Literature DB >> 28322859

Decoding the encoding of functional brain networks: An fMRI classification comparison of non-negative matrix factorization (NMF), independent component analysis (ICA), and sparse coding algorithms.

Jianwen Xie1, Pamela K Douglas2, Ying Nian Wu1, Arthur L Brody2, Ariana E Anderson3.   

Abstract

BACKGROUND: Brain networks in fMRI are typically identified using spatial independent component analysis (ICA), yet other mathematical constraints provide alternate biologically-plausible frameworks for generating brain networks. Non-negative matrix factorization (NMF) would suppress negative BOLD signal by enforcing positivity. Spatial sparse coding algorithms (L1 Regularized Learning and K-SVD) would impose local specialization and a discouragement of multitasking, where the total observed activity in a single voxel originates from a restricted number of possible brain networks. NEW
METHOD: The assumptions of independence, positivity, and sparsity to encode task-related brain networks are compared; the resulting brain networks within scan for different constraints are used as basis functions to encode observed functional activity. These encodings are then decoded using machine learning, by using the time series weights to predict within scan whether a subject is viewing a video, listening to an audio cue, or at rest, in 304 fMRI scans from 51 subjects. RESULTS AND COMPARISON WITH EXISTING
METHOD: The sparse coding algorithm of L1 Regularized Learning outperformed 4 variations of ICA (p<0.001) for predicting the task being performed within each scan using artifact-cleaned components. The NMF algorithms, which suppressed negative BOLD signal, had the poorest accuracy compared to the ICA and sparse coding algorithms. Holding constant the effect of the extraction algorithm, encodings using sparser spatial networks (containing more zero-valued voxels) had higher classification accuracy (p<0.001). Lower classification accuracy occurred when the extracted spatial maps contained more CSF regions (p<0.001).
CONCLUSION: The success of sparse coding algorithms suggests that algorithms which enforce sparsity, discourage multitasking, and promote local specialization may capture better the underlying source processes than those which allow inexhaustible local processes such as ICA. Negative BOLD signal may capture task-related activations.
Copyright © 2017 Elsevier B.V. All rights reserved.

Entities:  

Keywords:  Artifacts; Classification; FMRI; ICA; Image processing; Independent component analysis; K-SVD; L1 Regularized Learning; Machine learning; NMF; Negative BOLD signal; Non-negative matrix factorization; Pattern recognition; Random forests; Sparsity; Support vector machines

Mesh:

Substances:

Year:  2017        PMID: 28322859      PMCID: PMC5507942          DOI: 10.1016/j.jneumeth.2017.03.008

Source DB:  PubMed          Journal:  J Neurosci Methods        ISSN: 0165-0270            Impact factor:   2.987


  45 in total

1.  Learning the parts of objects by non-negative matrix factorization.

Authors:  D D Lee; H S Seung
Journal:  Nature       Date:  1999-10-21       Impact factor: 49.962

2.  Improved optimization for the robust and accurate linear registration and motion correction of brain images.

Authors:  Mark Jenkinson; Peter Bannister; Michael Brady; Stephen Smith
Journal:  Neuroimage       Date:  2002-10       Impact factor: 6.556

3.  Negative BOLD-fMRI signals in large cerebral veins.

Authors:  Marta Bianciardi; Masaki Fukunaga; Peter van Gelderen; Jacco A de Zwart; Jeff H Duyn
Journal:  J Cereb Blood Flow Metab       Date:  2010-09-22       Impact factor: 6.200

4.  Analysis of fMRI data by blind separation into independent spatial components.

Authors:  M J McKeown; S Makeig; G G Brown; T P Jung; S S Kindermann; A J Bell; T J Sejnowski
Journal:  Hum Brain Mapp       Date:  1998       Impact factor: 5.038

5.  An information-maximization approach to blind separation and blind deconvolution.

Authors:  A J Bell; T J Sejnowski
Journal:  Neural Comput       Date:  1995-11       Impact factor: 2.026

Review 6.  Visual object recognition.

