β-bursts reveal the trial-to-trial dynamics of movement ... · distribution of the number of...

β-bursts reveal the trial-to-trial dynamics of movement initiation and

cancellation

Shorttitle:β-burstsindexmovementinitiation&cancellation

JanR.Wessel1,2

1DepartmentofPsychologicalandBrainSciences,UniversityofIowa,IowaCityIA52245

2DepartmentofNeurology,UniversityofIowaHospitalandClinics,IowaCityIA52242

CorrespondingAuthor:

JanR.Wessel,Ph.D.

DepartmentofNeurology

UniversityofIowaHospitalsandClinics,444MRC

IowaCity,IA52242

[email protected]

+13193357745

www.wessellab.org

.CC-BY-NC-ND 4.0 International licenseacertified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under

The copyright holder for this preprint (which was notthis version posted May 21, 2019. ; https://doi.org/10.1101/644682doi: bioRxiv preprint

https://doi.org/10.1101/644682

http://creativecommons.org/licenses/by-nc-nd/4.0/

2

Abstract

Theneurophysiologicalbasisofmotorprocessesandtheircontrolisoftremendous

interest to basic researchers and clinicians alike. Notably, bothmovement initiation and

cancellationareaccompaniedbyprominentfieldpotentialchangesintheβ-frequencyband

(15-29Hz).Intrial-averages,movementinitiationisindexedbyβ-banddesynchronization

oversensorimotorsites,whilemovementcancellationissignifiedbyβ-powerincreasesover

(pre)frontal areas. However, averagingmisrepresents the true nature of the β-signal. As

recentworkhashighlighted,rawβ-bandactivityischaracterizedbyshort-lasting,burst-like

events, rather than by steady modulations. To investigate how such β-bursts relate to

movement initiation and cancellation in humans,we investigated scalp-recorded β-band

activity in 234healthy subjects performing the Stop-signal task. Four observationswere

made: First, both movement initiation and cancellation were indexed by systematic,

localized changes in β-bursting.While β-bursting at bilateral sensorimotor sites steadily

declinedduringmovementinitiation,β-burstingincreasedatfronto-centralsiteswhenStop-

signalsinstructedmovementcancellation.Second,theamountoffronto-centralβ-bursting

clearly distinguished successful from unsuccessful movement cancellation. Third, the

emergence of fronto-central β-bursting coincided with the latency of the movement

cancellationprocess,indexedbyStop-signalreactiontime.Fourth,individualfronto-central

β-burstsduringmovementcancellationwerefollowedbyalow-latencyre-instantiationof

bilateralsensorimotorβ-bursting.Thesefindingssuggestthatβ-burstingisafundamental

signature of the motor system, reflecting a steady inhibition of motor cortex that is

suppressedduringmovementinitiation,andcanberapidlyre-instantiatedbyfrontalareas

whenmovementshavetoberapidlycancelled.



https://doi.org/10.1101/644682


3

SignificanceStatement

Movement-relatedβ-frequency(15-29Hz)changesareamong themostprominent

features of neural recordings across species, scales, and methods. However, standard

averaging-basedmethodsobscurethetruedynamicsofβ-bandactivity,whichisdominated

byshort-lived,burst-likeevents.Here,wedemonstratethatbothmovement-initiationand

cancellation in humans are characterized by unique trial-to-trial patterns of β-bursting.

Movement initiation is characterized by steady reductions of β-bursting over bilateral

sensorimotorsites.Incontrast,duringrapidmovementcancellation,β–burstsfirstemerge

overfronto-centralsitestypicallyassociatedwithmotorcontrol,afterwhichsensorimotor

β–burstingre-initiates.Thesefindingssuggestafundamentallynovel,non-invasivemeasure

of theneural interactionunderlyingmovement-initiationand–cancellation,openingnew

avenuesforthestudyofmotorcontrolinhealthanddisease.



https://doi.org/10.1101/644682


4

Introduction

Activityintheβ-frequencyband(15-29Hz)isaprominentconstituentoftheneural

field potential. It can be observed at spatial scales ranging from extracellular to scalp

recordings, in species ranging fromrodents tohumans, andusingmethods ranging from

intracranial recordings to magnetoencephalography [1-4]. The β-frequency plays a

particularly important role in the functioning of themotor system. In particular, during

movementinitiation,aprominentdesynchronizationofβ-bandactivityisclearlyobservable

over sensorimotor areas [5, 6]. In contrast, the rapid cancellation of movement is

accompaniedbyβ-powerincreasesover(pre-)frontalcorticalareasgenerallyimplicatedin

cognitive control [7-11]. Moreover, movement-related changes in β-power can also be

observedinextrapyramidalpartsofthemotorsystem,includingthebasalganglia[12-15],

where abnormal β-rhythms are prominently observed in movement disorders such as

Parkinson’sDisease[16-19].

Recentstudiesofraw,unaveragedβ-bandactivity,however,haveledtoasignificant

reappraisalofthenatureofthisβ-bandactivity.Whiletrial-averagingapproachessuggest

that movement-related changes in the β-band reflect steady (de)synchronizations that

stretch over several hundred milliseconds, unaveraged β-band activity is primarily

characterized by rapid burst-like events, which typically last less than ~150ms [20-22].

Whiletheseburst-eventsappearasslow-evolving(de)synchronizationsinthetrial-average,

analysesofsingle-trialdatahavefoundthatthesimplepresenceorabsenceoftheseβ-bursts,

rather than overall changes in β-power, is the most reliable predictor of trial-to-trial

behavior[3].



https://doi.org/10.1101/644682


5

Here,wethereforeinvestigatethecharacteristicsofsingle-trialβ-burstinginhumans

duringboth the initiation,andparticularly,during therapidcancellationofmovement.A

largesampleofhealthyhumanparticipants(N=234)performedthestop-signaltask[23,24],

amotortaskthatincludesbothinstancesofmovementinitiation(followingaGo-signal)and

movement cancellation (on trials that include a subsequent Stop-signal). We used non-

invasive scalp-EEG recordings to investigate how β-bursting on individual trials indexes

both processes and the interaction between them. Specifically, we investigated five

questions:1.Ishumanβ-bandactivityduringmovementburst-like?2.Aretheresystematic

topographicalpatternsofβ-burstsduringbothmovement initiationand–cancellation?3.

