Temporal embedding and spatiotemporal feature selection boost multi-voxel pattern analysis decoding accuracy

doi:10.1016/j.jneumeth.2020.108836

	Temporal embedding and spatiotemporal feature selection boost multi-voxel pattern analysis decoding accuracy
	Choupan, Jeiran 2,3,4,5; Douglas, Pamela K.1,7,8; Gal, Yaniv 6; Cohen, Mark S.9,10,11,12,13,14,15,16; Reutens, David C.2; Yang, Zhengyi 2,6,17
刊名	JOURNAL OF NEUROSCIENCE METHODS
	2020-11-01
卷号	345 页码:13
关键词	fMRI Multi-variate pattern analysis Spatiotemporal feature selection Multiband EPI Random forest Support vector machine
ISSN号	0165-0270
DOI	10.1016/j.jneumeth.2020.108836
通讯作者	Choupan, Jeiran(choupan@usc.edu)
英文摘要	Background: In fMRI decoding, temporal embedding of spatial features of the brain allows the incorporation of brain activity dynamics into the multivariate pattern classification process, and provides enriched information about stimulus-specific response patterns and potentially improved prediction accuracy. New method: This study investigates the possibility of enhancing the classification performance by exploring temporal embedding, to identify the optimum combination of spatiotemporal features based on their classification performance. We investigated the importance of spatiotemporal feature selection using a slow event-related design adapted from the classic Haxby study (Haxby et al., 2001). Data were collected using a multiband fMRI sequence with temporal resolution of 0.568 s. Comparison with existing methods: A wide range of spatiotemporal observations were created as various combinations of spatiotemporal features. Using both random forest, and support vector machine, classifiers prediction accuracies for these combinations were then compared with the single spatial multivariate pattern approach that uses only a single temporal observation. Results: Our findings showed that, on average, spatiotemporal feature selection improved prediction accuracy. Moreover, the random forest algorithm outperformed the support vector machine and benefitted from temporal information to a greater extent. Conclusions: As expected, the most influential temporal durations were found to be around the peak of the hemodynamic response function, a few seconds after the stimuli onset until - 4 s after the peak of the hemodynamic response function. The superiority of spatiotemporal feature selection over single time-point spatial approaches invites future work to design optimal approaches that incorporate spatiotemporal dependencies into feature selection for decoding.
资助项目	W.M. Keck Foundation ; National Institute of Health[5T90DA022768] ; Staglin Center for Center for Cognitive Neuroscience ; University of Queensland International PhD Scholarship
WOS关键词	SUPPORT VECTOR MACHINES ; FMRI ; REPRESENTATIONS ; CLASSIFICATION ; RESPONSES ; STIMULUS ; DISCRIMINATION ; EXPECTATION ; INFORMATION ; REGRESSION
WOS研究方向	Biochemistry & Molecular Biology ; Neurosciences & Neurology
语种	英语
出版者	ELSEVIER
WOS记录号	WOS:000580629000001
资助机构	W.M. Keck Foundation ; National Institute of Health ; Staglin Center for Center for Cognitive Neuroscience ; University of Queensland International PhD Scholarship
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/42141]
专题	自动化研究所_脑网络组研究中心
通讯作者	Choupan, Jeiran
作者单位	1.Univ Calif Los Angeles, Ctr Cognit Neurosci, Los Angeles, CA USA 2.Univ Queensland, Ctr Adv Imaging, Brisbane, Qld, Australia 3.Univ Queensland, Queensland Brain Inst, Brisbane, Qld, Australia 4.Univ Southern Calif, USC Dornsife Coll Letters Arts & Sci, Dept Psychol, Los Angeles, CA 90007 USA 5.Univ Southern Calif, Keck Sch Med, USC Stevens Neuroimaging & Informat Inst, Lab Neuro Imaging, Los Angeles, CA 90007 USA 6.Univ Queensland, Sch Informat Technol & Elect Engn, Brisbane, Qld, Australia 7.UCF, Modeling & Simulat Dept, Orlando, FL USA 8.UCF, Comp Sci Dept, Orlando, FL USA 9.Univ Calif Los Angeles, Inst Neuropsychiat, 760 Westwood Plaza, Los Angeles, CA 90024 USA 10.Univ Calif Los Angeles, Dept Psychiat & Behav Sci, Los Angeles, CA USA
推荐引用方式 GB/T 7714	Choupan, Jeiran,Douglas, Pamela K.,Gal, Yaniv,et al. Temporal embedding and spatiotemporal feature selection boost multi-voxel pattern analysis decoding accuracy[J]. JOURNAL OF NEUROSCIENCE METHODS,2020,345:13.
APA	Choupan, Jeiran,Douglas, Pamela K.,Gal, Yaniv,Cohen, Mark S.,Reutens, David C.,&Yang, Zhengyi.(2020).Temporal embedding and spatiotemporal feature selection boost multi-voxel pattern analysis decoding accuracy.JOURNAL OF NEUROSCIENCE METHODS,345,13.
MLA	Choupan, Jeiran,et al."Temporal embedding and spatiotemporal feature selection boost multi-voxel pattern analysis decoding accuracy".JOURNAL OF NEUROSCIENCE METHODS 345(2020):13.