A multiple kernel learning approach to perform classification of groups from complex-valued fMRI data analysis: Application to schizophrenia

Eduardo Castro, Vanessa Gómez-Verdejo, Manel Martínez-Ramón, Kent A. Kiehl, Vince Daniel Calhoun

Research output: Contribution to journalArticle

Abstract

FMRI data are acquired as complex-valued spatiotemporal images. Despite the fact that several studies have identified the presence of novel information in the phase images, they are usually discarded due to their noisy nature. Several approaches have been devised to incorporate magnitude and phase data, but none of them has performed between-group inference or classification. Multiple kernel learning (MKL) is a powerful field of machine learning that finds an automatic combination of kernel functions that can be applied to multiple data sources. By analyzing this combination of kernels, the most informative data sources can be found, hence providing a better understanding of the analyzed learning task. This paper presents a methodology based on a new MKL algorithm (ν-MKL) capable of achieving a tunable sparse selection of features' sets (brain regions' patterns) that improves the classification accuracy rate of healthy controls and schizophrenia patients by 5% when phase data is included. In addition, the proposed method achieves accuracy rates that are equivalent to those obtained by the state of the art lp-norm MKL algorithm on the schizophrenia dataset and we argue that it better identifies the brain regions that show discriminative activation between groups. This claim is supported by the more accurate detection achieved by ν-MKL of the degree of information present on regions of spatial maps extracted from a simulated fMRI dataset. In summary, we present an MKL-based methodology that improves schizophrenia characterization by using both magnitude and phase fMRI data and is also capable of detecting the brain regions that convey most of the discriminative information between patients and controls.

Original languageEnglish (US)
Pages (from-to)1-17
Number of pages17
JournalNeuroImage
Volume87
DOIs
StatePublished - Feb 15 2014
Externally publishedYes

Fingerprint

Schizophrenia
Magnetic Resonance Imaging
Learning
Information Storage and Retrieval
Brain

Keywords

  • Complex-valued fMRI data
  • Feature selection
  • Independent component analysis
  • Multiple kernel learning
  • Schizophrenia
  • Support vector machines

ASJC Scopus subject areas

  • Cognitive Neuroscience
  • Neurology

Cite this

A multiple kernel learning approach to perform classification of groups from complex-valued fMRI data analysis : Application to schizophrenia. / Castro, Eduardo; Gómez-Verdejo, Vanessa; Martínez-Ramón, Manel; Kiehl, Kent A.; Calhoun, Vince Daniel.

In: NeuroImage, Vol. 87, 15.02.2014, p. 1-17.

Research output: Contribution to journalArticle

Castro, Eduardo ; Gómez-Verdejo, Vanessa ; Martínez-Ramón, Manel ; Kiehl, Kent A. ; Calhoun, Vince Daniel. / A multiple kernel learning approach to perform classification of groups from complex-valued fMRI data analysis : Application to schizophrenia. In: NeuroImage. 2014 ; Vol. 87. pp. 1-17.
@article{bd92d3e381d846ed81cde3728148258a,
title = "A multiple kernel learning approach to perform classification of groups from complex-valued fMRI data analysis: Application to schizophrenia",
abstract = "FMRI data are acquired as complex-valued spatiotemporal images. Despite the fact that several studies have identified the presence of novel information in the phase images, they are usually discarded due to their noisy nature. Several approaches have been devised to incorporate magnitude and phase data, but none of them has performed between-group inference or classification. Multiple kernel learning (MKL) is a powerful field of machine learning that finds an automatic combination of kernel functions that can be applied to multiple data sources. By analyzing this combination of kernels, the most informative data sources can be found, hence providing a better understanding of the analyzed learning task. This paper presents a methodology based on a new MKL algorithm (ν-MKL) capable of achieving a tunable sparse selection of features' sets (brain regions' patterns) that improves the classification accuracy rate of healthy controls and schizophrenia patients by 5{\%} when phase data is included. In addition, the proposed method achieves accuracy rates that are equivalent to those obtained by the state of the art lp-norm MKL algorithm on the schizophrenia dataset and we argue that it better identifies the brain regions that show discriminative activation between groups. This claim is supported by the more accurate detection achieved by ν-MKL of the degree of information present on regions of spatial maps extracted from a simulated fMRI dataset. In summary, we present an MKL-based methodology that improves schizophrenia characterization by using both magnitude and phase fMRI data and is also capable of detecting the brain regions that convey most of the discriminative information between patients and controls.",
keywords = "Complex-valued fMRI data, Feature selection, Independent component analysis, Multiple kernel learning, Schizophrenia, Support vector machines",
author = "Eduardo Castro and Vanessa G{\'o}mez-Verdejo and Manel Mart{\'i}nez-Ram{\'o}n and Kiehl, {Kent A.} and Calhoun, {Vince Daniel}",
year = "2014",
month = "2",
day = "15",
doi = "10.1016/j.neuroimage.2013.10.065",
language = "English (US)",
volume = "87",
pages = "1--17",
journal = "NeuroImage",
issn = "1053-8119",
publisher = "Academic Press Inc.",