Aremovementinitiationandcancellationaccompaniedbysystematictemporalpatternsof

β-burst activity? 4. Do specific patterns of β-burst activity distinguish successful from

unsuccessfulmovementcancellation?5.Aretheresystematicrelationshipsbetweeninitiation

andcancellation-relatedchangesinβ-burstingwhenmovementshavetoberapidlystopped?

Results

Systematicspatiotemporalpatternsofβ-burstscharacterizebothmovementinitiation

andcancellation

Figure1ashowsthatafterGo-signals(whichpromptthestartofmovementinitiation),

bilateralsensorimotorsites(electrodesC3andC4)initiallyshowlocalizedβ-bursting,which

immediatelybeginstodecreaseastimegetsclosertomovementexecution,resultingina

significantlineartrend(Figure1b,d,lineartrendforleft-handresponsesatC4:Z=-3.63,p

=.00028,left-handresponsesatC3:Z=-3.22,p=.001,right-handresponsesatC4:Z=-2.81,

p=.005,right-handresponsesatC3:Z=-3.23,p=.0013).Furthermore,inthelead-upto



https://doi.org/10.1101/644682


6

responseexecution, thispattern lateralizes,withsitescontralateral to theresponsehand

showingastrongersustainedreductioninβ-bursting(significantlateralizationatp<.0001

(FDR-corrected)attimefiveconsecutivewindowsfrom325to575msforelectrodeC4,t(233)

=-4.91,p=1.75*10-06,d=.38,t(233)=-8.25,p=1.2*10-14,d=.63,t(233)=-6.74,p=1.24*10-

10,d=.52,t(233)=-6.83,p=7.37*10-11,d=.61,t(233)=-5.75,p=2.81*10-08,d=.5;andat

three consecutive timewindows from325 to475ms for electrodeC3, t(233)=5.28, p=

2.92*10-07,d=.41,t(233)=6.02,p=6.6*10-09,d=.49,t(233)=4.81,p=2.65*10-06,d=.37).

An inspection of individual trial data shows that the single-trial β-band signal is indeed

dominatedbyshort,burst-likeevents,ratherthanbysteadymodulations(Figure1c,e;plots

of single trial data for each individual participant can be found in the Supplementary

Materials).



https://doi.org/10.1101/644682


7

Figure 1. β-burst properties during movement initiation on Go-trials. A) Topographical

distribution of the number of β-bursts on the scalp at different consecutive time-windows

followingtheGo-signal,forbothleft-hand(top)andright-hand(bottom)responses.Thereare

visible bilateral peaks over electrode sites C3 and C4 in both conditions until about 200ms

followingtheGo-signal.B)Temporaldistributionofβ-burstnumbersfollowingtheGo-Signal

atright lateralelectrodeC4duringboth left-andrighthandresponses.Asignificant linear



https://doi.org/10.1101/644682


8

trendcanbeobserved,suchthatthenumberofβ-burstssteadilydecreasesfollowingGo-signal

onset.Moreover,thereisasignificantlateralizationofthiseffectstartingat325msfollowing

theGo-signal,suchthattheβ-burstnumbersforthecontralateralhandkeepdiminishingwhile

thenumberasymptotesearlierfortheipsilateralresponsehand.C)Individualtrialβ-banddata

atelectrodeC4fromonerepresentativesubject,clearlyshowingburst-likeβ-events.D)AsB,

butforleft-lateralelectrodeC3.E)AsC,butforC3.

Figure2ashowsthatinthetimeperiodfollowingStop-signals(whichfollowedGo-

signalson1/3ofalltrialsatavariabledelayandpromptedtheparticipantstoattemptto

stop theirpendingmovement),nocoherentspatiotemporalorganizationof therateofβ-

burstscanbeobserveduntilaround200msafterthestop-signal.Atthatpoint,aclearradial

fronto-centraltopographicaldistributionemerges,centeredaroundelectrodeFCz.Justlike

Go-signal-relatedactivityatC3andC4,single-trialStop-signal-relatedactivityatFCzclearly

showsthepresenceofβ-bursting(Figure1d).



https://doi.org/10.1101/644682


9

Figure2.β-burstpropertiesduringmovementcancellationonStop-trials.A)Topographical

distribution of the number of β-bursts on the scalp at different consecutive time-windows

followingtheStop-signal,separatelyforsuccessful(top)andfailed(bottom)Stop-trials.Inthe

time-windowtowardstheendofSSRT(orangewindow),aclearfronto-centralorganizationof

β-burstingcenteredaroundelectrodeFCz isobservable.B)Temporaldistributionofβ-burst

numbers following the Stop-Signal at fronto-central electrode FCz. C) Comparison of the

numberofβ-burstsbetweensuccessfulandfailedStop-trialsintheStop-signal-to-SSRTperiod

foreachsubject.SuccessfulStop-trialsareaccompaniedbyasignificantlylargernumberofβ-

bursts.D)Individualtrialβ-banddataatelectrodeFCzfromonerepresentativesubject,clearly

showingburst-likeβ-events.

Fronto-centralβ-burstingindexessuccessfulmovementcancellation



https://doi.org/10.1101/644682


10

BehaviorintheStop-signaltaskwastypical(meanGo-trialreactiontime:533.51ms,

SEM:6.6;failedStop-trialreactiontime:459.59ms,SEM:5.79;stopaccuracy:.52,SEM:.002,

Stop-signaldelay:282.42,SEM:7.91;Stop-Signalreactiontime:244.98ms,SEM:3.62).

Theincreaseoffronto-centralβ-burstingafterStop-signals(comparedtoGo-signals)

is highly significant (Figure 3 shows the topographical difference plots between the

distribution of β-bursts following Stop- vs. Go-trials, thresholded at p < .0001, FDR-

corrected).Moreover,thetimeaftertheStop-signalatwhichthisfronto-centralorganization

ofβ-burstingdevelopsoverlapswiththeendofSSRT(~245ms).Theperiodbeforetheend

of SSRT is the exact time period during which neural activity reflecting movement

cancellationshouldbemaximalaccordingtocomputationalmodels[25].Moreover,adirect

comparisonofthepre-SSRTtimeperiodineachindividualparticipant(i.e.,thetimerange

between the Stop-signal and the end of that participants’ SSRT) revealed that successful

Stop-trialsyieldedanincreasedrateofβ-burstsatFCzcomparedtofailedStop-trials(t(233)

=3.47,p<.0007,d=.26,Figure2b,c).

