www.bjorl.org
Brazilian
Journal
of
OTORHINOLARYNGOLOGY
ORIGINAL
ARTICLE
Automated
cortical
auditory
evoked
potentials
threshold
estimation
in
neonates
夽
Lilian
Sanches
Oliveira
a,
Dayane
Domeneghini
Didoné
b,
Alessandra
Spada
Durante
a,∗aFaculdadedeCiênciasMédicasdaSantaCasadeSãoPaulo,EscoladeFonoaudiologiaeAudiologia,SãoPaulo,SP,Brazil bUniversidadeFederaldoRioGrandedoSul,ProgramadePós-graduac¸ãoemSaúdedaCrianc¸aedoAdolescente,RioGrandedo
Sul,RS,Brazil
Received4September2017;accepted2January2018 Availableonline2February2018
KEYWORDS Audiology; Auditoryevoked potentials; Infant,newborn; Electrophysiology Abstract
Introduction:Theevaluationofcorticalauditoryevokedpotentialhasbeenthefocusof sci-entificstudiesininfants.Some authorshavereportedthatautomatedresponsedetectionis effectiveinexploringthesepotentialsininfants,butfewhavereportedtheirefficacyinthe searchforthresholds.
Objective:Toanalyzethelatency,amplitudeandthresholdsofcorticalauditoryevoked poten-tialusinganautomaticresponsedetectiondeviceinaneonatalpopulation.
Methods:Thisisacross-sectional,observationalstudy.Corticalauditoryevokedpotentialswere recordedinresponsetopure-tonestimuliofthefrequencies500,1000,2000and4000Hz pre-sentedinanintensityrangebetween0and80dBHLusingasinglechannelrecording.P1was performedinanexclusivelyautomatedfashion,usingHotelling’sT2statisticaltest.Thelatency
andamplitudewereobtainedmanuallybythreeexaminers.Thestudycomprised39neonates upto28daysoldofbothsexeswithpresenceofotoacousticemissionsandnoriskfactorsfor hearingloss.
Results:Withtheprotocolused,corticalauditoryevokedpotentialresponsesweredetected inallsubjectsathighintensityandthresholds.Themeanthresholdswere24.8±10.4dBNA, 25±9.0dBNA,28±7.8dBNAand29.4±6.6dBHLfor500,1000,2000and4000Hz, respec-tively.
Conclusion:Reliableresponseswereobtainedintheassessmentofcorticalauditorypotentials intheneonatesassessedwithadeviceforautomaticresponsedetection.
© 2018 Associac¸˜ao Brasileira de Otorrinolaringologia e Cirurgia C´ervico-Facial. Published by Elsevier Editora Ltda. This is an open access article under the CC BY license (http:// creativecommons.org/licenses/by/4.0/).
夽 Pleasecitethisarticleas:OliveiraLS,DidonéDD,DuranteAS.Automatedcorticalauditoryevokedpotentialsthresholdestimationin
neonates.BrazJOtorhinolaryngol.2019;85:206---12.
∗Correspondingauthor.
E-mail:alessandra.durante@fcmsantacasasp.edu.br(A.S.Durante).
PeerReviewundertheresponsibilityofAssociac¸ãoBrasileiradeOtorrinolaringologiaeCirurgiaCérvico-Facial.
https://doi.org/10.1016/j.bjorl.2018.01.001
1808-8694/©2018Associac¸˜aoBrasileiradeOtorrinolaringologiaeCirurgiaC´ervico-Facial.PublishedbyElsevierEditoraLtda.Thisisanopen accessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).
PALAVRAS-CHAVE Audiologia; Potenciaisevocados auditivos; Lactente, recém-nascido; Eletrofisiologia
Estimativadolimiardepotenciaisevocadosauditivoscorticaisautomatizadosem recém-nascidos
Resumo
Introduc¸ão: Opotencialevocadoauditivocorticaltemsidoofocodeestudoscientíficos.Alguns autoresobservaramqueadetecc¸ãoautomatizadaderespostaséeficaznaexplorac¸ãodesses potenciaisemlactentes,maspoucosrelataramsuaeficácianabuscadelimiares.
Objetivo: Analisaralatência,aamplitudeeoslimiaresdopotencialevocadoauditivocortical emrecém-nascidos,comousodeumdispositivodedetecc¸ãoautomáticaderesposta.
