40硝酸用于选择性单硝化和苄位氧化王盈

。1。,,。1　3　3.1　(1),3。3.2　,ANN。,,,。,,,。,0.15,0.1。:[1]　BalanG,TimminsP,Greene,D,etal.Invivo-invitrocorrelation(IVIVC)modelsformetforminaf-teradministrationofmodified-release(MR)oraldossageformstohealthyhumanvolunteers[J].JPharmSci,2001,90(8):1176-1185.[2]　HuSR.TheapplicationtechnologyofANN[M].Changsha:NationalDefenceScienceandTechnologyUniversityPress,1993.73-437.[3]　ErbRJ.Thebackpropagationneuralnetwork—aBayesianclassifier.Introductionandapplicabilitytopharmacokinetics[J].ClinPharmacokinet,1995,29(2):69-79.[4]　,,　.[J].,1997,16(1):8-10.[5]　,,,.[J].,1981,13(4):45-50.Invivo-invitroCorrelationforErythromycinSustainedReleaseMicrocapsulesUsingArtificialNeuralNetworkLILing-Bing1,　LIXiu2(1.Dept.ofPharmaceutics,ShandongUniversity;2.AffiliatedHospitalofShandongUniversity,Ji'nan250012)ABSTRACT:Sustainedreleasegelatinmicrocapsulesoferythromycinwerepreparedandinvivo-invitrocorrelationwasinvestigatedusingartificialneuralnetwork(ANN).InvitrodataandinvivodatawereusedastheANNmodelinputandoutput,respectively.Internalpredictabilitywasassessedbycom-paringobservedandpredictedvaluesforinvivodata.TheresultsshowedthatANNhadsatisfactoryinvi-vo-invitrocorrelation.KeyWords:erythromycinsustainedreleasemicrocapsules;artificialneuralnetwork;invivo-invitrocorrelationS34-67　Knoevenagel　RenZJ[SynthCommun,2002,32:3475],Knoevenagel。168%,1180%～96%。,。[]S34-68　40%　Bahu-layanD[SynthCommun,2002,32:3565]/40%5h,,635%～99%。,541%～90.5%。[　]·612·　ChineseJournalofPharmaceuticals2003,34(12)