Figure 3. Statistical comparison of β-burst topographies following Stop- and Go-signals,

separatelyforsuccessful(top)andfailed(bottom)Stop-trials.Thresholdedforsignificanceat

p<.0001,FDR-correctedformultiplecomparisons.Itisevidentthatwhilemovementinitiation

(Go-trials)isaccompaniedbyasignificantlyincreasedβ-burstcountatlateralelectrodesuntil



https://doi.org/10.1101/644682


11

~175msaftertheGo-signal,movementcancellationisaccompaniedbyasignificantlyincreased

β-burstcountatfronto-centralelectrodes,startingat~225msaftertheStop-signal.

Fronto-centralβ-burstsarefollowedbyincreasedbilateralsensorimotorβ-bursting

Figure4ashowsthetemporaldevelopmentofβ-burstingatsensorimotorsitesipsi-

and contralateral to the to-be-stopped movement on successful stop-trials. Importantly,

these plots are time-locked to the latency of the first fronto-central β-burst event that

occurredwithinthepre-SSRTtimeperiod.ThisisthetimeperiodduringwhichtheStop-

processshouldbeactiveaccordingtherace-modelsoftheStop-signaltask[23,25,26].These

plotsshowasignificantincreaseinbilateralsensorimotorβ-burstingwithin25msofthefirst

fronto-central β-burst. To evaluate significance, these valueswere compared tomatched

time-periodsonsuccessfulstop-trialswithoutfronto-centralβ-burstsinthepre-SSRTperiod

(exactvaluesforthepairwisecomparisonsbetweentrialswithandwithoutβ-bursts:Z=

7.81,p=5.65*10-15andZ=4.36,p=1.28*10-05 for thetwosignificant time-windowsfor

contralateralsitesandZ=8.06,p=7.7*10-16andZ=4.8,p=1.55*10-06foripsilateralsites).

Atopographicalrepresentationofthiseffectrevealsthattheincreaseinβ-burstingfollowing

thefirstfronto-centralβ-burstislocalizedtobilateralsensorimotorsites(inadditiontothe

fronto-centralelectrodessurroundingFCz,Figure4b).



https://doi.org/10.1101/644682


12

Figure 4. Interaction between fronto-central and lateralized β-bursting during successful

movement cancellation. A) Number of contra- and ipsilateral β-bursts at electrodes C3/4

surrounding fronto-central β-bursts during the Stop-signal-to-SSRTperiod.During the time

window 25ms following the fronto-central β-burst (green highlighting), both contra- and

ipsilateral β-bursting over electrodes C3/C4 was significantly increased. B) Topographical

representationofβ-burstinginthehighlightedtimeperiod,25msfollowingthefronto-central

β-burstevent,showingthatincreasesinβ-burstactivityarelocalizedtobilateralsensorimotor

sites(inadditiontofronto-centralsitessurroundingFCz).



https://doi.org/10.1101/644682


13

Discussion

Thecurrentsetoffindingsshowsthatβ-burstingonindividualtrialsistightlyrelated

tomovementinhumans.Priortomovementinitiation,bilateralsensorimotorsitesshowed

localizedpatternsofβ-bursting,whichmayrepresentaninhibitedstateofthemotorsystem

[4,11,27].Thisβ-bursting is then steadily reducedwhenaGo-signal ispresentedanda

movement is initiated (reflecting a net-disinhibition of the motor system). Notably, this

reductioninburst-rateslateralizedjustpriortomovementexecution,whichmirrorsprior

observationsoflateralizedmovement-relatedβ-desynchronizationsinthetrialaverage[6,

28,29],andlikelyexplainsthesepatterns.

Subsequently, when inhibition had to be rapidly reinstated – i.e., when already

initiated movements had to be suddenly cancelled following Stop-signals – β-bursting

significantlyincreasedatfronto-centralscalpsites.Thisfronto-centralincreaseinβ-bursting

showeditsmostcoherentspatiotemporalorganizationat fronto-centralsitesduringtime

periodtowardstheendofSSRT.Thisistheexacttimeperiodthat,accordingtoapopular

computationalmodeloftheStop-signaltask,shouldyieldthehighestamountofinhibitory

activity[25].Finally,onehighlynotablepatterninourdatawasthatindividualinstancesof

fronto-centralβ-burstingwerefollowedbyarapid(<25ms)re-instantiationofβ-bursting

over sensorimotor sites,both ipsi- andcontralateral to the to-be-stoppedmovement.We

proposethatthisreflectsalow-latencyre-instantiationofinhibitionatthelevelofthemotor

system,triggeredbyafronto-centralcontrolsignal(reflectedinthefronto-centralβ-burst

events).

These features of our data show intriguing overlap with several of the

proposed properties of the neural cascade underlying movement cancellation that was



https://doi.org/10.1101/644682


14

identified in prior work using measures other than β-bursting. First, the topographical

distributionandtemporalevolutionofthefronto-centralβ-burstreportedhereparallelsthe

propertiesoftheStop-signalP3,anevent-relatedpotentialthathasbeenproposedtoreflect

motorinhibitionintrial-averagingstudiesofphase-lockedevent-relatedEEGactivity[30-

32]. Second, trial-averaged power in the β-band has been repeatedly implicated in

movementcancellationinstudiesoftrial-averagedtime-frequencyactivity[7-10,13].This

includes intracranial recordings from sites that could well underlie the fronto-centrally

distributedpatternofβ-burstingobserved inourstudy, includingthepre-supplementary

motor area [33]. Previous studiesmay havemissed these β -burst patterns due to trial-

averaging.Third,theprominentrace-modeloftheStop-signaltaskproposesthatbehavior

in this task canbemodeledby a race between twoprocesses; oneworking towards the

executionofthemotorresponseandtriggeredbytheGo-signal(theGo-process),andone

working towards the cancellation of that response and triggered by the Stop-signal (the

Stop-process[23]).Somecontroversystillremainsregardingwhetherthesetwoprocesses

operateindependently,orwhethertheStop-process,onceinitiated,directlyinfluencesthe

Go-process[25,26,34,35].Ourresultsshowthatstopping-relatedactivityatfronto-central

sitesisimmediatelyfollowedbyare-instantiationofβ-burstingoversensorimotorsites–

thesamesignaturewhoseinitialreductionrepresentsthestartoftheGo-process(Figure4).