}

TY - JOUR

T1 - A multiple kernel learning approach to perform classification of groups from complex-valued fMRI data analysis

T2 - Application to schizophrenia

AU - Castro, Eduardo

AU - Gómez-Verdejo, Vanessa

AU - Martínez-Ramón, Manel

AU - Kiehl, Kent A.

AU - Calhoun, Vince Daniel

PY - 2014/2/15

Y1 - 2014/2/15

N2 - FMRI data are acquired as complex-valued spatiotemporal images. Despite the fact that several studies have identified the presence of novel information in the phase images, they are usually discarded due to their noisy nature. Several approaches have been devised to incorporate magnitude and phase data, but none of them has performed between-group inference or classification. Multiple kernel learning (MKL) is a powerful field of machine learning that finds an automatic combination of kernel functions that can be applied to multiple data sources. By analyzing this combination of kernels, the most informative data sources can be found, hence providing a better understanding of the analyzed learning task. This paper presents a methodology based on a new MKL algorithm (ν-MKL) capable of achieving a tunable sparse selection of features' sets (brain regions' patterns) that improves the classification accuracy rate of healthy controls and schizophrenia patients by 5% when phase data is included. In addition, the proposed method achieves accuracy rates that are equivalent to those obtained by the state of the art lp-norm MKL algorithm on the schizophrenia dataset and we argue that it better identifies the brain regions that show discriminative activation between groups. This claim is supported by the more accurate detection achieved by ν-MKL of the degree of information present on regions of spatial maps extracted from a simulated fMRI dataset. In summary, we present an MKL-based methodology that improves schizophrenia characterization by using both magnitude and phase fMRI data and is also capable of detecting the brain regions that convey most of the discriminative information between patients and controls.

AB - FMRI data are acquired as complex-valued spatiotemporal images. Despite the fact that several studies have identified the presence of novel information in the phase images, they are usually discarded due to their noisy nature. Several approaches have been devised to incorporate magnitude and phase data, but none of them has performed between-group inference or classification. Multiple kernel learning (MKL) is a powerful field of machine learning that finds an automatic combination of kernel functions that can be applied to multiple data sources. By analyzing this combination of kernels, the most informative data sources can be found, hence providing a better understanding of the analyzed learning task. This paper presents a methodology based on a new MKL algorithm (ν-MKL) capable of achieving a tunable sparse selection of features' sets (brain regions' patterns) that improves the classification accuracy rate of healthy controls and schizophrenia patients by 5% when phase data is included. In addition, the proposed method achieves accuracy rates that are equivalent to those obtained by the state of the art lp-norm MKL algorithm on the schizophrenia dataset and we argue that it better identifies the brain regions that show discriminative activation between groups. This claim is supported by the more accurate detection achieved by ν-MKL of the degree of information present on regions of spatial maps extracted from a simulated fMRI dataset. In summary, we present an MKL-based methodology that improves schizophrenia characterization by using both magnitude and phase fMRI data and is also capable of detecting the brain regions that convey most of the discriminative information between patients and controls.

KW - Complex-valued fMRI data

KW - Feature selection

KW - Independent component analysis

KW - Multiple kernel learning

KW - Schizophrenia

KW - Support vector machines

UR - http://www.scopus.com/inward/record.url?scp=84890888665&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84890888665&partnerID=8YFLogxK

U2 - 10.1016/j.neuroimage.2013.10.065

DO - 10.1016/j.neuroimage.2013.10.065

M3 - Article

C2 - 24225489

AN - SCOPUS:84890888665

VL - 87

SP - 1

EP - 17

JO - NeuroImage

JF - NeuroImage

SN - 1053-8119

ER -