|
Activity Number:
|
308
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, August 5, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #300777 |
|
Title:
|
A Bayesian Image Analysis of Change in Tumor/Brain Contrast Uptake Induced by Radiation
|
|
Author(s):
|
Xiaoxi Zhang*+ and Timothy D. Johnson and Roderick J. Little and Yue Cao
|
|
Companies:
|
Pfizer, Inc. and The University of Michigan and The University of Michigan and The University of Michigan
|
|
Address:
|
235 E 42nd Street, New York, NY, 10017,
|
|
Keywords:
|
hidden Markov random fields ; Quantitative MRI ; reversible jump MCMC ; Swendsen-Wang algorithm ; trans-dimensional proposal
|
|
Abstract:
|
This work is motivated by a quantitative Magnetic Resonance Imaging study of the change in tumor/brain contrast uptake induced by radiation. The results show a transient period at which contrast uptake in the tumor is maximal; suggesting an optimal time to initiate chemotherapy during the course of radiotherapy. A notable feature of the data is spatial heterogeneity. We introduce a latent layer of discrete labels and employ a Gaussian hidden Markov random field (MRF) model that respects this feature. Conditional on the hidden labels, the observed data are assumed independent and normally distributed. We estimate the MRF regularization parameter, treat the number of MRF states as a parameter and estimate it via a reversible jump Markov chain Monte Carlo algorithm with a novel and nontrivial implementation. We examine the performance of our method in a simulation study and on real data.
|
