Information Technology Reference
In-Depth Information
Chapter 5
A Joint Bayesian Framework for MR Brain Scan
Tissue and Structure Segmentation Based on
Distributed Markovian Agents
Benoit Scherrer
1
,
3
,
4
,FlorenceForbes
2
, Catherine Garbay
3
,andMichelDojat
1
,
4
1
INSERM U836, 38706 La Tronche, France
{
benoit.scherrer,michel.dojat
}
@ujf-grenoble.fr
2
INRIA, Laboratoire Jean Kuntzman, MISTIS Team, 38041 Montbonnot, France
florence.forbes@inrialpes.fr
3
Laboratoire d'Informatique de Grenoble, 38041 France
catherine.garbay@imag.fr
4
Universite Joseph Fourier, Institut des Neurosciences Grenoble,
38706 La Tronche, France
Abstract.
In most approaches, tissue and subcortical structure segmen-
tations of MR brain scans are handled globally over the entire brain vol-
ume through two relatively independent sequential steps. We propose a
fully Bayesian joint model that integrates within a multi-agent frame-
work local tissue and structure segmentations and local intensity dis-
tribution modeling. It is based on the specification of three conditional
Markov Random Field (MRF) models. The first two encode coopera-
tions between tissue and structure segmentations and integrate
apriori
anatomical knowledge. The third model specifies a Markovian spatial
prior over the model parameters that enables local estimations while
ensuring their consistency, handling this way nonuniformity of inten-
sity without any bias field modeling. The complete joint model provides
then a sound theoretical framework for carrying out tissue and structure
segmentations by distributing a set of local agents that estimate coop-
eratively local MRF models. The evaluation, using a previously ane-
registered atlas of 17 structures, was performed using both phantoms and
real 3T brain scans. It shows good results and in particular robustness
to nonuniformity and noise with a low computational cost. The innova-
tive coupling of agent-based and Markov-centered designs appears as a
robust, fast and promising approach to MR brain scan segmentation.
Keywords:
Medical Imaging, Multi-Agents, Medical Image Processing.
1
Introduction
Diculties in automatic MR brain scan segmentation arise from various sources.
The nonuniformity of image intensity results in spatial intensity variations within
each tissue, which is a major obstacle to an accurate automatic tissue segmenta-
tion. The automatic segmentation of subcortical structures is a challenging task
Search WWH ::
Custom Search