ThisindicatesthattheactivityoftheStop-processisfollowedbyasubstantialchangeinthe

neuralrepresentationoftheGo-process,thusspeakinginfavoroftheinteractiverace-model

ofmovementcancellation.Finally,thepatternofβ-burstsduringmovementcancellationin

ourstudydovetailswellwitha fourthaspectof theexisting literature.Notably, theStop-

relatedincrease insensorimotorβ-burstingfollowingfronto-centralβ-burstswasequally



https://doi.org/10.1101/644682


15

observableatbothipsi-andcontralateralsensorimotorsites.Thisparallelsexistingfindings

fromthemotorphysiologyliteratureontheStop-signaltask,whichhaveusedmotorevoked

potentials (asignatureofcortico-spinalexcitabilityof specificmotor tracts) toshowthat

rapidmovement cancellation is non-selective. Specifically, whenmovements are rapidly

cancelled, redutions of cortico-spinal excitability are not limited to the specific motor

effectors that have to be stopped. Instead, suppression of the entire motor system is

observed,evenattask-unrelatedmuscles[36-39].Thesimultaneous,rapidre-activationof

β-burstingoverbothsensorimotorcorticesafterfronto-centralStop-relatedβ-burstscould

be the neurophysiological expression of that samenon-selective property in our current

dataset.

The current studyhas several key implications for future research, andmotivates

several immediate follow-up experiments. First, the precise neural origin of movement-

relatedβ-burstsneedstobeinvestigated.Computationalmodelinghassuggestedthatthese

eventsmayresultsfromtheintegrationofnear-synchronousburstsofexcitatorysynaptic

drive,targetingthedendritesofpyramidalneuronsinspecificcorticallayers[22].Whether

thisphysiologicalpropertydovetailswiththefeaturesofspecificbrainregionsthatarethe

purportedneuralgeneratorsofthescalp-recordeddatareportedhereremainstobetested.

Second,ifβ-burstingoversensorimotorsitesisindeedindicativeofa‘tonic’inhibitorymode

ofthemotorsystem–onethatisreducedduringmovementinitiationandre-initiatedduring

movementcancellation–itwouldmakeitaninterestingtargetforinvestigationsoftonic,

preparatory control activity, e.g., that found in proactive inhibition tasks [40-45]. For

example,increasesinproactiveinhibitionintaskcontextswithhigherrelativelikelihoodsof

Stop-signalscouldresultinincreasedbilateralsensorimotorβ-bursting.Third,thereisgreat



https://doi.org/10.1101/644682


16

interestinpathologicalfeaturesoftheβ-frequencybandinmovementdisorders,especially

Parkinson’sdisease[16-19].Todate,moststudiesofβ-bandactivityinPDarestillbasedon

trial-averages.However,recentstudieshavealreadyusedtrial-levelβ-burstmeasurements

in subcortical areas of the basal ganglia to identify gait problems in PD [46] and have

investigatedtheeffectofdeep-brainstimulationonsubcorticalβ-burst[47].Futurestudies

may aim to investigate the relationshipbetween abnormal patterns of β-bursting on the

scalpandthespecificimpairmentsinmovementcancellationthatarecommonlyobserved

inPD[48-50].Fourth,β-burstsmayprovideanewwindowintotheinteractionsbetweenthe

subcorticalaspectsoftheextrapyramidalmotorsystemandthecorticalaspectsunderlying

higherlevelsofmovement-planningand–control.Severalstudieshavealreadydescribedβ-

bursting in several basal ganglia nuclei [17, 20, 21, 51]. It is tempting to assume a

correspondencebetweenβ-burstinginthesebrainregionsandthecorticalregionsthatare

likelyunderlyingthepatternsobservedinthecurrentstudy. Indeed,β-burstingmaybea

‘universal’languageofthemotorsystem,signifyingdistributedprocessingthroughoutthe

bothpyramidalandextrapyramidalmotorpathways.Finally,investigatingfronto-centralβ-

burstscouldprovideafruitfultest-bedforstudiesofmotorinhibitionacrossmanydifferent

psychological paradigms. As has been proposed elsewhere, inhibitory control may be a

ratheruniversalcontrolmechanism,whichcouldbeinvolvedinregulatingbehaviorafter

unexpectedevents[52-54],response-conflict[55,56],errorprocessing[57,58],andmany

othercontrolprocesses.Thecurrentstudysuggeststhatevenalready-existingdatasetsof

experimentaltasksthatoperationalizetheseprocessesandmeasuretheneurophysiological

indicesmaybeworthre-investigatingwithafocusonβ-bandbursting.



https://doi.org/10.1101/644682


17

Given its nature as the first investigation of single-trial β-bursting during human

movement cancellation (and –initiation), the current study cannot answer all potential

questions,andhasseveralshortcomings.First,sincetheneuralactivityinthecurrentstudy

was non-invasively recorded from the scalp, only very limited conclusions can bemade

regarding the exact origins of the observed β-bursts. Second,while the fronto-central β-

bursts observed in the current study are clearly related to the success of movement

cancellation,theyarealsoclearlyneithernecessarynorsufficient.Infact,Figure2cshows

that even failed Stop-trials include fronto-central β-bursts (though at a reduced rate).