Método: Estudotransversal,observacional.Ospotenciaisevocadosauditivoscorticais foram registradosemrespostaaestímulosdetonspurosnasfrequênciasde500,1000,2000e4000Hz eapresentadosemumafaixadeintensidadeentre0---80dBNA,comousodegravac¸ãodecanal único.OP1foifeitodeformaexclusivamenteautomática,comousodotesteestatísticoT2
deHotelling.Alatênciaeaamplitudeforamobtidasmanualmenteportrêsexaminadores.O estudoincluiu39recém-nascidoscomaté28diasdeidadedeambosossexos,compresenc¸a deemissõesotoacústicasesemfatoresderiscoparaperdaauditiva.
Resultados: Comoprotocolousado,asrespostasdosPEACforamdetectadasemtodosos indi-víduosemaltaintensidadeelimiares.Oslimiaresmédiosforam24,8±10,4dBNA,25±9,0 dBNA,28±7,8dBNAe29,4±6,6dBNApara500,1000,2000e4000Hz,respectivamente.
Conclusão:Foramobtidasrespostasconfiáveisnaavaliac¸ãodospotenciaisauditivoscorticais emrecém-nascidoscomumdispositivoparadetecc¸ãoderespostaautomática.
© 2018 Associac¸˜ao Brasileira de Otorrinolaringologia e Cirurgia C´ervico-Facial. Publicado por Elsevier Editora Ltda. Este ´e um artigo Open Access sob uma licenc¸a CC BY (http:// creativecommons.org/licenses/by/4.0/).
Introduction
Hearingisakeyfunctionsupportingthecommunication pro-cessamongindividuals.Itisthroughhearingthatachildcan experiencetheworldofsound,promotingthedevelopment ofspokenlanguage.Theanatomicalandfunctionalintegrity of the peripheral and central auditory system, together withexposuretoauditoryexperiences,arebasicrequisites for thenormalacquisition anddevelopmentof language.1
Thus,assuring reliableestimates of auditorythresholdsin the infants is paramount. However, accurately investigat-ing hearingin the firstmonths of lifehas always posed a challengeforaudiologists.
Inthefirstmonthsoflife,itisnotpossibletoaccurately determine auditory thresholds by observation of behav-ioralresponsestosoundstimulialone.Therefore,objective measures of hearing are essential tools to properly ver-ify subjective observationsand integrate this cross-check investigation.OneimportantobjectivemeasureisAuditory EvokedPotential (AEP). AEP resultsfromneural activities in the auditory pathways in response toa sound stimulus andsometypesofAEPareusefulforestablishingauditory thresholdsininfants.2
Neuralelectrical activity generatedby acousticstimuli can be detected at many different levelsof theauditory pathway,accordingtotheresponselatencyinrelationtothe stimuli(milliseconds---ms).Theseresponsescanbe classi-fiedintothreegroups:shortlatencyAEP,i.e.thoseoccurring withinthe first10---12ms;medium latency AEP,generated within12---50ms,andlong latency(or cortical)AEPwhich occurat50---600ms.3
CorticalAuditoryEvokedPotentials(CAEP)were discov-eredin the1930sand thoroughly researchedin the 1960s and1970s.
The CAEP arerepresented, in adults, by a complexof waves called P1, N1 andP2. Innormal hearing neonates, CAEPresponseoccursatamarkedpositivepeakataround 200---300msafteracousticstimulus,4 withchangesin form
and latency of components taking place during the first 14---16yearsoflife.Morphologicalchangesandmaturation oftheCentralNervousSystem(CNS)improvesynaptic effec-tiveness and are responsible for these shifts in response latenciesduringthefirstyearsoflife.5,6
ThedetectionofCAEPhasnumerousbenefits,including theabilitytoassessthewholeauditorysystem,i.e.fromthe brainstemtotheauditorycortex.Anotheradvantageisthe examcanbeperformedwiththesubjectawake,i.e.inolder children,usinginsertearphonesorinfreefield,makingthe testmorewidelyapplicableandattractive.7,8Becauseitis
anexogenouspotential,theP1componentisrelatedtothe detectionoftheacousticstimulusintheprimaryauditory cortexand,becauseitiswidelythoughtnottobeaffected bysleep,canbeusedwithwakingorsleepingsubjects.9
The need for accurate, reliable assessments, coupled with the major technological developments in recent decades,haspavedthe wayfor progressin studiesinthis area.InAustralia,TheNationalAcousticLaboratories(NAL), theresearchdivisionof thestatutory authority Australian Hearing developed a device to record Cortical Auditory EvokedPotentials(CAEPs)andaprotocoltoobjectively ana-lyze the responses called the HearLab System, which has highsensitivityfordetectingresponses,reducingnoiseand
