A Bayesian hierarchical spatial model for dental caries assessment using non-Gaussian Markov random fields

Ick Hoon Jin, Ying Yuan, Dipankar Bandyopadhyay

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Research in dental caries generates data with two levels of hierarchy: that of a tooth overall and that of the different surfaces of the tooth. The outcomes often exhibit spatial referencing among neighboring teeth and surfaces, that is, the disease status of a tooth or surface might be influenced by the status of a set of proximal teeth/surfaces. Assessments of dental caries (tooth decay) at the tooth level yield binary outcomes indicating the presence/absence of teeth, and trinary outcomes at the surface level indicating healthy, decayed or filled surfaces. The presence of these mixed discrete responses complicates the data analysis under a unified framework. To mitigate complications, we develop a Bayesian two-level hierarchical model under suitable (spatial) Markov random field assumptions that accommodates the natural hierarchy within the mixed responses. At the first level, we utilize an autologistic model to accommodate the spatial dependence for the tooth-level binary outcomes. For the second level and conditioned on a tooth being nonmissing, we utilize a Potts model to accommodate the spatial referencing for the surface-level trinary outcomes. The regression models at both levels were controlled for plausible covariates (risk factors) of caries and remain connected through shared parameters. To tackle the computational challenges in our Bayesian estimation scheme caused due to the doubly-intractable normalizing constant, we employ a double Metropolis–Hastings sampler.We compare and contrast our model performances to the standard nonspatial (naive) model using a small simulation study, and illustrate via an application to a clinical dataset on dental caries.

Original languageEnglish
Pages (from-to)884-905
Number of pages22
JournalAnnals of Applied Statistics
Volume10
Issue number2
DOIs
Publication statusPublished - 2016 Jun

Fingerprint

Spatial Model
Hierarchical Model
Random Field
Binary Outcomes
Potts model
Normalizing Constant
Metropolis-Hastings
Spatial Dependence
Bayesian Estimation
Potts Model
Performance Model
Risk Factors
Spatial model
Random field
Complications
Covariates
Regression Model
Data analysis
Simulation Study
Decay

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Modelling and Simulation
  • Statistics, Probability and Uncertainty

Cite this

@article{4488fd31fad74a6bbd5355ee6fc3704e,
title = "A Bayesian hierarchical spatial model for dental caries assessment using non-Gaussian Markov random fields",
abstract = "Research in dental caries generates data with two levels of hierarchy: that of a tooth overall and that of the different surfaces of the tooth. The outcomes often exhibit spatial referencing among neighboring teeth and surfaces, that is, the disease status of a tooth or surface might be influenced by the status of a set of proximal teeth/surfaces. Assessments of dental caries (tooth decay) at the tooth level yield binary outcomes indicating the presence/absence of teeth, and trinary outcomes at the surface level indicating healthy, decayed or filled surfaces. The presence of these mixed discrete responses complicates the data analysis under a unified framework. To mitigate complications, we develop a Bayesian two-level hierarchical model under suitable (spatial) Markov random field assumptions that accommodates the natural hierarchy within the mixed responses. At the first level, we utilize an autologistic model to accommodate the spatial dependence for the tooth-level binary outcomes. For the second level and conditioned on a tooth being nonmissing, we utilize a Potts model to accommodate the spatial referencing for the surface-level trinary outcomes. The regression models at both levels were controlled for plausible covariates (risk factors) of caries and remain connected through shared parameters. To tackle the computational challenges in our Bayesian estimation scheme caused due to the doubly-intractable normalizing constant, we employ a double Metropolis–Hastings sampler.We compare and contrast our model performances to the standard nonspatial (naive) model using a small simulation study, and illustrate via an application to a clinical dataset on dental caries.",
author = "Jin, {Ick Hoon} and Ying Yuan and Dipankar Bandyopadhyay",
year = "2016",
month = "6",
doi = "10.1214/16-AOAS917",
language = "English",
volume = "10",
pages = "884--905",
journal = "Annals of Applied Statistics",
issn = "1932-6157",
publisher = "Institute of Mathematical Statistics",
number = "2",

}

A Bayesian hierarchical spatial model for dental caries assessment using non-Gaussian Markov random fields. / Jin, Ick Hoon; Yuan, Ying; Bandyopadhyay, Dipankar.