Furthermore,thefigurealsoshowsthatthemajorityofsuccessfulStop-trialsdidnotinclude

aβ-burstevent(atleastinthepre-SSRTperiod).Onepossiblereasonforthisisthattypically,

not all Stop-trials in the Stop-signal task actually involve the initiation of aGo-response,

whichisarequirementformotorinhibition[44,59,60].Thisisespeciallytrueontrialson

whichtheStop-signaldelayisveryshort,givinglittletimeforsignificantmovementinitiation

todevelop.Whilethisexplanationislikelyonlypartofthestory,anexploratoryanalysisof

our data confirms this impression, aswedid indeed find that successful Stop-trialswith

fronto-centralβ-burstshad longerStop-signaldelayscomparedtotrialswithoutβ-bursst

(t(229)=2.27,p=.024,d=.06)1.

In summary, we here provide the first report of β-burst activity underlying both

movementinitiationandcancellationinhumans.Futurestudiesshouldinvestigatetheexact

propertiesoftheseβ-burstsastheyrelatetohumanandnon-humanbehavior,movement

disorders,andtheneuralbasisofmovementcancellationincognitivecontrol.

1Notethattheeffectsizeisverysmall.ThisislikelyaneffectoftheadaptiveStop-signaldelay,whichseverelylimitsthevarianceinthisanalysis.



https://doi.org/10.1101/644682


18

Methods

Participants

234healthy adulthumans (meanage:22.7, SEM: .43, 137 female, 25 left-handed)

fromtheIowaCitycommunityparticipatedinthestudy,eitherforcoursecreditorforan

hourlypayment.123ofthosedatasetswerepublishedaspartofotherstudies,noneofwhich

focused on β-bursting [54, 61-63]. All procedures were approved by the local ethics

committeeattheUniversityofIowa(IRB#201511709).

Task

Thetaskwasidenticaltotheonedescribedin[54,61-63].Inshort,trialsbeganwith

afixationcross(500msduration),followedbyawhiteleft-orrightwardarrow(Go-signal).

Participantswereinstructedtorespondasfastandaccuratelyaspossibletothearrowusing

theirleftorrightindexfinger(therespectiveresponse-buttonswereqandponaQWERTY

keyboard).Onone-thirdoftrials,aStop-signaloccurred(thearrowturnedfromwhitetored)

atadelayafterthego-stimulus(stop-signaldelay,SSD).TheSSD,whichwasinitiallysetto

200ms, was dynamically adjusted in 50ms increments to achieve a p(stop) of .5: after

successfulstops,theSSDwasprolonged;afterfailedstops,itwasshortened.Thiswasdone

independently for left-andrightwardgo-stimuli.Trialdurationwas fixedat3,000ms.Six

blocksof50trialswereperformed(200Go,100Stop).

Dataavailability



https://doi.org/10.1101/644682


19

Alldata,procedures,andanalysisroutinescanbedownloadedontheOpenScience

Frameworkat[URLtobeinsertedafteracceptance].

Behavioralanalysis

Meanswereextractedforeachsubjectforthefollowingmeasures:Go-trialreaction

time, failedStop-trial reaction time,Stop-signaldelay,Stoppingaccuracy,andStop-signal

reactiontime,whichwascalculatedviatheintegrationmethod[24].

EEGrecording

Scalp-EEG was recorded using two different BrainProducts (Garching, Germany)

systems – one active (actiChamp) and one passive (MR plus). In both cases, 62-channel

electrodecapswithtwoadditionalelectrodesontheleftcanthus(overthelateralpartofthe

orbitalboneofthelefteye)andoverthepartoftheorbitalbonedirectlybelowthelefteye

wereused.ThegroundwasplacedatelectrodeFz,andthereferencewasplacedatelectrode

Pz.EEGwasdigitizedatasamplingrateof500Hz,withhardwarefilterssetto10stime-

constanthigh-passand1000Hzlow-pass.

EEGdatapreprocessing

DatawerepreprocessedusingcustomroutinesinMATLAB,incorporatingfunctions

from the EEGLAB toolbox [64]. The electrode * time-seriesmatrices for each taskwere

imported into MATLAB and then filtered using symmetric two-way least-squares finite

impulseresponsefilters(high-passcutoff: .3Hz,low-passcutoff:30Hz).Non-stereotyped

artifactswereautomaticallyremovedfromfurtheranalysisusingsegmentstatisticsapplied



https://doi.org/10.1101/644682


20

toconsecutiveone-secondsegmentsofdata[jointprobabilityandjointkurtosis,withboth

cutoffs set to 5 SD, cf., 65]. After removal of non-stereotypic artifacts, the datawere re-

referenced to common average and subjected to a temporal infomax ICA decomposition

algorithm[66],withextensiontosubgaussiansources[67].Theresultingcomponentmatrix

was screened for components representing eye-movement and electrode artifacts using

outlierstatisticsandnon-dipolarcomponents[residualvariancecutoffat15%,68],which

wereremovedfromthedata.Theremainingcomponentsweresubjectedtofurtheranalyses.

Forallstatisticalanalysesandallplots(exceptthetopographicalplotsinFigure1aand2a),

thedatawere subsequently re-referencedusing the current-sourcedensitymethod [69],

whichminimizestheeffectsofvolumeconductiononthescalp-measuredactivity.Thiswas

donetopermitβ-eventdetectionatfronto-centralelectrodeFCzandlateralsensorimotor

electrodesC3andC4withoutcross-contaminationbyeitherside.

β-burstdetection

β-burstdetectionwasperformedasdescribedinShinandcolleagues’work[3].The

descriptionisadaptedfromtherein.

First,eachelectrode’sdatawasconvolvedwithacomplexMorletwaveletoftheform:

𝑤 �, 𝑓 = 𝐴 exp 𝑡+

2𝜎.+ exp(2𝑖𝜋𝑓𝑡)

with𝜎 = 3+45

,𝐴 = 678

2𝜋,

andm=7(cycles)foreachof15evenlyspacedfrequenciesspanningtheβ-band(15-29Hz).

Time-frequencypowerestimateswereextractedbycalculatingthesquaredmagnitudeof

thecomplexwavelet-convolveddata.Thesepowerestimateswerethenepochedrelativeto



https://doi.org/10.1101/644682


21

theeventsinquestion(rangingfrom-500to+1,000mswithrespecttoStop-/Go-signals).