Authors:  N K Logothetis; D L Sheinberg
Journal:  Annu Rev Neurosci       Date:  1996       Impact factor: 12.449

7.  Sparse dictionary learning of resting state fMRI networks.

Authors:  Harini Eavani; Roman Filipovych; Christos Davatzikos; Theodore D Satterthwaite; Raquel E Gur; Ruben C Gur
Journal:  Int Workshop Pattern Recognit Neuroimaging       Date:  2012-07-02

Review 8.  Independent component analysis of functional MRI: what is signal and what is noise?

Authors:  Martin J McKeown; Lars Kai Hansen; Terrence J Sejnowsk
Journal:  Curr Opin Neurobiol       Date:  2003-10       Impact factor: 6.627

9.  Independent component analysis for brain FMRI does indeed select for maximal independence.

Authors:  Vince D Calhoun; Vamsi K Potluru; Ronald Phlypo; Rogers F Silva; Barak A Pearlmutter; Arvind Caprihan; Sergey M Plis; Tülay Adalı
Journal:  PLoS One       Date:  2013-08-29       Impact factor: 3.240

10.  Single trial decoding of belief decision making from EEG and fMRI data using independent components features.

Authors:  Pamela K Douglas; Edward Lau; Ariana Anderson; Austin Head; Wesley Kerr; Margalit Wollner; Daniel Moyer; Wei Li; Mike Durnhofer; Jennifer Bramen; Mark S Cohen
Journal:  Front Hum Neurosci       Date:  2013-07-31       Impact factor: 3.169

View more
  9 in total

1.  Link prediction based on non-negative matrix factorization.

Authors:  Bolun Chen; Fenfen Li; Senbo Chen; Ronglin Hu; Ling Chen
Journal:  PLoS One       Date:  2017-08-30       Impact factor: 3.240

2.  Shift-Invariant Canonical Polyadic Decomposition of Complex-Valued Multi-Subject fMRI Data With a Phase Sparsity Constraint.

Authors:  Li-Dan Kuang; Qiu-Hua Lin; Xiao-Feng Gong; Fengyu Cong; Yu-Ping Wang; Vince D Calhoun
Journal:  IEEE Trans Med Imaging       Date:  2019-08-19       Impact factor: 10.048

3.  Sparse coding reveals greater functional connectivity in female brains during naturalistic emotional experience.

Authors:  Yudan Ren; Jinglei Lv; Lei Guo; Jun Fang; Christine Cong Guo
Journal:  PLoS One       Date:  2017-12-22       Impact factor: 3.240

4.  Altered Functional Connectivity in Children With Low-Function Autism Spectrum Disorders.

Authors:  Shoujun Xu; Meng Li; Chunlan Yang; Xiangling Fang; Miaoting Ye; Lei Wei; Jian Liu; Baojuan Li; Yungen Gan; Binrang Yang; Wenxian Huang; Peng Li; Xianlei Meng; Yunfan Wu; Guihua Jiang
Journal:  Front Neurosci       Date:  2019-08-02       Impact factor: 4.677

5.  A Novel Signal Separation Method Based on Improved Sparse Non-Negative Matrix Factorization.

Authors:  Huaqing Wang; Mengyang Wang; Junlin Li; Liuyang Song; Yansong Hao
Journal:  Entropy (Basel)       Date:  2019-04-28       Impact factor: 2.524

Review 6.  An Overview of ICA/BSS-Based Application to Alzheimer's Brain Signal Processing.

Authors:  Wenlu Yang; Alexander Pilozzi; Xudong Huang
Journal:  Biomedicines       Date:  2021-04-06

7.  Co-sparse Non-negative Matrix Factorization.

Authors:  Fan Wu; Jiahui Cai; Canhong Wen; Haizhu Tan
Journal:  Front Neurosci       Date:  2022-01-12       Impact factor: 4.677

8.  Brain disorder prediction with dynamic multivariate spatio-temporal features: Application to Alzheimer's disease and autism spectrum disorder.

Authors:  Jianping Qiao; Rong Wang; Hongjia Liu; Guangrun Xu; Zhishun Wang
Journal:  Front Aging Neurosci       Date:  2022-08-30       Impact factor: 5.702

9.  A method to compare the discriminatory power of data-driven methods: Application to ICA and IVA.

Authors:  Yuri Levin-Schwartz; Vince D Calhoun; Tülay Adalı
Journal:  J Neurosci Methods       Date:  2018-10-30       Impact factor: 2.390

  9 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.