Premium
A shrinkage method for causal network detection of brain regions
Author(s) -
Ahmad Fayyaz,
Lee Namgil,
Kim Eunwoo,
Kim SungHo,
Park HyunWook
Publication year - 2013
Publication title -
international journal of imaging systems and technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.359
H-Index - 47
eISSN - 1098-1098
pISSN - 0899-9457
DOI - 10.1002/ima.22047
Subject(s) - autoregressive model , shrinkage , computer science , artificial intelligence , covariance , pattern recognition (psychology) , covariance matrix , functional magnetic resonance imaging , neuroimaging , statistics , machine learning , algorithm , mathematics , psychology , neuroscience
We present a computationally as well as statistically efficient method of inferring causal networks for the brain regions. It is based on James‐Stein‐type shrinkage estimation of covariance matrix, suggested by (Opgen‐Rhein and Strimmer, BMC Syst Biol 1 ([R. Opgen‐Rhein, 2007]), 37‐40), among different brain regions of interest of the functional magnetic resonance imaging (fMRI) experiment, that enhance the accuracy of vector autoregressive (VAR) model coefficient estimates. We have shown that this approach is well suited for the small number of samples in time and large number of brain regions encountered in real fMRI experiments of seventeen healthy individuals. © 2013 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 23, 140–146, 2013