Individualβ-burstsweredefinedaslocalmaximainthetrial-by-trialβ-bandtime-frequency

powermatrixforwhichthepowerexceededasetcutoffof6xthemedianpoweroftheentire

time-frequency powermatrix for that electrode. Localmaximawere identified using the

MATLABfunctionimregional().

Topographicaldistributionofβ-bursts(Figures1a,2a,3)

Tovisualizethetopographicaldistributionofβ-burstswithrespecttoStop-andGo-

signals,12windowsof25mslength,startingat25msaftertheevent,weredefined.Foreach

subject,thenumberofβ-burstsineachwindowateachelectrodefollowingtherespective

stimulus(Go/Stop)wascounted.Theaveragenumberofβ-burstsineachtimewindowfor

eachofthethreetrialtypes(CorrectGo,SuccessfulStop,FailedStop)wasthenplottedina

topographicalgridrepresentingthescalpsurface(Figures1aand2a).Figure3depictsthe

differencebetweenthenumberofβ-burstsonGo-andStop-trials ineachwindowandat

eachelectrode,testedforsignificanceusingpaired-samplest-tests.Theresultingelectrode

*timewindowmatrixofp-valueswascorrectedformultiplecomparisonstoasignificance

levelofp=.0001usingthefalse-discoveryrateprocedure(FDR,[70]).

Temporaldevelopmentofβ-bursts(Figure1b,2b,2d)

Tovisualizethetemporaldevelopmentofβ-burstsonGo-trialsatthetwoelectrodes

ofinterest(C3andC4),11windowsof50mslengthrangingfrom25mspost-eventto575ms

post-eventweredefined.Thistimerangespannedtheentirepost-Go-signalperiodleading

up tomean reaction time. To test the linear decreasing trend observed at C3/C4 during



https://doi.org/10.1101/644682


22

movement initiation for significance, we submitted the means for left- and right-hand

responsestotheMann-Kendalltest.Totestthelateralizationofthelineartrendtowardsthe

endof theresponseperiod(i.e.,prior tomeanRT),wecomparedthemeans for left-and

right-hand responsesatbothelectrodesusingpaired-samples t-tests, again corrected for

multiple comparisons to a significance level of p = .0001 using the false-discovery rate

procedure.Tovisualizethetemporaldevelopmentofβ-burstsonStop-trialsattheelectrode

ofinterest(FCz),6windowsof50mslengthrangingfrom25mspost-eventto325mspost-

eventweredefined.Thistimerangespannedtheentirepost-Stop-signalperiodleadingup

toStop-signalreactiontime.

Comparisonofpre-SSRTβ-events(Figure2c)

Totestthedifferenceintheamountofβ-burstsatfronto-centralelectrodeFCzfor

significance,wecountedthenumberofβ-burstsinthetimeperiodrangingfromtheStop-

signal to each individualparticipants’ SSRTestimate, separately for successful and failed

Stop-trials.Themeanswerecomparedacrosssubjectsusingapaired-samplest-tests.

Lateralizedβ-burstingafterfronto-centralβ-bursts(Figure4)

To investigate the pattern of β-bursting over bilateral electrodes C3/4 following

fronto-central β-bursts in the Stop-Signal-to-SSRT period on successful Stop-trials, we

identifiedeachtrialinwhichsuchafronto-centralβ-bursteventoccurred,andcountedthe

amount of β-bursts at electrodes C3/4 both contra- and ipsilateral to the to-be-stopped

responseineighttimewindowsof25msdurationrangingfrom-100to+100msaroundthe

fronto-centralβ-burstevent.Incasemorethanoneβ-bursteventwasfound,wechosethe



https://doi.org/10.1101/644682


23

latencyofthefirstofthoseevents.Tocomparetheseβ-burstcountstotrials inwhichno

fronto-centralβ-burstwasfound,arandomtime-pointintheStop-signal-to-SSRTinterval

waschosenfromauniformdistributionandC3/4β-burstswerequantifiedinanidentical

timewindowaroundthatrandomtimepointinthepre-SSRTperiod.Wethencomparedthe

meansburst-countsforthetwoconditionsinthefourtime-windowsfollowingtheevent(or

the ‘pseudo-event’ in the case of the trial without an actual fronto-central burst) using

signed-rank tests (the non-parametrical equivalent of the paired-samples t-test, chosen

because of the large skew of the means towards zero in these samples), corrected for

multiplecomparisonstoacriticalp-valueofp<.0001usingtheFDR-method.

Acknowledgements

Theauthorwouldliketothankthefollowingresearchassistantsthatwereinvolved

incollectingthesedata:DanielThayer,JulianScheffer,CaileyParker,HaileyBillings,Carly

Ryder,MeganHynd,Carly Iacullo,BrynneDochterman, IsabellaDutra,AlecMather,Kylie

Dolan, Nathan Chalkley, Darcy Waller, Tobin Dykstra, and Cheol Soh. Furthermore, the

authorwouldliketothankMarkBlumberg,JohnFreeman,andBobMcMurrayforhelpful

discussions.This researchwassupportedbygrants from theNational InstituteofHealth

(R01NS102201),theNationalScienceFoundation(NSFCAREER1752355),andtheRoyJ

CarverFoundation(JuniorResearchProgramofExcellence).



https://doi.org/10.1101/644682


24

References

1.Murthy,V.N.andFetz,E.E.(1992)Coherent25-to35-Hzoscillationsinthesensorimotor

cortexofawakebehavingmonkeys.ProcNatlAcadSciUSA89(12),5670-4.

2.Sanes,J.N.andDonoghue,J.P.(1993)Oscillationsinlocalfieldpotentialsoftheprimate

motorcortexduringvoluntarymovement.ProcNatlAcadSciUSA90(10),4470-4.

3.Shin,H.etal.(2017)Therateoftransientbetafrequencyeventspredictsbehavioracross

tasksandspecies.Elife6.

4.Engel,A.K.andFries,P.(2010)Beta-bandoscillations--signallingthestatusquo?CurrOpin

Neurobiol20(2),156-65.

5. Pfurtscheller, G. et al. (2003) Spatiotemporal patterns of beta desynchronization and

gamma synchronization in corticographic data during self-paced movement. Clin

Neurophysiol114(7),1226-36.

