An external field prior for the hidden Potts model, with application to cone-beam computed tomography

By Moores Matt, Hargrave Catriona, Deegan Timothy, Poulsen Michael, Harden Fiona, and Mengersen Kerrie
Computational Statistics and Data Analysis (2015)

6 companion websites

A New Nonlinear Wavelet-Based Unit Root Test with Structural Breaks

Munich Personal RePEc Archive (2019)

Aydin Mucahit

Are shocks to disaggregated renewable energy consumption permanent or temporary for the USA? Wavelet based unit root test with smooth structural shifts

Energy (2020)

Aydin Mucahit

Statistical inference of locally stationary functional coefficient models

Journal of Statistical Planning and Inference (2020)

Ding Hao, Hu Jianhua, Liu Liangyuan, and Feng Jingyan

A Simple Test for Zero Multiple CorrelationCoefficient in High-Dimensional Normal Data UsingRandom Projection

Working Paper (2019)

Najarzadeh Dariush

Specification tests in semiparametric transformation models - a multiplier bootstrap approach

Working Paper (2019)

Kloodt Nick and Neumeyer Natalie

Quantifying the risk of heat waves using extreme value theory and spatio-temporal functional data

Computational Statistics and Data Analysis (2018)

French Joshua

Simultaneous estimation of quantile regression functions using B-splines and total variation penalty

Computational Statistics and Data Analysis (2017)

Jhong Jae-Hwan and Koo Ja-Yong

Mode jumping MCMC for Bayesian variable selection in GLMM

Computational Statistics and Data Analysis (2018)

Storvik Geir and Hubin Aliaksandr

Multinomial VaR Backtests: A simple implicit approach to backtesting expected shortfall

Journal of Banking and Finance (2018)

Kratz Marie, Lok Yen H., and McNeil Alexander J.

Abstract

In images with low contrast-to-noise ratio (CNR), the information gain from the observed pixel values can be insufficient to distinguish foreground objects. A Bayesian approach to this problem is to incorporate prior information about the objects into a statistical model. This paper introduces a method for representing spatial prior information as an external field in a hidden Potts model of the image lattice. The prior distribution of the latent pixel labels is a mixture of Gaussian fields, centred on the positions of the objects at a previous point in time. This model is particularly applicable in longitudinal imaging studies, where the manual segmentation of one image can be used as a prior for automatic segmentation of subsequent images. The model is demonstrated by application to cone-beam computed tomography (CT), an imaging modality that exhibits distortions in pixel values due to X-ray scatter. The external field prior results in a substantial improvement in segmentation accuracy, reducing the mean pixel misclassification rate on our test images from 87% to 6%.

See paper

Suggested citation

Moores M., Hargrave C., Deegan T., Poulsen M., Harden F., and Mengersen K. (2015) An external field prior for the hidden Potts model, with application to cone-beam computed tomography. Computational Statistics and Data Analysis, 86, 27-41.

See paper