artifacts. In addition, the detection of responses is per-formed in an automated fashion by the device, thereby reducingexaminersubjectivity.Inearlylatencyresponses, the evoked potentials are relatively stable, but in late latencyevokedpotentials,detectioncanbeimpairedbythe instability of the true evoked potential, aswell as resid-ualnoise.10 CAEPs inyoungchildrenshowmorevariability
thanthoseofadultsbecauseofincreased electrophysiolog-icalnoise.11 Asaresult,thecommonmethodforresponse
detection,withvisualobservationofresponse,aplausible latency for keyresponse componentsand response track-ing(i.e.,increasedlatencyanddecreasedamplitudeofthe responsewithdecreasingstimuluspresentationlevels)may beinadequateinthispopulation.So,methodsthatreduce residualnoise,suchasthosepresentintheHearLabSystem, areneeded.10
Some authors have shown that automated response detection can make a valuable contribution in exploring these potentials in the infant population, but few have reportedtheirefficacyinestablishingthresholds.12---15
Thisstudyisparticularlyrelevantinthecurrentcontext of infant audiology assessment as no studies using auto-mated CAEP analyzers or standardizing responses in the neonatalpopulationareavailable.
Therefore,ourobjectiveistoanalyzetheCAEPinterms oflatencies,amplitudesat80dBHLandthresholdsofCAEP at the frequencies 500, 1000, 2000 and 4000Hz using an automaticresponsedetectiondeviceinneonateswith pres-enceofotoacousticemissionsandnoriskfactorsforhearing loss.
Methods
ThemethodwassubmittedtoandapprovedbytheResearch EthicsCommitteeofSantaCasadeSaoPaulo(n◦ 951.829). The parents of the participants were informed about the studyobjectivesand,havingagreedtotheneonatestaking part,freelysignedaninformedconsentform.
The subjects included in the study were neonates recruitedfromthematernity wardbetween Februaryand September 2015. The families were invited to take part in the study during their stay in the accommodation unit attachedtothematernityward.Neonateswithupto28days oflifethatpassedtheNeonatalAuditoryScreeningwitha gestationalageofat least37weeksandnoriskindicators for hearing loss, according to the guidelines of the Joint CommitteeonInfantHearing,16wereselected.
Neonatesthathadneurologicalsyndromesor abnormali-ties,pre-andperinatalcomplications,orextremeagitation andexcessmovementduringtheexamprecluding comple-tionoftheassessmentwereexcluded.
Initially, 45 neonates were selected to take part, six ofwhich weresubsequentlyexcluded for notmeetingthe studyinclusioncriteria.Thisgaveafinalstudysampleof39 neonates,comprising19femalesand20males.
CAEPs were recorded in response to pure-tonestimuli withfrequenciesof500,1000,2000and4000Hzpresented inanintensityrangebetween0and80dBHLusingasingle channelrecording,withinsertearphoneER-3A,alternating polarity,andaninter-stimulusintervalof1.125ms.
Priortostudycommencement,thedevicewascalibrated indBHLaccordingtothetechnicalcriteriaestablishedby themanufacturer.
Anotoscope wasusedto inspect theexternal acoustic meatusofthesubjectsassessedtoexcludeexcessearwax andensurethattherewerenocontra-indicationsforuseof insertearphones.
DisposabletypeelectrodeswereaffixedatFpz(ground), Cz (active) and M1 or M2 (reference) positions and the impedancewaskeptatalevelunder5kohms.
Examswereperformedinasound-proofedroomwiththe subjectsplacedinthemother’slaporinanappropriatechair fortheageoftheinfant.Duringtheassessment,newborns remainedinlightsleep.Thebehavioralstateoftheneonates wasmonitoredbytwoexamineraudiologists,whomonitored the newborns throughout the evaluation.Because itis an exogenous potential, the P1 component is related to the detectionof theacousticstimulusin theprimaryauditory cortex and, most studies agree that it has little relation to the stages of sleepor wakefulness.9 Therefore, in the
presentstudyneonatesremainedinlightsleep.