6.McFarland,D.J.etal.(2000)Muandbetarhythmtopographiesduringmotorimageryand

actualmovements.BrainTopogr12(3),177-86.

7. Swann, N. et al. (2011) Deep brain stimulation of the subthalamic nucleus alters the

cortical profile of response inhibition in the beta frequency band: a scalp EEG study in

Parkinson'sdisease.JNeurosci31(15),5721-9.

8.Swann,N.etal.(2009)IntracranialEEGrevealsatime-andfrequency-specificroleforthe

right inferior frontal gyrus and primary motor cortex in stopping initiated responses. J

Neurosci29(40),12675-85.

9.Swann,N.C.etal.(2012)Rolesforthepre-supplementarymotorareaandtherightinferior

frontal gyrus in stopping action: electrophysiological responses and functional and

structuralconnectivity.Neuroimage59(3),2860-70.



https://doi.org/10.1101/644682


25

10.Wagner,J.etal.(2018)EstablishingaRightFrontalBetaSignatureforStoppingActionin

Scalp EEG: Implications for Testing Inhibitory Control in Other Task Contexts. J Cogn

Neurosci30(1),107-118.

11.Picazio,S. etal. (2014)Prefrontal controlovermotorcortexcyclesatbeta frequency

duringmovementinhibition.CurrBiol24(24),2940-5.

12. Kuhn, A.A. et al. (2004) Event-related beta desynchronization in human subthalamic

nucleuscorrelateswithmotorperformance.Brain127(Pt4),735-46.

13.Ray,N.J.etal.(2012)Theroleofthesubthalamicnucleusinresponseinhibition:evidence

fromlocalfieldpotentialrecordingsinthehumansubthalamicnucleus.Neuroimage60(1),

271-8.

14.Wessel, J.R. et al. (2016) Stop-related subthalamic beta activity indexes globalmotor

suppressioninParkinson'sdisease.MovDisord31(12),1846-1853.

15.Brittain,J.S.andBrown,P.(2014)Oscillationsandthebasalganglia:motorcontroland

beyond.Neuroimage85Pt2,637-47.

16.Hammond,C.etal.(2007)PathologicalsynchronizationinParkinson'sdisease:networks,

modelsandtreatments.TrendsNeurosci30(7),357-64.

17.Jenkinson,N.andBrown,P.(2011)Newinsightsintotherelationshipbetweendopamine,

betaoscillationsandmotorfunction.TrendsNeurosci34(12),611-8.

18. Bronte-Stewart, H. et al. (2009) The STN beta-band profile in Parkinson's disease is

stationaryandshowsprolongedattenuationafterdeepbrainstimulation.ExpNeurol215

(1),20-8.

19. Quinn, E.J. et al. (2015) Beta oscillations in freely moving Parkinson's subjects are

attenuatedduringdeepbrainstimulation.MovDisord30(13),1750-8.



https://doi.org/10.1101/644682


26

20.Feingold,J.etal.(2015)Burstsofbetaoscillationdifferentiatepostperformanceactivity

inthestriatumandmotorcortexofmonkeysperformingmovementtasks.ProcNatlAcad

SciUSA112(44),13687-92.

21.Leventhal,D.K.etal. (2012)Basalgangliabetaoscillationsaccompanycueutilization.

Neuron73(3),523-36.

22.Sherman,M.A.etal.(2016)Neuralmechanismsoftransientneocorticalbetarhythms:

Convergingevidencefromhumans,computationalmodeling,monkeys,andmice.ProcNatl

AcadSciUSA113(33),E4885-94.

23. Logan, G.D. and Cowan,W.B. (1984)On the Ability to Inhibit Thought andAction: A

TheoryofanActofControl.PsychRev91,295-327.

24.Verbruggen,F.etal.(2019)Aconsensusguidetocapturingtheabilitytoinhibitactions

andimpulsivebehaviorsinthestop-signaltask.Elife8.

25.Boucher,L.etal.(2007)Inhibitorycontrolinmindandbrain:aninteractiveracemodel

ofcountermandingsaccades.PsycholRev114(2),376-97.

26.Verbruggen,F.andLogan,G.D.(2008)Responseinhibitioninthestop-signalparadigm.

TrendsCognSci.

27.Rossiter,H.E.etal.(2014)Betaoscillationsreflectchangesinmotorcortexinhibitionin

healthyageing.Neuroimage91,360-5.

28.Doyle,L.M.etal.(2005)Lateralizationofevent-relatedbetadesynchronizationintheEEG

duringpre-cuedreactiontimetasks.ClinNeurophysiol116(8),1879-88.

29.Kaiser,J.etal.(2001)Event-relatedbetadesynchronizationindicatestimingofresponse

selectioninadelayed-responseparadigminhumans.NeurosciLett312(3),149-52.



https://doi.org/10.1101/644682


27

30. Kok, A. et al. (2004) ERP components associated with successful and unsuccessful

stoppinginastop-signaltask.Psychophysiology41(1),9-20.

31.Wessel, J.R. andAron, A.R. (2015) It's not too late: the onset of the frontocentral P3

indexessuccessfulresponseinhibitioninthestop-signalparadigm.Psychophysiology52(4),

472-80.

32. Kenemans, J.L. (2015) Specific proactive and generic reactive inhibition. Neurosci

BiobehavRev56,115-26.

33. Swann, N.C. et al. (2011) Roles for the pre-supplementarymotor area and the right

inferiorfrontalgyrusinstoppingaction:Electrophysiologicalresponsesandfunctionaland

structuralconnectivity.Neuroimage.

34. Schmidt, R. et al. (2013) Canceling actions involves a race between basal ganglia

pathways.NatureNeuroscience16(8),1118-24.

35.Schall,J.D.etal.(2017)Modelsofinhibitorycontrol.PhilosTransRSocLondBBiolSci

372(1718).

36.Badry,R.etal.(2009)Suppressionofhumancortico-motoneuronalexcitabilityduring

theStop-signaltask.ClinNeurophysiol120(9),1717-23.

37. Majid, D.S. et al. (2012) Transcranial magnetic stimulation reveals dissociable

mechanismsforglobalversusselectivecorticomotorsuppressionunderlyingthestopping

ofaction.CerebralCortex22(2),363-71.

