Dimensionality of ICA in resting-state fMRI investigated by feature optimized classification of independent components with SVM

Different machine learning algorithms have recently been used for assisting automated classification of independent component analysis (ICA) results from resting-state fMRI data. The success of this approach relies on identification of artifact components and meaningful functional networks. A limiting factor of ICA is the uncertainty of the number of independent components (NIC). We aim to develop a framework based on support vector machines (SVM) and optimized feature-selection for automated classification of independent components (ICs) and use the framework to investigate the effects of input NIC on the ICA results. Seven different resting-state fMRI datasets were studied. 18 features were devised by mimicking the empirical criteria for manual evaluation. The five most significant (p < 0.01) features were identified by general linear modeling and used to generate a classification model for the framework. This feature-optimized classification of ICs with SVM (FOCIS) framework was used to classify both group and single subject ICA results. The classification results obtained using FOCIS and previously published FSL-FIX were compared against manually evaluated results. On average the false negative rate in identifying artifact contaminated ICs for FOCIS and FSL-FIX were 98.27 and 92.34%, respectively. The number of artifact and functional network components increased almost linearly with the input NIC. Through tracking, we demonstrate that incrementing NIC affects most ICs when NIC < 33, whereas only a few limited ICs are affected by direct splitting when NIC is incremented beyond NIC > 40. For a given IC, its changes with increasing NIC are individually specific irrespective whether the component is a potential resting-state functional network or an artifact component. Using FOCIS, we investigated experimentally the ICA dimensionality of resting-state fMRI datasets and found that the input NIC can critically affect the ICA results of resting-state fMRI data.

[1]  S. Petersen,et al.  Concepts and principles in the analysis of brain networks , 2011, Annals of the New York Academy of Sciences.

[2]  Xiangyu Long,et al.  Functional segmentation of the brain cortex using high model order group PICA , 2009, Human brain mapping.

[3]  Tülay Adali,et al.  Estimating the number of independent components for functional magnetic resonance imaging data , 2007, Human brain mapping.

[4]  Stephen M Smith,et al.  Correspondence of the brain's functional architecture during activation and rest , 2009, Proceedings of the National Academy of Sciences.

[5]  V. Haughton,et al.  Mapping functionally related regions of brain with functional connectivity MR imaging. , 2000, AJNR. American journal of neuroradiology.

[6]  S. Rombouts,et al.  Consistent resting-state networks across healthy subjects , 2006, Proceedings of the National Academy of Sciences.

[7]  Graeme D. Jackson,et al.  An Automated Method for Identifying Artifact in Independent Component Analysis of Resting-State fMRI , 2013, Front. Hum. Neurosci..

[8]  C. Windischberger,et al.  Group ICA of resting-state data: a comparison , 2010, Magnetic Resonance Materials in Physics, Biology and Medicine.

[9]  V. Calhoun,et al.  A Robust Classifier to Distinguish Noise from fMRI Independent Components , 2014, PloS one.

[10]  Koene R. A. Van Dijk,et al.  Template based rotation: A method for functional connectivity analysis with a priori templates , 2014, NeuroImage.

[11]  Rex E. Jung,et al.  A Baseline for the Multivariate Comparison of Resting-State Networks , 2011, Front. Syst. Neurosci..

[12]  Chih-Jen Lin,et al.  LIBSVM: A library for support vector machines , 2011, TIST.

[13]  Xiaoping Xie,et al.  Estimating intrinsic dimensionality of fMRI dataset incorporating an AR(1) noise model with cubic spline interpolation , 2009, Neurocomputing.

[14]  Stephen M. Smith,et al.  Probabilistic independent component analysis for functional magnetic resonance imaging , 2004, IEEE Transactions on Medical Imaging.

[15]  D. F. Andrews,et al.  PLOTS OF HIGH-DIMENSIONAL DATA , 1972 .

[16]  Vince D. Calhoun,et al.  An ICA-based method for the identification of optimal FMRI features and components using combined group-discriminative techniques , 2009, NeuroImage.

[17]  Chih-Jen Lin,et al.  Newton's Method for Large Bound-Constrained Optimization Problems , 1999, SIAM J. Optim..

[18]  T. Adali,et al.  Ieee Workshop on Machine Learning for Signal Processing Semi-blind Ica of Fmri: a Method for Utilizing Hypothesis-derived Time Courses in a Spatial Ica Analysis , 2022 .

