000865865 001__ 865865
000865865 005__ 20210130003204.0
000865865 0247_ $$2doi$$a10.1016/j.neuroimage.2019.116276
000865865 0247_ $$2ISSN$$a1053-8119
000865865 0247_ $$2ISSN$$a1095-9572
000865865 0247_ $$2Handle$$a2128/23997
000865865 0247_ $$2altmetric$$aaltmetric:68252714
000865865 0247_ $$2pmid$$apmid:31610298
000865865 0247_ $$2WOS$$aWOS:000507987000012
000865865 037__ $$aFZJ-2019-05152
000865865 082__ $$a610
000865865 1001_ $$0P:(DE-HGF)0$$aHe, Tong$$b0
000865865 245__ $$aDeep neural networks and kernel regression achieve comparable accuracies for functional connectivity prediction of behavior and demographics
000865865 260__ $$aOrlando, Fla.$$bAcademic Press$$c2020
000865865 3367_ $$2DRIVER$$aarticle
000865865 3367_ $$2DataCite$$aOutput Types/Journal article
000865865 3367_ $$0PUB:(DE-HGF)16$$2PUB:(DE-HGF)$$aJournal Article$$bjournal$$mjournal$$s1579705295_19654
000865865 3367_ $$2BibTeX$$aARTICLE
000865865 3367_ $$2ORCID$$aJOURNAL_ARTICLE
000865865 3367_ $$00$$2EndNote$$aJournal Article
000865865 520__ $$aThere is significant interest in the development and application of deep neural networks (DNNs) to neuroimaging data. A growing literature suggests that DNNs outperform their classical counterparts in a variety of neuroimaging applications, yet there are few direct comparisons of relative utility. Here, we compared the performance of three DNN architectures and a classical machine learning algorithm (kernel regression) in predicting individual phenotypes from whole-brain resting-state functional connectivity (RSFC) patterns. One of the DNNs was a generic fully-connected feedforward neural network, while the other two DNNs were recently published approaches specifically designed to exploit the structure of connectome data. By using a combined sample of almost 10,000 participants from the Human Connectome Project (HCP) and UK Biobank, we showed that the three DNNs and kernel regression achieved similar performance across a wide range of behavioral and demographic measures. Furthermore, the generic feedforward neural network exhibited similar performance to the two state-of-the-art connectome-specific DNNs. When predicting fluid intelligence in the UK Biobank, performance of all algorithms dramatically improved when sample size increased from 100 to 1000 subjects. Improvement was smaller, but still significant, when sample size increased from 1000 to 5000 subjects. Importantly, kernel regression was competitive across all sample sizes. Overall, our study suggests that kernel regression is as effective as DNNs for RSFC-based behavioral prediction, while incurring significantly lower computational costs. Therefore, kernel regression might serve as a useful baseline algorithm for future studies.
000865865 536__ $$0G:(DE-HGF)POF3-571$$a571 - Connectivity and Activity (POF3-571)$$cPOF3-571$$fPOF III$$x0
000865865 588__ $$aDataset connected to CrossRef
000865865 7001_ $$0P:(DE-HGF)0$$aKong, Ru$$b1
000865865 7001_ $$0P:(DE-HGF)0$$aHolmes, Avram J.$$b2
000865865 7001_ $$0P:(DE-HGF)0$$aNguyen, Minh$$b3
000865865 7001_ $$0P:(DE-HGF)0$$aSabuncu, Mert R.$$b4
000865865 7001_ $$0P:(DE-Juel1)131678$$aEickhoff, Simon B.$$b5
000865865 7001_ $$0P:(DE-Juel1)136848$$aBzdok, Danilo$$b6
000865865 7001_ $$0P:(DE-HGF)0$$aFeng, Jiashi$$b7
000865865 7001_ $$0P:(DE-HGF)0$$aThomas Yeo, B. T.$$b8$$eCorresponding author
000865865 773__ $$0PERI:(DE-600)1471418-8$$a10.1016/j.neuroimage.2019.116276$$gp. 116276 -$$p116276$$tNeuroImage$$v206$$x1053-8119$$y2020
000865865 8564_ $$uhttps://juser.fz-juelich.de/record/865865/files/1-s2.0-S1053811919308675-main.pdf$$yOpenAccess
000865865 8564_ $$uhttps://juser.fz-juelich.de/record/865865/files/1-s2.0-S1053811919308675-main.pdf?subformat=pdfa$$xpdfa$$yOpenAccess
000865865 909CO $$ooai:juser.fz-juelich.de:865865$$pdnbdelivery$$pdriver$$pVDB$$popen_access$$popenaire
000865865 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)131678$$aForschungszentrum Jülich$$b5$$kFZJ
000865865 9131_ $$0G:(DE-HGF)POF3-571$$1G:(DE-HGF)POF3-570$$2G:(DE-HGF)POF3-500$$3G:(DE-HGF)POF3$$4G:(DE-HGF)POF$$aDE-HGF$$bKey Technologies$$lDecoding the Human Brain$$vConnectivity and Activity$$x0
000865865 9141_ $$y2020
000865865 915__ $$0StatID:(DE-HGF)0200$$2StatID$$aDBCoverage$$bSCOPUS
000865865 915__ $$0StatID:(DE-HGF)1030$$2StatID$$aDBCoverage$$bCurrent Contents - Life Sciences
000865865 915__ $$0StatID:(DE-HGF)0600$$2StatID$$aDBCoverage$$bEbsco Academic Search
000865865 915__ $$0LIC:(DE-HGF)CCBYNCND4$$2HGFVOC$$aCreative Commons Attribution-NonCommercial-NoDerivs CC BY-NC-ND 4.0
000865865 915__ $$0StatID:(DE-HGF)0100$$2StatID$$aJCR$$bNEUROIMAGE : 2017
000865865 915__ $$0StatID:(DE-HGF)0150$$2StatID$$aDBCoverage$$bWeb of Science Core Collection
000865865 915__ $$0StatID:(DE-HGF)0110$$2StatID$$aWoS$$bScience Citation Index
000865865 915__ $$0StatID:(DE-HGF)0111$$2StatID$$aWoS$$bScience Citation Index Expanded
000865865 915__ $$0StatID:(DE-HGF)0510$$2StatID$$aOpenAccess
000865865 915__ $$0StatID:(DE-HGF)0030$$2StatID$$aPeer Review$$bASC
000865865 915__ $$0StatID:(DE-HGF)9905$$2StatID$$aIF >= 5$$bNEUROIMAGE : 2017
000865865 915__ $$0StatID:(DE-HGF)0310$$2StatID$$aDBCoverage$$bNCBI Molecular Biology Database
000865865 915__ $$0StatID:(DE-HGF)1050$$2StatID$$aDBCoverage$$bBIOSIS Previews
000865865 915__ $$0StatID:(DE-HGF)0300$$2StatID$$aDBCoverage$$bMedline
000865865 915__ $$0StatID:(DE-HGF)0420$$2StatID$$aNationallizenz
000865865 915__ $$0StatID:(DE-HGF)0199$$2StatID$$aDBCoverage$$bClarivate Analytics Master Journal List
000865865 920__ $$lyes
000865865 9201_ $$0I:(DE-Juel1)INM-7-20090406$$kINM-7$$lGehirn & Verhalten$$x0
000865865 980__ $$ajournal
000865865 980__ $$aVDB
000865865 980__ $$aUNRESTRICTED
000865865 980__ $$aI:(DE-Juel1)INM-7-20090406
000865865 9801_ $$aFullTexts