Differential analog amplification was 1210 times using a12dBhigh-passfilterwithoctavesof4000Hzanda6dB low-passfilterwithoctavesbelow3000Hz.Thetoneburst was40ms,withalternatingpolarity, acosine envelope,a 0.5Hzstimulusspeed,arise-fallof10ms,andaplateauof 30ms.Therejectionofartifactswasbasedonthecurrent difference of the active-reference electrodes established by the device. Responses present with a minimum of 50 stimulideliveredwereacceptedwhentheobjective detec-tionstatisticwasp<0.001.Otherwise,aCAEPresponsewas judged tobe present if the p-value reached the level of
p<0.05afterreachingtheacceptednumberof150epochs. Residual noise wascontrolled during allassessments, and the HearLab display indicated the quality of the cortical responserecordedinrelationtothenoiselevelofthesignal. Aresidualnoiselevelvaluelessorequalto3.2Vindicatesa goodqualityrecording;avaluebetween3.2and3.6V indi-catesaslightlycompromised recordingandavaluehigher than3.6Vindicatesapoorqualityrecording.Inthisstudy, the maximum value allowed for noise was 3.6V andfor thisreasonparticipantswithextreme agitationandexcess movementwereexcluded.Theambientnoiseleveldidnot exceed35dBSPL.
Physiological assessment was performed by studying cortical auditory evoked potentials by monaural acoustic stimulation,withthetest earchosen randomly.A totalof 15rightearsand25leftearsweretestedatfrequenciesof 500,1000,2000and4000Hzwithamplitudes,latenciesand electrophysiologicalthresholdsrecordedforeachfrequency assessed.
The stimuli werepresented, with minormodifications, accordingtotheacousticstimulidecisionprotocolproposed by Van Dun et al.17 for greater ease-of-use and speed in
CAEPresponses. The initialmaximum intensity wassetat 80dB HL to assess the integrity of the auditory pathway atthecentrallevel.Subsequently,thestimuluswastested at an intensityof 30dBHL.In theabsenceofresponseat thisintensity,thestimuluswasincreasedin5dBincrements untilthethresholdwasdetected.Intheeventofaresponse at 30dB HL,thestimuluswasdeliveredat an intensity of 15dBHL,5dBHLand0dBHL.Intheabsenceofresponseat
Figure1 An example ofautomatic cortical auditory evoked potentialthreshold estimation to4000Hz. In the example,the equipmentconsideredresponseat80and35dBHL.Theblackcontinuous lineindicatestheP1latencyconsideredbythethree examiners.
5dBHL,thestimuluswasincreasedin 5dBHLincrements untiltheelectrophysiologicthresholdwasattained.All sub-jectsweretestedusingthismodifiedprotocol.Thestimuliin thedifferentfrequencieswerepresentedrandomlyinorder to avoid the habituation of the central auditory nervous system.7
The presence of P1 was objectively tested using Hotelling’s T2 test, a multivariate extension of the
ordi-nary one-sample t-test. Hotelling’s T2 test takes vector
dataandtestsanullhypothesis thatthetruemeanvector equalsaspecifiedvector,inthiscasethezerovector.Each ‘‘datapoint’’wasanine-dimensionalbinnedepoch,andthe nullhypothesisbeingtestedwasthattheaveragedcortical responseineverybinwaszero,whichconsidersresponses presentwhenp<0.05.Thismethodhasshowngreat sensi-tivityandspecificityinthedetectionofcorticalresponses.8
When therewere highlevelsofnoise ordoubts about the P1 responses, the assessment was performed more than once.
The recording of the latency and amplitude of the P1 responses at 80dB HL intensity was performed by three researcherswithexperienceinelectrophysiology,sincethe equipmentdoes notautomaticallydothis.P1was consid-eredinthehighestpeakinlatencyof100---500ms.
Theamplitudeandlatencywereconsideredonlyat80dB HLsincetheobjectiveofthepresentstudywastoverifythe speedandrecruitmentofneuronsoftheauditorycortexata strongintensity.Inaddition,atlowerintensitiesthelatency andamplitudemaybevariable,whichcouldcompromisethe comparisonbetweenthegroups.
Fig.1showsanexampleofanautomatedcortical audi-toryevokedpotentialthresholdestimatesinaneonate.
The Mann---Whitney, Wilcoxon and Repeated-Measures Analysis of Variance (ANOVA) testswere used for statisti-calanalysis.Inalltests,asignificancelevelof0.05(or5%) wasadoptedforrejectionofthenullhypothesis.
