Name
Abstract | ||
|---|---|---|
In this paper, we present a new Markovian scheme for MRI segmentation using a priori knowledge obtained from prob- ability maps. Indeed we propose to use both triplet Markov chain and a brain atlas containing prior expectations about the spatial localization of the different tissue classes, to seg- ment the brain in gray matter, white matter and cerebro-spinal fluid in an unsupervised way. Experimental results on real data are included to validate this approach. Comparison with other previously used techniques demonstrates the advan- tages (robustness, low computational complexity) of this new Markovian segmentation scheme using a probabilistic atlas. Magnetic Resonance Imaging (MRI) provides three-dimen- sional description of the human brain at a high resolution which serves as a reference for clinical investigations as well as for functional studies. Quantitative analysis of anatomical structures in D images has recently become of great inter- est for the study of different pathologies (multiple sclerosis, schizophrenia). Moreover D visualization of brain struc- tures, with surface rendering methods at interactive speeds, is highly desirable for routine use in clinical settings and surgical planning. Therefore, automatic segmentation of D brain images is becoming an increasingly important process- ing step for diagnosis, treatment and surgery planning. Because of issues such as partial volume effect, noise or acquisition artifacts, segmenting cerebral structures remains a challenging task that will be hardly accomplished by meth- ods that rely only on information present in the image. For these reasons, segmentation methods, that make use of some models (a priori information) of the targeted cerebral struc- tures, were proposed (1). In (2, 3), Asbhurner and Friston proposed a modified maximum likelihood mixture model algorithm using proba- bilistic images for the segmentation. The probability images they used, are the means of binary images of white matter (WM), gray matter (GM), and cerebrospinal fluid (CSF) ob- tained from the segmentation of the original images. How- ever this method does not take into account neighborhood information, and therefore the quality of the segmentation is degraded if noise is present in the images. Other methods |
Year | DOI | Venue |
|---|---|---|
2006 | 10.1109/ISBI.2006.1624934 | Arlington, VA |
Keywords | Field | DocType |
Markov processes,biological tissues,biomedical MRI,brain,image segmentation,medical image processing,3D MRI brain segmentation,brain tissue,cerebro-spinal fluid,gray matter,low computational complexity,probabilistic atlas,robustness,spatial localization,triplet Markov chain,white matter | Brain segmentation,Computer vision,Brain atlas,Markov process,Pattern recognition,Segmentation,Computer science,Markov chain,Robustness (computer science),Image segmentation,Artificial intelligence,Computational complexity theory | Conference |
ISSN | ISBN | Citations |
1945-7928 | 0-7803-9576-X | 11 |
PageRank | References | Authors |
0.71 | 8 | 3 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Stephanie Bricq | 1 | 51 | 3.82 |
| Christophe Collet | 2 | 246 | 35.46 |
| Jean-paul Armspach | 3 | 81 | 5.20 |