Two-way principal component analysis for matrix-variate data, with an application to functional magnetic resonance imaging data

Lei Huang, Philip T. Reiss, Luo Xiao, Vadim Zipunnikov, Martin Lindquist, Ciprian M Crainiceanu

Research output: Contribution to journalArticle

Abstract

Many modern neuroimaging studies acquire large spatial images of the brain observed sequentially over time. Such data are often stored in the forms of matrices. To model these matrix-variate data we introduce a class of separable processes using explicit latent process modeling. To account for the size and two-way structure of the data, we extend principal component analysis to achieve dimensionality reduction at the individual level. We introduce necessary identifiability conditions for each model and develop scalable estimationprocedures.Themethodismotivatedbyandappliedtoafunctionalmagneticresonanceimaging study designed to analyze the relationship between pain and brain activity.

Original languageEnglish (US)
Pages (from-to)214-229
Number of pages16
JournalBiostatistics
Volume18
Issue number2
DOIs
StatePublished - Apr 1 2017

Fingerprint

Functional Magnetic Resonance Imaging
Principal Component Analysis
Magnetic Resonance Imaging
Brain
Neuroimaging
Latent Process
Identifiability
Matrix Models
Pain
Dimensionality Reduction
Process Modeling
Necessary
Principal component analysis
Functional magnetic resonance imaging
Model
Dimensionality reduction
Process modeling

Keywords

  • fMRI
  • Latent process modeling
  • Matrix-variate
  • Principal component analysis
  • Separability

ASJC Scopus subject areas

  • Statistics and Probability
  • Medicine(all)
  • Statistics, Probability and Uncertainty

Cite this

Two-way principal component analysis for matrix-variate data, with an application to functional magnetic resonance imaging data. / Huang, Lei; Reiss, Philip T.; Xiao, Luo; Zipunnikov, Vadim; Lindquist, Martin; Crainiceanu, Ciprian M.

In: Biostatistics, Vol. 18, No. 2, 01.04.2017, p. 214-229.

Research output: Contribution to journalArticle

@article{ca06e6d42b534a0daafc90b187f56eca,
title = "Two-way principal component analysis for matrix-variate data, with an application to functional magnetic resonance imaging data",
abstract = "Many modern neuroimaging studies acquire large spatial images of the brain observed sequentially over time. Such data are often stored in the forms of matrices. To model these matrix-variate data we introduce a class of separable processes using explicit latent process modeling. To account for the size and two-way structure of the data, we extend principal component analysis to achieve dimensionality reduction at the individual level. We introduce necessary identifiability conditions for each model and develop scalable estimationprocedures.Themethodismotivatedbyandappliedtoafunctionalmagneticresonanceimaging study designed to analyze the relationship between pain and brain activity.",
keywords = "fMRI, Latent process modeling, Matrix-variate, Principal component analysis, Separability",
author = "Lei Huang and Reiss, {Philip T.} and Luo Xiao and Vadim Zipunnikov and Martin Lindquist and Crainiceanu, {Ciprian M}",
year = "2017",
month = "4",
day = "1",
doi = "10.1093/biostatistics/kxw040",
language = "English (US)",
volume = "18",
pages = "214--229",
journal = "Biostatistics",
issn = "1465-4644",
publisher = "Oxford University Press",
number = "2",

}

TY - JOUR

T1 - Two-way principal component analysis for matrix-variate data, with an application to functional magnetic resonance imaging data

AU - Huang, Lei

AU - Reiss, Philip T.

AU - Xiao, Luo

AU - Zipunnikov, Vadim

AU - Lindquist, Martin

AU - Crainiceanu, Ciprian M

PY - 2017/4/1

Y1 - 2017/4/1

N2 - Many modern neuroimaging studies acquire large spatial images of the brain observed sequentially over time. Such data are often stored in the forms of matrices. To model these matrix-variate data we introduce a class of separable processes using explicit latent process modeling. To account for the size and two-way structure of the data, we extend principal component analysis to achieve dimensionality reduction at the individual level. We introduce necessary identifiability conditions for each model and develop scalable estimationprocedures.Themethodismotivatedbyandappliedtoafunctionalmagneticresonanceimaging study designed to analyze the relationship between pain and brain activity.

AB - Many modern neuroimaging studies acquire large spatial images of the brain observed sequentially over time. Such data are often stored in the forms of matrices. To model these matrix-variate data we introduce a class of separable processes using explicit latent process modeling. To account for the size and two-way structure of the data, we extend principal component analysis to achieve dimensionality reduction at the individual level. We introduce necessary identifiability conditions for each model and develop scalable estimationprocedures.Themethodismotivatedbyandappliedtoafunctionalmagneticresonanceimaging study designed to analyze the relationship between pain and brain activity.

KW - fMRI

KW - Latent process modeling

KW - Matrix-variate

KW - Principal component analysis

KW - Separability

UR - http://www.scopus.com/inward/record.url?scp=85020182340&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85020182340&partnerID=8YFLogxK

U2 - 10.1093/biostatistics/kxw040

DO - 10.1093/biostatistics/kxw040

M3 - Article

VL - 18

SP - 214

EP - 229

JO - Biostatistics

JF - Biostatistics

SN - 1465-4644

IS - 2

ER -