Combining images across multiple subjects: A study of direct cortical electrical interference

Diana L Miglioretti, C. McCulloch, S. L. Zeger

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

This article introduces a Bayesian hierarchical model for combining information across multiple images. Our work was motivated by an invasive functional brain mapping technique called direct cortical electrical interference that gives a sparse set of binary observations of an underlying "true" region at multiple sites on the brain surface. To model region shapes that may vary widely across individuals, we use mixtures of simple templates, for example, circles. These subject-specific templates are treated as random effects, governed by a set of population templates that make up a population region. The numbers of subject-specific and population templates are treated as unknown variables to be estimated from the data. Conditional on the subject-specific regions, the observed data are modeled using logistic regression. To estimate the variability among images across patients, we develop a measure based on Baddeley's error measure for binary images. Because the dimension of the parameter space changes as the numbers of subject-specific and population templates change, inference is made using reversible jump Markov chain Monte Carlo. Using a hierarchical approach, we may better estimate each individual's region by borrowing strength from other subjects' data, we can estimate a population region by pooling information across subjects, and we can use a collection of data from previous patients to predict the location of a future patient's region of interest. The approach is illustrated with DCEI data collected on 20 patients for two motor tasks: tongue and hand movements.

Original languageEnglish (US)
Pages (from-to)125-135
Number of pages11
JournalJournal of the American Statistical Association
Volume97
Issue number457
DOIs
StatePublished - Mar 2002
Externally publishedYes

Fingerprint

Interference
Template
Estimate
Reversible Jump Markov Chain Monte Carlo
Bayesian Hierarchical Model
Pooling
Binary Image
Logistic Regression
Region of Interest
Random Effects
Parameter Space
Circle
Vary
Binary
Predict
Unknown
Brain
Model

Keywords

  • Binary image variation
  • Functional brain mapping
  • Hierarchical model
  • Reversible jump Markov chain Monte Carlo
  • Template mixture model

ASJC Scopus subject areas

  • Mathematics(all)
  • Statistics and Probability

Cite this

Combining images across multiple subjects : A study of direct cortical electrical interference. / Miglioretti, Diana L; McCulloch, C.; Zeger, S. L.

In: Journal of the American Statistical Association, Vol. 97, No. 457, 03.2002, p. 125-135.

Research output: Contribution to journalArticle

@article{4d8a2e63ee524910a75e1d5100421d97,
title = "Combining images across multiple subjects: A study of direct cortical electrical interference",
abstract = "This article introduces a Bayesian hierarchical model for combining information across multiple images. Our work was motivated by an invasive functional brain mapping technique called direct cortical electrical interference that gives a sparse set of binary observations of an underlying {"}true{"} region at multiple sites on the brain surface. To model region shapes that may vary widely across individuals, we use mixtures of simple templates, for example, circles. These subject-specific templates are treated as random effects, governed by a set of population templates that make up a population region. The numbers of subject-specific and population templates are treated as unknown variables to be estimated from the data. Conditional on the subject-specific regions, the observed data are modeled using logistic regression. To estimate the variability among images across patients, we develop a measure based on Baddeley's error measure for binary images. Because the dimension of the parameter space changes as the numbers of subject-specific and population templates change, inference is made using reversible jump Markov chain Monte Carlo. Using a hierarchical approach, we may better estimate each individual's region by borrowing strength from other subjects' data, we can estimate a population region by pooling information across subjects, and we can use a collection of data from previous patients to predict the location of a future patient's region of interest. The approach is illustrated with DCEI data collected on 20 patients for two motor tasks: tongue and hand movements.",
keywords = "Binary image variation, Functional brain mapping, Hierarchical model, Reversible jump Markov chain Monte Carlo, Template mixture model",
author = "Miglioretti, {Diana L} and C. McCulloch and Zeger, {S. L.}",
year = "2002",
month = "3",
doi = "10.1198/016214502753479284",
language = "English (US)",
volume = "97",
pages = "125--135",
journal = "Journal of the American Statistical Association",
issn = "0162-1459",
publisher = "Taylor and Francis Ltd.",
number = "457",

}

TY - JOUR

T1 - Combining images across multiple subjects

T2 - A study of direct cortical electrical interference

AU - Miglioretti, Diana L

AU - McCulloch, C.

AU - Zeger, S. L.

PY - 2002/3

Y1 - 2002/3

N2 - This article introduces a Bayesian hierarchical model for combining information across multiple images. Our work was motivated by an invasive functional brain mapping technique called direct cortical electrical interference that gives a sparse set of binary observations of an underlying "true" region at multiple sites on the brain surface. To model region shapes that may vary widely across individuals, we use mixtures of simple templates, for example, circles. These subject-specific templates are treated as random effects, governed by a set of population templates that make up a population region. The numbers of subject-specific and population templates are treated as unknown variables to be estimated from the data. Conditional on the subject-specific regions, the observed data are modeled using logistic regression. To estimate the variability among images across patients, we develop a measure based on Baddeley's error measure for binary images. Because the dimension of the parameter space changes as the numbers of subject-specific and population templates change, inference is made using reversible jump Markov chain Monte Carlo. Using a hierarchical approach, we may better estimate each individual's region by borrowing strength from other subjects' data, we can estimate a population region by pooling information across subjects, and we can use a collection of data from previous patients to predict the location of a future patient's region of interest. The approach is illustrated with DCEI data collected on 20 patients for two motor tasks: tongue and hand movements.

AB - This article introduces a Bayesian hierarchical model for combining information across multiple images. Our work was motivated by an invasive functional brain mapping technique called direct cortical electrical interference that gives a sparse set of binary observations of an underlying "true" region at multiple sites on the brain surface. To model region shapes that may vary widely across individuals, we use mixtures of simple templates, for example, circles. These subject-specific templates are treated as random effects, governed by a set of population templates that make up a population region. The numbers of subject-specific and population templates are treated as unknown variables to be estimated from the data. Conditional on the subject-specific regions, the observed data are modeled using logistic regression. To estimate the variability among images across patients, we develop a measure based on Baddeley's error measure for binary images. Because the dimension of the parameter space changes as the numbers of subject-specific and population templates change, inference is made using reversible jump Markov chain Monte Carlo. Using a hierarchical approach, we may better estimate each individual's region by borrowing strength from other subjects' data, we can estimate a population region by pooling information across subjects, and we can use a collection of data from previous patients to predict the location of a future patient's region of interest. The approach is illustrated with DCEI data collected on 20 patients for two motor tasks: tongue and hand movements.

KW - Binary image variation

KW - Functional brain mapping

KW - Hierarchical model

KW - Reversible jump Markov chain Monte Carlo

KW - Template mixture model

UR - http://www.scopus.com/inward/record.url?scp=0036489068&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0036489068&partnerID=8YFLogxK

U2 - 10.1198/016214502753479284

DO - 10.1198/016214502753479284

M3 - Article

AN - SCOPUS:0036489068

VL - 97

SP - 125

EP - 135

JO - Journal of the American Statistical Association

JF - Journal of the American Statistical Association

SN - 0162-1459

IS - 457

ER -