多因素分析法用于熒光光譜識別胃癌的研究.pdf

1、分類號南絨研究生學位論文多因素分析法用于熒光光譜識別胃癌的研究研究生姓名史睦凰指導教師姓名呈墨(數(shù)授)申請學位級別亟專業(yè)名稱迸堂王猩論文答辯日期!QQ!生旦3旦學位授予日期2QQ生!旦中國海洋大學多因素判別法用于熒光光譜識別胃癌的研究StudyOilMultivariateStatisticalMethodsUsedforFluorescenceSpectrumforStomachCancerDiagnosisAbstractMorea

2、ndmoreattentionhasbeenpaidtofluorescencespectroscopyforitsadvantagesofrealtimedetection，impersonalityinnocuityandSOonAlthoughthecanceridentificationusingspectroscopyhasbeenstudiedextensivelyitisstillatprimarystageEspecia

3、llytheslowspectrumanalysistechnologyhaskeptbackthedevelopmentoffluorescencespectrumforstomachcancerdiagnosisLightinducedfluorescencewasusedinthethesis308rim，337nm，365nmand405nmwerechosenasexcitationwavelengthThe58normals

4、amplesand58cancersamplesweretakenfrom58stomachcancerpatientsThe116samplesinvitrohavebeenmeasuredThestudyonthefluorescencespectrarevealedthat308nmand337nmmightbetheoptimalexcitationwavelengthsandtheywereselectedforthedisc

5、riminatinganalysisThedoublefactorsdiscrimination，thepartialLeast—squares(PLS)methodandtheartificialneuralnetworkbasedonthepartialLeastsquares(ANNPLS)wereusedtoanalyzefluorescencespectraofstomachtissuesThesenskivityspecif

6、icityandthepositivepredictivevalueofthethreeclassificationalgorithmswere684％，947％，929％：81，O％，966％，959％；862％，100％，100％inseparatelySoweconcludedthattheANNPLSmethodwasthemosteffectivemethodforidetatificationofstomachcancelK