Magnetic resonance image tissue classification using a partial volume model

被引:795
作者
Shattuck, DW [1 ]
Sandor-Leahy, SR
Schaper, KA
Rottenberg, DA
Leahy, RM
机构
[1] Univ So Calif, Signal & Image Proc Inst, Los Angeles, CA 90089 USA
[2] TRW Co Inc, Redondo Beach, CA 90278 USA
[3] VA Med Ctr, PET Imaging Ctr, Minneapolis, MN 55417 USA
[4] Univ Minnesota, Dept Neurol, Minneapolis, MN 55455 USA
[5] Univ Minnesota, Dept Radiol, Minneapolis, MN 55455 USA
关键词
D O I
10.1006/nimg.2000.0730
中图分类号
Q189 [神经科学];
学科分类号
071006 ;
摘要
We describe a sequence of low-level operations to isolate and classify brain tissue within T1-weighted magnetic resonance images (MRI). Our method first removes nonbrain tissue using a combination of anisotropic diffusion filtering, edge detection, and mathematical morphology. We compensate for image nonuniformities due to magnetic field inhomogeneities by fitting a tricubic B-spline gain field to local estimates of the image nonuniformity spaced throughout the MRI volume. The local estimates are computed by fitting a partial volume tissue measurement model to histograms of neighborhoods about each estimate point. The measurement model uses mean tissue intensity and noise variance values computed from the global image and a multiplicative bias parameter that is estimated for each region during the histogram fit. Voxels in the intensity-normalized image are then classified into six tissue types using a maximum a posteriori classifier. This classifier combines the partial volume tissue measurement model with a Gibbs prior that models the spatial properties of the brain. We validate each stage of our algorithm on real and phantom data. Using data from the 20 normal MRI brain data sets of the Internet Brain Segmentation Repository, our method achieved average kappa indices of kappa = 0.746 +/- 0.114 for gray matter (GM) and kappa = 0.798 +/- 0.089 for white matter (WM) compared to expert labeled data. