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 A. Lindquist, Ciprian M. Crainiceanu

Research output: Contribution to journalArticlepeer-review

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

Keywords

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

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Fingerprint Dive into the research topics of 'Two-way principal component analysis for matrix-variate data, with an application to functional magnetic resonance imaging data'. Together they form a unique fingerprint.

Cite this