Table1 P1latencyvalues(ms)at80dBHLforthe frequen-ciestested. P1latency 500Hz 1000Hz 2000Hz 4000Hz p-value Mean 242.79 225.54 232.74 244.51 Median 243 226 231 241 Standard deviation 51.30 36.31 39.84 46.26 0.411 Minimum 137 125 160 157 Maximum 419 307 353 370 CI 33.27 23.54 25.83 30 N 39 39 39 39
CI,confidenceinterval;N,numberofsubjects.
Results
Averagedurationofanexamwas73.3min,withtheshortest being 38min and the longest 111min. This duration var-iednot only due to the responses but also asa result of movementandagitationofthesubjects.
Latencyand amplitudevaluesat 80dB HLintensity for thefrequenciestestedaregiveninTables1and2, respec-tively,showingnostatisticallysignificantdifferenceamong frequencies.
The mean threshold obtained was 24.8±10.4dB HL, 25±9.0dBHL,28.72±7.84dBHLand29.4±6.6dBHLfor 500,1000,2000 and4000Hz, respectively.Nostatistically significantdifferencesamongthefrequencies testedwere found(Table3).
Fig.2showsthelatencyasafunctionofintensity,being thelatency ofthecomponentP1inverselyproportionalto theintensityoftheacousticstimulus.
Table2 P1amplitudevalues(V)at80dBHLforthe fre-quenciestested. P1amplitude 500Hz 1000Hz 2000Hz 4000Hz p-value Mean 6.41 7.36 6.31 5.88 Median 6.07 7.15 5.43 5.78 Standard deviation 3.39 4.10 3.31 2.83 0.550 Minimum 1.30 2.48 2.13 1.94 Maximum 18.78 21.83 14.49 15.95 CI 2.2 2.66 2.14 1.84 N 39 39 39 39
CI,confidenceinterval;N,numberofsubjects.
Table3 Descriptionofcorticalthresholdsforthe frequen-ciestested(dBHL). Thresholds 500Hz 1000Hz 2000Hz 4000Hz p-value Mean 24.87 25 28.72 29.49 Median 25 30 30 30 Standard deviation 10.41 9.03 7.84 6.66 0.085 Minimum 0 0 0 15 Maximum 40 35 40 50 CI 6.76 5.86 5.08 4.32 N 39 39 39 39
CI,confidenceinterval;N,numberofsubjects.
80 70 60 50 40 30 20 10 0 0 50 100 150 200 Latency Intensity 250 300 350 400 90
Figure2 Representationofthelatencyfunctionofthe corti-calpotentialP1byintensity.
Discussion
The need to increase knowledge on hearing in the first months of life has prompted researchers to dedicate a greatdeal of attention tothe search for newprocedures inthis area.Objectivemethods of assessinghearing have proveneffectivefor thispurpose, particularlyinneonates andyoung infants. With thetechnicaladvances of recent decades,CAEPperformedbyautomateddeviceshasgained importance in this field.However, the paucity of studies using this method and the absence of parameters in the infantpopulationpromptedthepresentstudy.
The effectiveness of automated statistical detection comparedtoexperiencedexaminersindetectingthe pres-ence of infant CAEP has already been studied in other research,evidencingitsreliability.10 The protocol devised
inthisstudyforassessinghearinginneonatesusingadevice forautomaticresponsedetectionofCAEP,accordingtothe modificationsappliedtotheVanDunmodel,17proved
effec-tiveandviableindetectingcorticalresponsesin100%ofthe subjectstested.Theinitialstudyat80dBHLallowed deter-minationofP1latencyandamplitudeinthegroupassessed, enablinganalysisoftheprocessingof soundstimuliat the centrallevel.
