基于磁共振的大腦網絡分析方法研究及其應用.pdf

1、電子科技大學碩士學位論文基于磁共振的大腦網絡分析方法研究及其應用姓名：潘正勇申請學位級別：碩士專業(yè)：應用數(shù)學指導教師：陳華富20100501ABSTRACTABSTRACTBeingthecommandcenterofahumanbodybrainhasacomplexorganizationinboth咖ctl鵬andfunctionalmechanismGenerallyfunctionsamongseveralbrainregio

2、nsmaybesimilarwhichmeanstheseregionsformanetwork，usuallycalledstaticnetworkOntheotherhand，interactionOccursbetweensomebrainregionswhentaskperformance，suchasmodulationandinhibition，formingadynamicnetworkWi也theprogressofth

3、etimes，thehumanbeingnotonlyexploreshislivingenvironment，butalsomakesagoodstudyofhisownbodyespeciallythebrainInthiscase，brainmapping，especiallytheinventionofmagneticresonanceimaging(MRl)technologyisavaluabletoolforhumanto

4、studytheirbrainsRecentlythebrainnetwork，basedontheresearchoftheMRI，hasbeenwidelystudied111estaticanddynamicnetworks，wementionedabove，舡ethemosttwoimportantandwellknownnetworksinthisfieldTheseresearchesnotonlyalehelpfultos

5、tudyourselves，moreimportantlymeyplayaimportantroleindiseasestudyInthispaperbasedonthemagneticresonanceimagingdatasetweconsiderthemethodsinbrainnetworkanalysisandtheirapplicationsThenumericalresultsdemonstratetheeffective

6、nessofourmethodsThreeaspectsofthisdissertationareasfollows：1ProposeaBayesianstudybasedradialbasisfunction(RBF)neuralnetworkforlocatingthefunctionnetworkswhentaskperformancesS吖eralbrainnetworks，suchasvisual，imaginaryandha

7、ndmove，arewellseparatedwhichindicatestheeffectivenessofthismethodinthelocalizationoffunctionnetworksItalsodemonstratesthatthemethodcandiscriminatetheresponsesofdifferenttasksThatmeans，thisresearchprovidesUSanewtoolforfun

8、ctionlocalizationwhenmultipletasksperformances2Combiningthefunctionalandthestructuralconnectivitytostudythedefaultmodenetworkofpatientswithmesialtemporallobeepilepsywefindthefunctionalconnectivityandthestructuralconnecti

9、vitybetweenPCCandbilateralhippocampusarebothdecreasedMoreoverboththefunctionalconnectivityandthestructuralconnectivityarCsignificantlypositivelycorrelatedwhichm瑚lilSthatcombiningthefunctionalconnectivityandthestructuralc