Independent Component Analysis for Brain fMRI Does Indeed Select for Maximal Independence

Vince D. Calhoun; Vamsi K. Potluru; Ronald Phlypo; Rogers F. Silva; Barak A. Pearlmutter; Arvind Caprihan; Sergey M. Plis; Tülay Adali

doi:10.1371/journal.pone.0073309

Independent Component Analysis for Brain fMRI Does Indeed Select for Maximal Independence

Vince D. Calhoun
, Vamsi K. Potluru
, Ronald Phlypo
, Rogers F. Silva
, Barak A. Pearlmutter
, Arvind Caprihan
, Sergey M. Plis
, Tülay Adali

Research output: Contribution to journal › Article › peer-review

59 Scopus citations

Abstract

A recent paper by Daubechies et al. claims that two independent component analysis (ICA) algorithms, Infomax and FastICA, which are widely used for functional magnetic resonance imaging (fMRI) analysis, select for sparsity rather than independence. The argument was supported by a series of experiments on synthetic data. We show that these experiments fall short of proving this claim and that the ICA algorithms are indeed doing what they are designed to do: identify maximally independent sources.

Original language	English (US)
Article number	e73309
Journal	PloS one
Volume	8
Issue number	8
DOIs	https://doi.org/10.1371/journal.pone.0073309
State	Published - Aug 29 2013
Externally published	Yes

ASJC Scopus subject areas

General Biochemistry, Genetics and Molecular Biology
General Agricultural and Biological Sciences
General

Access to Document

10.1371/journal.pone.0073309

Cite this

@article{f80522424b1f4e08a6c69765d0c7c93c,

title = "Independent Component Analysis for Brain fMRI Does Indeed Select for Maximal Independence",

abstract = "A recent paper by Daubechies et al. claims that two independent component analysis (ICA) algorithms, Infomax and FastICA, which are widely used for functional magnetic resonance imaging (fMRI) analysis, select for sparsity rather than independence. The argument was supported by a series of experiments on synthetic data. We show that these experiments fall short of proving this claim and that the ICA algorithms are indeed doing what they are designed to do: identify maximally independent sources.",

author = "Calhoun, \{Vince D.\} and Potluru, \{Vamsi K.\} and Ronald Phlypo and Silva, \{Rogers F.\} and Pearlmutter, \{Barak A.\} and Arvind Caprihan and Plis, \{Sergey M.\} and T{\"u}lay Adali",

year = "2013",

month = aug,

day = "29",

doi = "10.1371/journal.pone.0073309",

language = "English (US)",

volume = "8",

journal = "PloS one",

issn = "1932-6203",

publisher = "Public Library of Science",

number = "8",

}

TY - JOUR

T1 - Independent Component Analysis for Brain fMRI Does Indeed Select for Maximal Independence

AU - Calhoun, Vince D.

AU - Potluru, Vamsi K.

AU - Phlypo, Ronald

AU - Silva, Rogers F.

AU - Pearlmutter, Barak A.

AU - Caprihan, Arvind

AU - Plis, Sergey M.

AU - Adali, Tülay

PY - 2013/8/29

Y1 - 2013/8/29

N2 - A recent paper by Daubechies et al. claims that two independent component analysis (ICA) algorithms, Infomax and FastICA, which are widely used for functional magnetic resonance imaging (fMRI) analysis, select for sparsity rather than independence. The argument was supported by a series of experiments on synthetic data. We show that these experiments fall short of proving this claim and that the ICA algorithms are indeed doing what they are designed to do: identify maximally independent sources.

AB - A recent paper by Daubechies et al. claims that two independent component analysis (ICA) algorithms, Infomax and FastICA, which are widely used for functional magnetic resonance imaging (fMRI) analysis, select for sparsity rather than independence. The argument was supported by a series of experiments on synthetic data. We show that these experiments fall short of proving this claim and that the ICA algorithms are indeed doing what they are designed to do: identify maximally independent sources.

UR - https://www.scopus.com/pages/publications/84883245774

UR - https://www.scopus.com/pages/publications/84883245774#tab=citedBy

U2 - 10.1371/journal.pone.0073309

DO - 10.1371/journal.pone.0073309

M3 - Article

C2 - 24009746

AN - SCOPUS:84883245774

SN - 1932-6203

VL - 8

JO - PloS one

JF - PloS one

IS - 8

M1 - e73309

ER -

Independent Component Analysis for Brain fMRI Does Indeed Select for Maximal Independence

Abstract

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this