嗓音的客观多参数分析与主观听觉分析相关性的研究

《嗓音的客观多参数分析与主观听觉分析相关性的研究》由会员上传分享，免费在线阅读，更多相关内容在行业资料-天天文库。

1、听力学及言语疾病杂志2003年第11卷第3期161]1a1床研究·嗓音的客观多参数分析与主观听觉分析相关性的研究于萍，GiovanniAntoine【摘要」目的探讨嗓音的客观多参数检测与主观听觉评估的相关性，建立嗓音客观多参数评估模式，实现嗓音评估的客观化、数据化。方法声音样本采自135例声音嘶哑患者和29例嗓音正常者。客观检测采用法国开发的，"EVA'”嗓音工作站，测试对象发长元音/a:/时，检测下列参数:基频(En)、音强((intensity)、基频微扰(jitter)、信噪比(signal一l。一、,sr,ratin,SNR),场apunov系数、口腔气流量(oralan

2、flw,OAF),音域(range)及最大发声时间(maxnmumphonaturytime,MPP);发/1./音时，检测声门下It(estimatedsubglotticpressure,ESGP)。主观听觉评估采用日本言语语音学会声音嘶W评估GRRAS系统中的总嘶哑度(grade,G)4级评估标准。所有受试者以自然的音调及音强读一段统一的标准文字，听评委山4名嗓音学家组成。结果应用判别分析方法，建立了7参数客观评估模式(音域、Lya-punoti系数、声门下压、最大发声时间、口腔气流量、信噪比和基频)。客观评估结果与主观评估结果的一致性达到94%。结论嗓音的客观检测是多参数

3、的，客观评估能够反映出主观听觉印象。【关键词】嗓音客观多参数分析主观听觉分析判别分析【中图分类号】R767.92【文献标识码】A【文章编号】1006一7299(2003)03一。161一04CorrelationBetweenObjectiveMufti一parametricAnalysisandPerceptualAnalysisforDysphoniaYuPing',GiovanniAntoine(‘InstituteofOtorhinolaryngology,ChinesePLAGeneralHospital,Beijing,100853)[Abstract]ObjeCtl

4、VeTodevelopamulti一parametricobjectiveevaluationprotanlandinvestigateitscorrelationwithperceptualanalysisofcontinuousspeechforobjectivityandquantitatively.MethodsThevoicesampleswerecollectedfrom135patientswithdyaphouiaand29subjectswithnonnalvoice.Theobjectiveparameters,includingfundamentalfreq

5、uency(E,),in-tensity,jitter,signal一to一noiseratio,Lyapmuvcoefficient(LG),oralairflow(OAF),vocalrange(Range)andnaximunpho-natorytime(MPT)which~measuredmainlyonasustained/a/andestimatedsuhglotticpressure(ESGP)whichwasdeter-minedona。。of/pa/,wererecordedonan"EVA"workstationwhich。developedinFrench.

6、AccordingtotheG(ove-calldysphonia)componentoftheGRBASscoreona4一pointscalerangingfrom0fornormalto3forseveredysphonia,theperceptualanalysi，二donebyajurycomposedof4experiencedlistenersandthevoicesamples(continuousspeech)werefromthe--patientswithnormalvoicesandnormalintensity.ResultsUsingthediscri

7、minateanalysis,asevenparametersprotocol(Range.LC,F-S(;P,MPr,OAF,signal一to一noiseratioando)wasdeveloped.Resultshowedthatthisprotocolallowed84%cuncor-dancewithjuryclassificati-ConclusionTheobjectivevoiceevaluationismulti-parametricandisabletoref