Population Value Decomposition, a Framework for the Analysis of Image Populations

Images, often stored in multidimensional arrays, are fast becoming ubiquitous in medical and public health research. Analyzing populations of images is a statistical problem that raises a host of daunting challenges. The most significant challenge is the massive size of the datasets incorporating im...

Full description

Saved in:
Bibliographic Details
Published inJournal of the American Statistical Association Vol. 106; no. 495; pp. 775 - 790
Main Authors Crainiceanu, Ciprian M., Caffo, Brian S., Luo, Sheng, Zipunnikov, Vadim M., Punjabi, Naresh M.
Format Journal Article
LanguageEnglish
Published Alexandria, VA American Statistical Association 01.09.2011
Subjects
Applications
Applications and Case Studies
Data processing
Eigenfunctions
Eigenvalues
Eigenvectors
Exact sciences and technology
Frequency ranges
General topics
Image analysis
Inference from stochastic processes; time series analysis
Magnetic resonance imaging
Mathematics
Matrices
Medical care
Medical sciences
Methodology
Multivariate analysis
Principal components analysis
Probability and statistics
Public health
Sciences and techniques of general use
Sleep
Statistical models
Statistical variance
Statistics
Biometrics
Decomposition method
Population analysis
Decomposition
Electroencephalography
Statistical estimation
Signal extraction
Multivariate analysis
Stochastic process
Statistical method
Random field
Computational geometry
Image analysis
Reduction method
Medical science
Cohort study
Computer graphics
Public health
Online AccessGet full text

Cover

More Information
Summary:Images, often stored in multidimensional arrays, are fast becoming ubiquitous in medical and public health research. Analyzing populations of images is a statistical problem that raises a host of daunting challenges. The most significant challenge is the massive size of the datasets incorporating images recorded for hundreds or thousands of subjects at multiple visits. We introduce the population value decomposition (PVD), a general method for simultaneous dimensionality reduction of large populations of massive images. We show how PVD can be seamlessly incorporated into statistical modeling, leading to a new, transparent, and rapid inferential framework. Our PVD methodology was motivated by and applied to the Sleep Heart Health Study, the largest community-based cohort study of sleep containing more than 85 billion observations on thousands of subjects at two visits. This article has supplementary material online.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ObjectType-Article-2
ObjectType-Feature-1
ISSN:0162-1459
1537-274X
DOI:10.1198/jasa.2011.ap10089