects of motor imagery, movement and coil orientation. ClinNeurophysiol 2005;116(7):1601–10.

Oishi M, Mochizuki Y, Du C, Takasu T. Contingent negative-variationand movement-related cortical potentials in Parkinsonism. Electroen-cephalogr Clin Neurophysiol 1995;95(5):346–9.

Oldfield RC. The assessment and analysis of handedness: the Edinburghinventory. Neuropsychologia 1971;9:77–113.

Oullier O, Jantzen KJ, Steinberg FL, Kelso JAS. Neural substrates of realand imagined sensorimotor coordination. Cereb Cortex2005;15(7):975–85.

Porro CA, Cettolo V, Francescato MP, Baraldi P. Ipsilateral involvementof primary motor cortex during motor imagery. Eur J Neurosci2000;12(8):3059–63.

Rao SM, Binder JR, Bandettini PA, Hammeke TA, Yetkin FZ,Jesmanowicz A, et al. Functional magnetic resonance imaging ofcomplex human movements. Neurology 1993;43(11):2311–8.

Samuel M, CeballosBaumann AO, Blin J, Uema T, Boecker H, Passing-ham RE, et al. Evidence for lateral premotor and parietal overactivityin Parkinson’s disease during sequential and bimanual movements – aPET study. Brain 1997;120:963–76.

Solodkin A, Hlustik P, Chen EE, Small SL. Fine modulation in networkactivation during motor execution and motor imagery. Cereb Cortex2004;14(11):1246–55.

Stephan KM, Fink GR, Passingham RE, Silbersweig D, Ceballosbau-mann AO, Frith CD, et al. Functional-anatomy of the mentalrepresentation of upper extremity movements in healthy-subjects. JNeurophysiol 1995;73(1):373–86.

Stevens JA, Stoykov MEP. Using motor imagery in the rehabilitation ofhemiparesis. Arch Phys Med Rehabil 2003;84(7):1090–2.

Stinear CM, Byblow WD, Steyvers M, Levin O, Swinnen SP. Kinesthetic,but not visual, motor imagery modulates corticomotor excitability.Exp Brain Res 2005:1–8.

Thobois S, Dominey PF, Decety J, Pollak P, Gregoire MC, Le Bars D,et al. Motor imagery in normal subjects and in asymmetricalParkinson’s disease – a PET study. Neurology 2000;55(7):996–1002.

Walter WG, Winter AL, Cooper R, McCallum WC, Aldridge VJ.Contingent negative variation – electric sign of sensorimotor associ-ation + expectancy in human brain. Nature 1964;203(494):380.