赵仑_09_March_ERP高级培训班讲座(4)

ExplicitcategorizationtaskImplicitcategorizationtask行为数据测量反应时、正误率反应时的重要性（举例）平均波幅测量固定法/连续测量具体法或间断法ERN，error-relatednegativity－ConflictcontrolSubsequentvisualscrutinyensuredthatthemostnegativevaluesrepresentedrealpeaksratherthanendpointsoftheepoch.（1）峰潜伏期、始潜时、波幅（2）头皮分布????（3）定位分析RepeatedmeasuresANOVA－ERPs研究主要比较不同条件下的成份差异。统计和结果分析PrioriorplannedcomparisonPosteriori,post-hoc,unplannedcomparisonForfactorswith????levels,P-valueswerecorrectedfornon-sphericityusingtheGreenhouse-GeissercorrectionBonferroni脑区和电极的选择－N170－Word/Face/ObjectERP时频分析Basar等（1984，1985）time-frequencyanalysisBandAnalysisEvoked－time-phaselockedInduced－time-,non-phaselockedERD/ERSEvent-relateddesynchronizationEvent-relatedsynchronizationCoherenceAnalysis…..Neurophysiologymustdevelopnewwaystomeasuretheactivityofnetworks.Thiswillprobablyinvolvecorrelationalstudiesandattentiontorhythmsaswellastoisolatedpotentials.-----Picton,Alain,&McIntoshVEP-2DfrequencyVEP-2DVEP-3D脑电地形图源分析偶极子（Dipole）是一对数值相等、符号相反的电荷，彼此相隔一定距离时形成的体系。ERPs成份的源即脑内的偶极子。神经传导递质作用于突触后神经元受体的过程中产生的突触后电位导致了ERPs的产生(Luck,1998)。头皮电位的分布取决于电流偶极子的位置与方向。源定位可根据偶极子的数目、时间特征、位置和偶极子量等因素推导。偶极子模型－single/multipledipolemodels偶极子定位电流密度（currentdensity）可看作一定容积内的偶极子量每个偶极子代表一定容积的电流密度,偶极子量与所代表的电流密度是成比例的。电流密度•ERP溯源分析的相关问题怎样确定用于溯源分析的时间范围？如何有效地使用PCA/ICA？什么情况下需要将某些ICA成分滤除？对自己的数据，应该选择哪一种源重构模型最适当？如何判断得到的溯源结果是否真实？Auditory-VisualSpeechIntegrationMcGurkEffectbySourceReconstructionAuditory/ba/ReconstructionTimePoints(foreachidentifiedpeakinthemGFP)Visual/ba/AuditoryVisual/ba/70ms165ms150ms204ms90ms212msAuditory/ba/CurrentDensityReconstruction(MinimumNormLeastSquares)Visual/ba/AuditoryVisual/ba/Latency-70ms(auditoryPI)Auditory/ba/CurrentDensityReconstruction(MinimumNormLeastSquares)Visual/ba/AuditoryVisual/ba/Latency-150ms(visualPI)WhyisSNRImportant?WhatisSNR?•Signal-noise-ratio(manyoptionsforcalculation)•WheredoIfindit?•Transforms•Whyisitimportant?•Tendencytooverestimatepeakamplitude•e.g.,IfSNRis2:1,asmuchashalfmeasureamplitudemayreflectnoise.•Problemenhancebyusingextremesaspeaks.•example:ms-100.0-50.00.050.0100.0150.0200.0250.0300.0350.03µVElectrodeCz4.2µVp.p.5.3µVp.p.25%increaseinmeasureamplitudeDUETONOISECONCLUSIONSConfidenceEllipsoids•Descriptionoferror/variabilityrelatedtoSNRofdata•eliminatespointsourcedescriptionofdipolelocation•PrecisesourcelocalizationwithhighdegreeofconfidencerequireshighSNR•requiressignificantamountsofindividualorgroupdata•SNRcanbeimprovedbyselectionoffilteringtechniquestooptimizeSNRforcomponentofinterest•bandpass,wavelet,ICA•Methodishelpfulindeterminingnumberofsourcestoexplaindata•superfluoussourceshavelargeconfidencevolumes•Confidenceellipsoidsdonotincorporateerrorsbasedon•mislocationofsensors/electrodesonscalp•digitizepositions•inappropriate/inaccuratesourcemodel•useboundary/finiteelementmodelsERPs研究成分研究应用研究CNVP300（P3a）N400？MMN特点及其影响因素起源及产生机制心理生理临床应用特因条件功能评估注意、记忆、语言加工、运动知觉等神经精神科、昏迷愈后、辅助诊断等航空、航天、航海、恶劣环境条件等驾驶疲劳、功能评估、音乐认知能力、健康评估等ERPs的研究分类EyemovementCNV标准CNV范式，即警告刺激（S1）－命令刺激（S2）－运动反应（MR，Motorresponse）。S1与S2之间脑电发生负向偏转—关联性负变(CNV,contingentnegativevariation)。（Walt&Cooper,1964,Nature）S1:声/光/其他S2:声/光/其他EEGCNV（魏景汉；JinghanWei,etal）Emotion&ERP*解脱波情绪研究的重要指标RecognitionMemoryOld-neweffectRetrieval(郭春彦，等.2003)DMeffect(Paller,etal,1987)DifferencesbasedonlatermemoryperformanceEncodingMMN（mismatchnegativity)Traditionalparadigm•新范式及其意义Control：Deviant，Standard，Voice1-3Oddball：Deviant，Standard(Jacobsen&Schröger,2001)Oddball-MMN=Sensoryrefractory+SensorymemoryControl-MMN=SensorymemoryOddball-MMNControl-MMNThenatureofauditoryMMNPolarityreverse---noseasthereferenceelectrode.Thefirsttrialasdeviants---noMMN(e.g.,Cowan,etal,1993)DelayedMMNwithlongertimewindowasthedeviantdecreased(Näätänen,etal.,1989).DecreasedMMNastheSOAprolonged(e.g.,Näätänen,etal.,1993;Rüsseler,etal.,2001).Pre-attentivesensory-memoryaswellaslong-termmemory,e.g.,semanticMMN,syntaxMMN,musiclanguageMMN(Näätänen,etal,2007,forareview).………