Someaudiologistshavedifficultiesindetectingand inter-pretingelectrophysiologicalresponsesbasedexclusivelyon visual analysisofresponses, withsuchmethods relyingon subjectivity and the clinical experience of the examiner, particularlyatlevelsthatareloworborderthresholds.8,15
TheautomaticdeviceforCAEPresponsedetectionandthe statistical methodadopted bythedevice provedsensitive andreliablefordetectingcorticalresponses,corroborating theresultsofpreviousstudies.7,8,10
Studies of cortical thresholds have revealed a correla-tionbetweenelectrophysiologicalthresholdsandbehavioral methods of assessing hearing both in normal hearing and hearing-impairedpopulations.13,17,18
In the present study, normal hearing neonates were assessedatspecificfrequencies.Thresholdsof24.8,25,28.7 and29.4dBHLwerefoundforthefrequenciesof500,1000, 2000 and 4000Hz, respectively. Nosignificant differences were observed among the frequencies tested, rendering this procedure attractive compared to other methods of establishing electrophysiological thresholds, since other studies havereportedhigher thresholdsat lower frequen-cies of500Hz.19 Ina meta-analysis,Stapells20 reports the
thresholdsforBrainStemAuditoryEvokedPotentialin nor-mal hearing children to be: 19.5(±0.5)dB HL at 500Hz, 17.4(±0.7)dBHLat1000Hz,13.6(±0.9)dBHLat2000Hz and15.5(±0.7)dBHLat4000Hz.CAEPthresholdsobtained in this study were higher than those describe by Stapells in Brain Stem Auditory Evoked Potentials.20 Further
stud-ies are required to confirm these findings, but the rapid maturation ofthe brainstemin comparisontothe central nervoussystemmayinfluencethethresholdsobtained.One study7 reported that the corrections and standard
devi-ations between the auditory cortical thresholds and the behavioralthresholdsinadultswere17.2±7.4for500Hz, 15.5±6.0for1000Hz,16.8±7.8for2000Hzand16.0±8.4 for 4000Hz. These values can be subtracted from the thresholdsfoundfromafullyautomatedCAEPtoestimate behavioralthresholds.
AlthoughtheaverageCAEPthresholdswerebelow30dB HL,thevariationofthresholdswas0to50dBNA.Thisfact evidences and corroborates withthe scientific literature, whichshowsthatinsomecasesofnormalhearingthe cor-ticalthresholdscanbehigherthanbehavioralthresholds.10
ConeandWhitaker21evidencedthecorticalauditory
poten-tial P1in 30dB HL in 77% of infants, demonstrating that theresponsesaremoredifficulttovisualizeinthresholds, relatedtotheimmaturityofcentralareas.Someresearchers suggest that about 33% of CAEP responses maybe absent evenwhenthe stimuliareaudible14 and thatinnewborns
andyoungchildren,theimmatureresponsecanbe record-ableforstimuliwellabovethreshold.22
The present analysis of latencies for the frequencies tested yielded values of 242.79 for 500Hz; 225.54 for 1000Hz;232.74for2000Hzand244.51for4000Hzin80dB
HL.ThesefindingsmirrorthosereportedbySharma,4who
found a marked positive peak, P1, at around200---300ms aftera soundstimulusin normalhearingneonates. In the present study,no statisticallysignificant differences were observedamongthefrequencies,corroboratingtheresults of Golding,5 whodespite havingusedspeech stimuli,also
foundnodifferencesamongstimuli.However,someauthors suggestthatsoundsaredecodeddifferentlyintheauditory córtex,23,24afindingnotobservedinthepresentstudy,
pos-sibly explainedbythe factthat onlythe abilitytodetect stimuliintheauditorycortexwasshown.Someresearches describe longerlatencies for the tone burst stimuli when compared with speech sounds.21 Although speech stimuli
betterrepresentstimulusprocessingatthecentrallevel,it wasnotpossibletocomparespeechstimuliwithtoneburst in thepresent study,sincethemoduleusedin the equip-ment does not have speechsound.Regarding amplitudes, valueswere6.41,7.36,6.31and5.88Vforthefrequencies 500,1000,2000and4000Hz,respectively.Nosimilarstudies werefound in theliterature forcomparison againstthese results, renderingthem reference valuesfor theprotocol used.
Inthisstudy,therewasanincreaseinlatencyas inten-sitydecreased.Thisfactcorroboratesanotherstudy,which showed that like the other auditory evoked potentials, latencyisinfluencedbyacousticintensityduetolower neu-ronalstimulation.22
Also, the study of CAEP can objectively contribute to increasingknowledgeonthematurationprocessofthe audi-torypathway.5,6,25 Thedevelopmentof thehearingsystem
commencesduringgestationandcontinuesthroughto ado-lescence,fromperipheraltocentralstructures.24,26Recent
studiesonbrainstemAEPusingspeechsounds27haveshown
shiftingpatternofresponsesofBAEPacrossdifferentphases ofthelifespan,establishingthattheplasticityofbrainstem developmentcontinuesbeyondthe firsttwoyearsof life. FutureinvestigationofBAEPcanhelpelucidatetheprocess ofplasticityandstabilityoftheauditorypathways.