In: Annals of Applied Statistics, Vol. 10, No. 2, 06.2016, p. 884-905.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A Bayesian hierarchical spatial model for dental caries assessment using non-Gaussian Markov random fields

AU - Jin, Ick Hoon

AU - Yuan, Ying

AU - Bandyopadhyay, Dipankar

PY - 2016/6

Y1 - 2016/6

N2 - Research in dental caries generates data with two levels of hierarchy: that of a tooth overall and that of the different surfaces of the tooth. The outcomes often exhibit spatial referencing among neighboring teeth and surfaces, that is, the disease status of a tooth or surface might be influenced by the status of a set of proximal teeth/surfaces. Assessments of dental caries (tooth decay) at the tooth level yield binary outcomes indicating the presence/absence of teeth, and trinary outcomes at the surface level indicating healthy, decayed or filled surfaces. The presence of these mixed discrete responses complicates the data analysis under a unified framework. To mitigate complications, we develop a Bayesian two-level hierarchical model under suitable (spatial) Markov random field assumptions that accommodates the natural hierarchy within the mixed responses. At the first level, we utilize an autologistic model to accommodate the spatial dependence for the tooth-level binary outcomes. For the second level and conditioned on a tooth being nonmissing, we utilize a Potts model to accommodate the spatial referencing for the surface-level trinary outcomes. The regression models at both levels were controlled for plausible covariates (risk factors) of caries and remain connected through shared parameters. To tackle the computational challenges in our Bayesian estimation scheme caused due to the doubly-intractable normalizing constant, we employ a double Metropolis–Hastings sampler.We compare and contrast our model performances to the standard nonspatial (naive) model using a small simulation study, and illustrate via an application to a clinical dataset on dental caries.

AB - Research in dental caries generates data with two levels of hierarchy: that of a tooth overall and that of the different surfaces of the tooth. The outcomes often exhibit spatial referencing among neighboring teeth and surfaces, that is, the disease status of a tooth or surface might be influenced by the status of a set of proximal teeth/surfaces. Assessments of dental caries (tooth decay) at the tooth level yield binary outcomes indicating the presence/absence of teeth, and trinary outcomes at the surface level indicating healthy, decayed or filled surfaces. The presence of these mixed discrete responses complicates the data analysis under a unified framework. To mitigate complications, we develop a Bayesian two-level hierarchical model under suitable (spatial) Markov random field assumptions that accommodates the natural hierarchy within the mixed responses. At the first level, we utilize an autologistic model to accommodate the spatial dependence for the tooth-level binary outcomes. For the second level and conditioned on a tooth being nonmissing, we utilize a Potts model to accommodate the spatial referencing for the surface-level trinary outcomes. The regression models at both levels were controlled for plausible covariates (risk factors) of caries and remain connected through shared parameters. To tackle the computational challenges in our Bayesian estimation scheme caused due to the doubly-intractable normalizing constant, we employ a double Metropolis–Hastings sampler.We compare and contrast our model performances to the standard nonspatial (naive) model using a small simulation study, and illustrate via an application to a clinical dataset on dental caries.

UR - http://www.scopus.com/inward/record.url?scp=84979936602&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84979936602&partnerID=8YFLogxK

U2 - 10.1214/16-AOAS917

DO - 10.1214/16-AOAS917

M3 - Article

AN - SCOPUS:84979936602

VL - 10

SP - 884

EP - 905

JO - Annals of Applied Statistics

JF - Annals of Applied Statistics

SN - 1932-6157

IS - 2

ER -