TY - JOUR
T1 - MRI Volumetric Analysis of Multiple Sclerosis
T2 - Methodology and Validation
AU - Li, Lihong
AU - Li, Xiang
AU - Lu, Hongbing
AU - Huang, Wei
AU - Christodoulou, Christopher
AU - Tudorica, Alina
AU - Krupp, Lauren B.
AU - Liang, Zhengrong
N1 - Funding Information:
Manuscript received December 5, 2002; revised June 20, 2003. This work was supported in part by the National Institutes of Health Grant #CA82402 and the National Multiple Sclerosis Society Grant #RG3042-A-2. L. Li is with the Electrical Engineering and Radiology Departments, State University of New York, Stony Brook, NY 11794 USA (e-mail: lli@mil.sunysb.edu). X. Li, H. Lu, W. Huang, and A. Tudorica are with the Radiology Department, State University of New York, Stony Brook, NY 11794 USA. C. Christodoulou and L. B. Krupp are with the Neurology Department, State University of New York, Stony Brook, NY 11794 USA. Z. Liang is with the Radiology and Computer Science Departments, State University of New York, Stony Brook, NY 11794 USA. Digital Object Identifier 10.1109/TNS.2003.817334
PY - 2003/10
Y1 - 2003/10
N2 - We present an automatic mixture-based algorithm for segmentation of brain tissues (white and gray matters-WM and GM), cerebral spinal fluid (CSF), and brain lesions to quantitatively analyze multiple sclerosis. The method performs intensity-based tissue classification using multispectral magnetic resonance (MR) images based on a stochastic model. With the existence of white Gaussian noise and spatially invariant blurring in acquired MR images, a Karhunen-Loéve (K-L) domain Wiener filter is applied for accurate noise reduction and resolution restoration on blurred and noisy images to minimize the partial volume effect (PVE), which is a major limiting factor for the quantitative analysis. Following that, we utilize a Markov random field Gibbs model to integrate the local spatial information into the well-established expectation-maximization model-fitting algorithm. Each voxel is then classified by a maximum a posterior (MAP) criterion, indicating its probabilities of belonging to each class, i.e., each voxel is labeled as a mixel with different tissue percentages, leading to further minimization of the PVE. The volumes of WM, GM, CSF, and brain lesions are extracted from the mixture-based segmentation and the corresponding brain atrophies are computed. In this study, we have investigated the accuracy and repeatability of the algorithm with inclusion of noise analysis and point spread function for image resolution enhancement. Experimental results on phantom, healthy volunteer, and patient studies are presented.
AB - We present an automatic mixture-based algorithm for segmentation of brain tissues (white and gray matters-WM and GM), cerebral spinal fluid (CSF), and brain lesions to quantitatively analyze multiple sclerosis. The method performs intensity-based tissue classification using multispectral magnetic resonance (MR) images based on a stochastic model. With the existence of white Gaussian noise and spatially invariant blurring in acquired MR images, a Karhunen-Loéve (K-L) domain Wiener filter is applied for accurate noise reduction and resolution restoration on blurred and noisy images to minimize the partial volume effect (PVE), which is a major limiting factor for the quantitative analysis. Following that, we utilize a Markov random field Gibbs model to integrate the local spatial information into the well-established expectation-maximization model-fitting algorithm. Each voxel is then classified by a maximum a posterior (MAP) criterion, indicating its probabilities of belonging to each class, i.e., each voxel is labeled as a mixel with different tissue percentages, leading to further minimization of the PVE. The volumes of WM, GM, CSF, and brain lesions are extracted from the mixture-based segmentation and the corresponding brain atrophies are computed. In this study, we have investigated the accuracy and repeatability of the algorithm with inclusion of noise analysis and point spread function for image resolution enhancement. Experimental results on phantom, healthy volunteer, and patient studies are presented.
KW - MRI
KW - Markov random field
KW - Mixture
KW - Multispectral
KW - Partial volume effect
KW - Segmentation
UR - http://www.scopus.com/inward/record.url?scp=0142094719&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0142094719&partnerID=8YFLogxK
U2 - 10.1109/TNS.2003.817334
DO - 10.1109/TNS.2003.817334
M3 - Article
AN - SCOPUS:0142094719
SN - 0018-9499
VL - 50
SP - 1686
EP - 1692
JO - IEEE Transactions on Nuclear Science
JF - IEEE Transactions on Nuclear Science
IS - 5 II
ER -