Assessment durationin the pediatricpopulation is also a key factor for successful diagnosis. This study showed shorter times than other similarclinical procedures,with an average test timeof 1h 13min, shorter than electro-physiological assessmentbyother methods.19 Theaverage
durationoftheexamwas73.3min.Otherresearches10have
described an average duration of the cortical recording sessionofaround43min,whileotherresearchers7reported
atimeof 40min in testingthe corticalthresholdof hear-ingimpaired subjects.Thisdifferencecanbeexplainedby thefactthattheresearchersuseddifferenttransducersto evoketheresponses.
The results ofthis studyhave demonstratedthe effec-tiveness of using automatic response detection when estimating cortical thresholds in an infant population. Despite the positive results of this study, further studies areneeded toexamine theconcordance of the resultsof measuringcorticalthresholdsautomatically,comparedwith behavioralassessments,bothinnormalhearingchildrenand in those withhearing loss. Some studies10 have reported
thatcortical thresholdsaremore easily visualizedin chil-drenwithhearinglossduetothepresenceofrecruitment, whichincreasestheamplitudeof responses.However,the sameresearchersaffirmthatHotelling’sT2test,partofthe
HearLabSystem, candetectnormal auditorythresholdsin children.
Although some established proceduresare widely used forassessingelectrophysiologicalthresholds,suchasBrain StemAuditoryEvokedPotentialandthemorerecent Stable-StateAuditoryEvokedPotential,nosinglemethodshouldbe analyzedalone.Whileobjectivetechniques areneeded in thisage group,monitoring hearingshouldprimarily entail behavioralaudiology assessment. In thisstudy, the corre-lationbetweencorticalandbehavioralassessmentwasnot performedbecauseofthelimitationsofbehavioral assess-mentinchildrenyoungerthansixmonths.
CAEP allows determination, both at the age group assessedandalsoinindividualsthatcannotbeassessedby the behavioral approach, of the way in which the sound stimulusreachestheauditorycortex,therebyenrichingthe objectivehearingassessment protocol. Reliable responses wereobtained in the assessment of cortical potentials in theneonatesassessedwithadeviceforautomaticresponse detection.
Conclusion
Themethoddevisedwaseffectiveforautomatedresponse analysisinestimatingauditorythresholdsofnormalhearing neonatesusingspecificfrequencystimuli,withmeanvalues of24.8dB HLfor 500Hz;25dB HLfor 1000Hz;28.7dB HL for2000Hzand29.4dBHLfor4000Hz.
Latency and amplitude values at 80dB HL intensity showed no statistically significant difference among the frequencies tested. The latency of the component P1 is inverselyproportionaltotheintensityoftheacoustic stim-ulus.
Funding
The HEARLab System was purchased with research funds (FAPESP2011/19556-3).
Conflicts
of
interest
Theauthorsdeclarenoconflictsofinterest.
Acknowledgements
TheauthorsthankProf.ErikaFukunaga,fromtheStatistical CenterofFacultyofMedicalSciencesofSantaCasaofSao Paulo,forthestatisticalanalysis.
References
1.MoonC.Theroleofearlyauditorydevelopmentinattachment andcommunication.ClinPerinatol.2011;38:657---69.
2.PaulrajMP,Subramaniam K,YaccobSB,Adom AH,HemaCR. Auditoryevokedpotentialresponseandhearingloss:areview. OpenBiomedEngJ.2015;9:17---24.
3.HallJW.Newhandbookofauditoryevokedresponses.Boston: Pearson,AllynandBacon;2007.
4.Sharma A, Dorman MF, Spahr AJ. A sensitive period for the developmentofthecentral auditorysystemin childrenwith
cochlear implants: implications for age ofimplantation. Ear Hear.2002;23:532---9.
5.GoldingM,PurdyS,SharmaM,DillonH.Theeffectof stimu-lusdurationandinterstimulusintervaloncorticalresponsesin infants.AustNZJAudiol.2006;28:122---36.
6.Rotteveel JJ,de GraafR, StegemanDF,Colon EJ, Visco YM. Thematurationofthecentralauditoryconductioninpreterm infantsuntilthreemonthsposttermV.Theauditorycortical response(ACR).HearRes.1987;27:95---110.
7.Bardy F, VanDun B, Dillon H, Seeto M,Qin H, Loi T, et al. Thecorticalautomaticthresholdestimationinadults.HearJ. 2016;69:32---7.
8.GoldingM,DillonH,SeymourJ,CarterL.Thedetectionofadult corticalauditoryevokedpotentials(CAEPs)usinganautomated statisticandvisualdetection.IntJAudiol.2009;48:833---42.
