Generative Embedding for Model-Based Classification of fMRI Data 英文参考文献.docVIP

GenerativeEmbeddingforModel-BasedClassificationof fMRIData KayH.Brodersen1,2*,ThomasM.Schofield3,AlexanderP.Leff3,ChengSoonOng1,EkaterinaI. Lomakina1,2,JoachimM.Buhmann1,KlaasE.Stephan2,3 1DepartmentofComputerScience,ETHZurich,Zurich,Switzerland,2LaboratoryforSocialandNeuralSystemsResearch,DepartmentofEconomics,UniversityofZurich, Zurich,Switzerland,3WellcomeTrustCentreforNeuroimaging,UniversityCollegeLondon,London,UnitedKingdom Abstract Decoding models, such as those underlying multivariate classification algorithms, have been increasingly used to infer cognitiveorclinicalbrainstatesfrommeasuresofbrainactivityobtainedbyfunctionalmagneticresonanceimaging(fMRI). The practicality of current classifiers, however, is restricted by two major challenges. First, due to the high data dimensionalityandlowsamplesize,algorithmsstruggletoseparateinformativefromuninformativefeatures,resultingin poorgeneralizationperformance.Second,populardiscriminativemethodssuchassupportvectormachines(SVMs)rarely afford mechanistic interpretability. In this paper, we address these issues by proposing a novel generative-embedding approachthatincorporatesneurobiologicallyinterpretablegenerativemodelsintodiscriminativeclassifiers.Ourapproach extendspreviousworkontrial-by-trialclassificationforelectrophysiologicalrecordingstosubject-by-subjectclassification forfMRIandofferstwokeyadvantagesoverconventionalmethods:itmayprovidemoreaccuratepredictionsbyexploiting discriminativeinformationencodedin‘hidden’physiologicalquantitiessuchassynapticconnectionstrengths;anditaffords mechanistic interpretability of clinical classifications. Here, we introduce generative embedding for fMRI using a combination of dynamic causal models (DCMs) and SVMs. We propose a general procedure of DCM-based generative embedding for subject-wise classification, provide a concrete implementation, and suggest good-practice guidelines for unbiasedapplicationofgenerativeembeddinginthecontextoffMRI.Weillustratetheutilityofourappro