38.Wessel,J.R.etal.(2013)Saccadesuppressionexertsglobaleffectsonthemotorsystem.

JournalofNeurophysiology.

39.Duque,J.etal.(2017)PhysiologicalMarkersofMotorInhibitionduringHumanBehavior.

TrendsNeurosci40(4),219-236.



https://doi.org/10.1101/644682


28

40.Elchlepp,H.etal.(2016)Proactiveinhibitorycontrol:Ageneralbiasingaccount.Cogn

Psychol86,27-61.

41.Greenhouse,I.etal.(2012)Stoppingaresponsehasglobalornonglobaleffectsonthe

motorsystemdependingonpreparation.JournalofNeurophysiology107(1),384-92.

42. Jaffard,M. et al. (2008)Proactive inhibitory control ofmovement assessedby event-

relatedfMRI.Neuroimage42(3),1196-206.

43.Stuphorn,V.andEmeric,E.E.(2012)Proactiveandreactivecontrolbythemedialfrontal

cortex.FrontNeuroeng5,9.

44.Verbruggen,F.andLogan,G.D.(2009)Proactiveadjustmentsofresponsestrategiesin

thestop-signalparadigm.JExpPsycholHumPerceptPerform35(3),835-54.

45.Vink,M.etal. (2015)Theroleof stop-signalprobabilityandexpectation inproactive

inhibition.EurJNeurosci41(8),1086-94.

46.Anidi,C.etal.(2018)Neuromodulationtargetspathologicalnotphysiologicalbetabursts

duringgaitinParkinson'sdisease.NeurobiolDis120,107-117.

47.Tinkhauser,G.etal.(2017)Themodulatoryeffectofadaptivedeepbrainstimulationon

betaburstsinParkinson'sdisease.Brain140(4),1053-1067.

48.Gauggel, S. et al. (2004) Inhibitionof ongoing responses inpatientswithParkinson's

disease.JNeurolNeurosurgPsychiatry75(4),539-44.

49.Obeso,I.etal.(2011)Deficitsininhibitorycontrolandconflictresolutiononcognitive

andmotortasksinParkinson'sdisease.ExpBrainRes212(3),371-84.

50.vandenWildenberg,W.P.M.etal.(2006)Stimulationofthesubthalamicregionfacilitates

the selection and inhibitionofmotor responses inParkinson’s disease. Journal Cognitive

Neuroscience.



https://doi.org/10.1101/644682


29

51. Bartolo, R. and Merchant, H. (2015) beta oscillations are linked to the initiation of

sensory-cued movement sequences and the internal guidance of regular tapping in the

monkey.JNeurosci35(11),4635-40.

52.Wessel,J.R.(2018)Surprise:AMoreRealisticFrameworkforStudyingActionStopping?

TrendsCognSci22(9),741-744.

53.Wessel, J.R.andAron,A.R.(2017)OntheGlobalityofMotorSuppression:Unexpected

EventsandTheirInfluenceonBehaviorandCognition.Neuron93(2),259-280.

54.Dutra,I.etal.(2018)Perceptualsurpriseimprovesactionstoppingbynon-selectively

suppressingmotoractivityviaaneuralmechanismformotorinhibition.JNeurosci.

55. Frank, M.J. et al. (2007) Hold your horses: impulsivity, deep brain stimulation, and

medicationinparkinsonism.Science318(5854),1309-12.

56.Wessel, J.R. et al. (2019) Non-selective inhibition of inappropriate motor-tendencies

duringresponse-conflictbyafronto-subthalamicmechanism.Elife8.

57.Wessel,J.R.(2018)Anadaptiveorientingtheoryoferrorprocessing.Psychophysiology

55(3).

58. Ridderinkhof, K.R. (2002) Micro- and macro-adjustments of task set: activation and

suppressioninconflicttasks.PsycholRes66(4),312-23.

59.Chikazoe, J. et al.,Preparation to inhibit a responsecomplements response inhibition

duringperformanceofastop-signaltask,JNeurosci,2009,pp.15870-7.

60.Bissett,P.G.andLogan,G.D. (2012)Post-stop-signaladjustments: inhibition improves

subsequentinhibition.JExpPsycholLearnMemCogn38(4),955-66.

61.Wessel,J.R.(2017)Prepotentmotoractivityandinhibitorycontroldemandsindifferent

variantsofthego/no-goparadigm.Psychophysiology.



https://doi.org/10.1101/644682


30

62. Wessel, J.R. (2016) A Neural Mechanism for Surprise-related Interruptions of

VisuospatialWorkingMemory.CerebCortex.

63. Waller, D.A. et al. (2019) Common neural processes during action-stopping and

infrequentstimulusdetection:ThefrontocentralP3asanindexofgenericmotorinhibition.

IntJPsychophysiol.

64.Delorme,A.andMakeig,S.(2004)EEGLAB:anopensourcetoolboxforanalysisofsingle-

trialEEGdynamicsincludingindependentcomponentanalysis.JNeurosciMethods134(1),

9-21.

65.Delorme,A.etal.(2007)EnhanceddetectionofartifactsinEEGdatausinghigher-order

statisticsandindependentcomponentanalysis.Neuroimage34(4),1443-9.

66. Bell, A.J. and Sejnowski, T.J. (1995) An information-maximization approach to blind

separationandblinddeconvolution.NeuralComput7(6),1129-59.

67. Lee, T.W. et al. (1999) Independent component analysis using an extended infomax

algorithmformixedsubgaussianandsupergaussiansources.NeuralComput11(2),417-41.

68.Delorme,A.etal.(2012)IndependentEEGsourcesaredipolar.PLoSOne7(2),e30135.

69.Tenke,C.E.andKayser,J.(2005)Reference-freequantificationofEEGspectra:combining

current source density (CSD) and frequency principal components analysis (fPCA). Clin

Neurophysiol116(12),2826-46.

70. Benjamini, Y. et al. (2006) Adaptive linear step-up procedures that control the false

discoveryrate.Biometrika93(3),491-507.



https://doi.org/10.1101/644682


β-bursts reveal the trial-to-trial dynamics of movement ... · distribution of the number of...

Documents