9.NirY, Vyazovskiy VV,Cirelli C,Banks MI, TononiG. Auditory responsesandstimulus-specificadaptationinratauditory cor-texarepreservedacrossNREMand REMsleep.CerebCortex. 2015;25:1362---78.
10.CarterL,GoldingM,DillonH,SeymourJ.Thedetectionofinfant corticalauditoryevokedpotentials(CAEPs)usingstatisticaland visualdetectiontechniques.JAmAcadAudiol.2010;21:347---56.
11.HydeM.Theslowvertexpotential:propertiesandclinical appli-cation.In:JacobsonJT,editor.Principlesand applicationsin auditoryevokedpotentials.NeedhamHeights:AllynandBacon; 1994.p.179---218.
12.Cardon G, Campbell J, Sharma A. Plasticity in the develop-ingauditorycortex:evidencefromchildrenwithsensorineural hearinglossandauditoryneuropathyspectrumdisorder.JAm AcadAudiol.2012;23:396---411.
13.ChangHW,DillonH,CarterL,vanDunB,YoungST.The rela-tionship between cortical auditory evoked potential (CAEP) detectionandestimatedaudibilityininfantswithsensorineural hearingloss.IntJAudiol.2012;51:663---70.
14.Van Dun B, Carter L, Dillon H. Sensitivity of cortical audi-toryevokedpotentialdetectionforhearing-impairedinfantsin responsetoshortspeechsounds.AudiolRes.2012;2:e13.
15.Purdy SC, Sharma M, Munro KJ, Morgan CL. Stimulus level effects onspeech-evoked obligatorycorticalauditoryevoked potentialsininfants withnormalhearing. ClinNeurophysiol. 2013;124:474---80.
16.American Academy ofPediatrics, JointCommittee on Infant Hearing.Year2007position statement: Principlesand guide-linesfor early hearingdetection and interventionprograms. Pediatrics.2007;120:898---921.
17.VanDunB,DillonH,SeetoM.Estimatinghearingthresholdsin hearing-impairedadultsthroughobjectivedetectionofcortical auditoryevokedpotentials.JAmAcadAudiol.2015;26:370---83.
18.Durante AS, Wieselberg MB, Carvalho S, Costa N, Pucci B, Gudayol N,et al. Cortical auditoryevoked potential: evalu-ation ofspeechdetection in adulthearingaid users.Codas. 2014;26:367---73.
19.RibeiroFM, CarvalhoRM,MarcouxAM. Auditorysteady-state evokedresponsesforpretermandtermneonates.Audiol Neu-rotol.2010;15:97---110.
20.StapellsDR.Thresholdestimationbythetone-evokedauditory brainstemresponse:aliteraturemeta-analysis.JSpeechLang PathAudiol.2004;24:74---83.
21.ConeW,WhitakerR.Dynamicsofinfantcorticalauditoryevoked potentials(CAEPs)for toneand speech tokens.IntJPediatr Otorhinolaryngol.2013;77:1162---73.
22.BritishSocietyofAudology.Recommendedprocedure---cortical auditoryevokedpotential(CAEP)testing;2016.https://www. thebsa.org.uk/wp-content/uploads/2016/05/Cortical-ERA.pdf
23.AgungK, Purdy SC, McMahon CM,Newall P. Theuse of cor-ticalauditoryevokedpotentialsto evaluateneural encoding of speech sounds in adults. J Am Acad Audiol. 2006;17: 559---72.
24.LahavA,SkoeE.AnacousticgapbetweentheNICUandwomb: apotentialriskforcompromisedneuroplasticityoftheauditory systeminpreterminfants.FrontNeurosci.2014;8:381.
25.Cardon G, Sharma A. Central auditory maturation and behavioral outcome in children with auditory neuropathy spectrum disorder who use cochlear implants. Int J Audiol. 2013;52:577---86.
26.EggermontJJ,MooreJK.Morphologicalandfunctional devel-opmentof theauditory nervous system. In: Werner LA, Fay RR,PopperNA,editors.Humanauditorydevelopment.London: SpringerNewYorkDordrechtHeidelberg;2012.p.61---105.
27.SkoelE,KrizmanJ,AndersonS,KrausN.Stabilityand plastic-ityofauditorybrainstemfunction acrossthelifespan.Cereb Cortex.2015;25:1415---26.
