贺勇fMRI研究设计和论文写作.ppt

贺勇fMRI研究设计和论文写作.ppt

《贺勇fMRI研究设计和论文写作.ppt》由会员分享，可在线阅读，更多相关《贺勇fMRI研究设计和论文写作.ppt（35页珍藏版）》请在冰豆网上搜索。

研究设计与论文写作贺永北京师范大学认知神经科学与学习国家重点实验室一个“好”的研究原创性/创新性Originality/Creativity具有重要性Significance具备可行性Feasibility可重复性Reproducibility/Reliability发表后引起关注（正面或者负面）Influence/impact:

5-10years作者、编辑、审稿人、读者科学研究的步骤-作者选题Selectingandformulatingresearchproblem检索文献Surveyingliterature建立假说Developinghypotheses设计实验Designingexperiment采集数据Collectingdata处理和分析数据Processingandanalyzingdata检验假说Testinghypotheses解释结果Interpretingresultsandgeneralizationofresults结论Makingconclusion写作Paperwriting审稿及发表（编辑、审稿人、读者）投稿修改出版“ThisstudywasthefirsttoderiveastructuralnetworkofthehumanbrainonthebasisofcorrelationsincorticalgreymatterthicknessmeasuredusingMRI.”（BullmoreandSporns,NatRevNeurosci2009）Citation:

137（fromISI）Reviewer1:

“Thisisaninterestingwell-writtenandcreativelyconceivedpapershowingforthefirsttimethathumanbrainanatomicalconnectivitymaybetopologicallyorganizedaccordingtosmallworldprinciples.”Reviewer2:

“Thispaperdealswithanimportantandverysignificantresearcharea,thecharacterizationoftheanatomicalnetworkoflong-rangeconnectionsinthehumanbrain.Thiscrucialdatasetiscurrentlylackingandanyprogresstowardsamorecompleteanatomicaldescriptionofthehumancortexishighlydesirable.Thepaperattemptstoaddressthisissuebyusinganovelandpotentiallyinterestingapproachtowardsmappinganatomicalpathwaysandprovidesananalysisoftheresultingnetwork.”StructuralMRI-basedBrainConnectomeAnalysisMethods（Chenetal.,CerebCortex2009;Sanabria-Diazetal.,NeuroImage2010）Normaldevelopment（Fanetal.,NeuroImage2011）Normalaging（Wuetal.,HumBrainMapp2011;Zhuetal.,NeurobiologyofAging2011;Chenetal.,NeuroImage2011）Alzheimersdisease（Heetal.,JNeurosci2008;Yaoetal.,PLoSComputationalBiology2010）Multiplesclerosis（Heetal.,Brain2009）Epilepsy（Rajetal.,NeuroImage2010;Bernhardtetal.,CerebralCortex2011）Schizophrenia（Bassettetal.,JNeurosci2008）Genetic（Schmittetal.,CerebralCortex2009）Howtoformtheidea?

2003年IPMI2003:

reject2003年2subjectsfromDr.XuchuWengIPMI2003:

rejectSmall-worldhumanbrainfunctionalnetworksbyresting-statefMRISalvadoretal（2005）CerebCortex2005.7.20MeetingwithAlanEvansandKeithWorsley说服你的老板！

MontrealLerchetal（2006）NeuroImage2005.09Question:

canweusestructuralMRItoconstructbrainnetworks?

Hypothesis:

humanbrainstructuralnetworkshaveasmall-worldorganization.Workhard!

Remember:

youonlyhaveonechance!

RegionsRegionsNetworkconstructionCorticalthicknessmeasurements0.60.40.20-0.2-0.4RegionsCorticalparcellation（ANIMAL+INSECT）Cross-correlationmatrixFDRHumancorticalthicknessnetworkBinarizedmatrix12407Subjects54CorticalthicknessdatamatrixHeetal（2007）CerebCortexMethodsParticipants/experimentaldesign/MRI（ICBM152）ImagingprocessingStatisticalanalysis要求：

合理性/准确性/新颖性/完整性方法描述必须准确，逻辑清晰关键的方法简要描述其合理性尽量避免任意选择参数，否则要评价参数变化对结果的稳定性充分利用你的data引述公式和采用别人的方法必须给出出处整个部分具有完整性Methods要求：

方法描述必须准确，逻辑清晰ThenativeMRimageswerefirstregisteredintostereotaxicspace（TalairachandTournoux1988）usinga9-parameterlineartransformation（Collinsetal.1994）.Simultaneously,imageswerecorrectedfornonuniformityartifactsusingtheN3algorithms（Sledetal.1998）.Theregisteredandcorrectedimageswerefurthersegmentedintograymatter,whitematter,cerebrospinalfluid,andbackgroundusinganadvancedneuralnetclassifier（Zijdenbosetal.2002）.TheinnerandoutergraymattersurfaceswerethenautomaticallyextractedfromeachMRvolumeusingtheConstrainedLaplacian-basedAutomatedSegmentationwithProximities（CLASP）algorithm（MacDonaldetal.2000;Kimetal.2005）andcorticalthicknesswasmeasuredinnativespaceusingthelinkeddistancebetween2surfacesat40962verticesthroughoutthecortex.Methods要求：

方法描述必须准确，逻辑清晰Finally,temporalfiltering（0.01Hzf0.08Hz）wasperformedonthetimeseriesofeachvoxeltoreduce（notavoid!

）theeffectoflow-frequencydriftsandhighfrequencynoise（Biswaletal.,1995;Loweetal.,1998;Jiangetal.,2004）.Methods要求：

方法描述必须准确，逻辑清晰关键的方法简要描述其合理性TheCLASPcorticalthicknessalgorithmhasbeenvalidatedusingbothmanualmeasurements（Kabanietal.2001）andsimulationapproaches（LerchandEvans2005;Leeetal.2006）,andrecentlyappliedtoAlzheimersdisease（Lerchetal.2005）andcorticaldevelopmentstudies（Shawetal.2006）.Methods要求：

方法描述必须准确，逻辑清晰关键的方法简要描述其合理性尽量避免任意选择参数，否则要评价参数变化对结果的稳定性Toexploretheinfluenceofthresholdingonthesmall-worldattributes,inthisstudy,weappliedthemaximumrangeofcorrelationthresholds（0RRmax,withanincrementalintervalof0.01）tothecorrelationmatrix,andrepeatedthesmall-worldanalysisforeveryRvalue.Methods要求：

方法描述必须准确，逻辑清晰关键的方法简要描述其合理性尽量避免任意选择参数，否则要评价参数变化对结果的稳定性充分利用你的data引述公式和采用别人的方法必须给出出处整个部分具有完整性Results通常分标题进行要求：

逻辑清晰，主次分明（避免简单重复）注意所投期刊的风格：

是否需要简单讨论合理使用SupplementalMaterialsCorpuscallosumShort-distancefibersSuperiorlongitudinalfasciculusThemostsignificant15corticalthicknesscorrelationsAABC（CorrectedPCpbrainCprandomLpr