[19]  Bernhard Schölkopf,et al.  A tutorial on support vector regression , 2004, Stat. Comput..

[20]  O. Tervonen,et al.  The effect of model order selection in group PICA , 2010, Human brain mapping.

[21]  Yihong Yang,et al.  A new approach to estimating the signal dimension of concatenated resting-state functional MRI data sets. , 2010, Magnetic resonance imaging.

[22]  Ludovica Griffanti,et al.  Automatic denoising of functional MRI data: Combining independent component analysis and hierarchical fusion of classifiers , 2014, NeuroImage.

[23]  Rainer Goebel,et al.  Classification of fMRI independent components using IC-fingerprints and support vector machine classifiers , 2007, NeuroImage.

[24]  Wen-Ming Luh,et al.  Differentiating BOLD and non-BOLD signals in fMRI time series using multi-echo EPI , 2012, NeuroImage.

[25]  O. Tervonen,et al.  Neuroglial Plasticity at Striatal Glutamatergic Synapses in Parkinson's Disease , 2011, Front. Syst. Neurosci..

[26]  Yanlu Wang,et al.  Analysis of Whole-Brain Resting-State fMRI Data Using Hierarchical Clustering Approach , 2013, PloS one.

[27]  Tohru Kiryu,et al.  Fast and precise independent component analysis for high field fMRI time series tailored using prior information on spatiotemporal structure , 2002, Human brain mapping.

[28]  Hang Joon Jo,et al.  Mapping sources of correlation in resting state FMRI, with artifact detection and removal , 2010, NeuroImage.

[29]  David R. Musicant,et al.  Successive overrelaxation for support vector machines , 1999, IEEE Trans. Neural Networks.

[30]  Stephen M. Smith,et al.  Investigations into resting-state connectivity using independent component analysis , 2005, Philosophical Transactions of the Royal Society B: Biological Sciences.

[31]  Marisa O. Hollinshead,et al.  The organization of the human cerebral cortex estimated by intrinsic functional connectivity. , 2011, Journal of neurophysiology.

[32]  Mark W. Woolrich,et al.  Advances in functional and structural MR image analysis and implementation as FSL , 2004, NeuroImage.

[33]  Christian F. Beckmann,et al.  Modelling with independent components , 2012, NeuroImage.

[34]  Mark W. Woolrich,et al.  Bayesian analysis of neuroimaging data in FSL , 2009, NeuroImage.

[35]  Rainer Goebel,et al.  Spatial independent component analysis of functional magnetic resonance imaging time-series: characterization of the cortical components , 2002, Neurocomputing.

[36]  Andreas Bartels,et al.  Brain dynamics during natural viewing conditions—A new guide for mapping connectivity in vivo , 2005, NeuroImage.

[37]  Vladimir Vapnik,et al.  The Support Vector Method , 1997, ICANN.

[38]  Allen R. Braun,et al.  Denoising the speaking brain: Toward a robust technique for correcting artifact-contaminated fMRI data under severe motion , 2014, NeuroImage.

[39]  Alan L. Yuille,et al.  Performance comparison of machine learning algorithms and number of independent components used in fMRI decoding of belief vs. disbelief , 2011, NeuroImage.

[40]  Habib Benali,et al.  CORSICA: correction of structured noise in fMRI by automatic identification of ICA components. , 2007, Magnetic resonance imaging.

[41]  J. Pekar,et al.  A method for making group inferences from functional MRI data using independent component analysis , 2001, Human brain mapping.

[42]  Alexander J. Smola,et al.  Support Vector Method for Function Approximation, Regression Estimation and Signal Processing , 1996, NIPS.

[43]  H. Akaike.  A new look at the statistical model identification , 1974 .

[44]  Rajesh Nandy,et al.  Estimation of the intrinsic dimensionality of fMRI data , 2006, NeuroImage.

[45]  Timothy O. Laumann,et al.  An approach for parcellating human cortical areas using resting-state correlations , 2014, NeuroImage.

[46]  Steen Moeller,et al.  ICA-based artefact removal and accelerated fMRI acquisition for improved resting state network imaging , 2014, NeuroImage.

[47]  Kurt Hornik,et al.  kernlab - An S4 Package for Kernel Methods in R , 2004 .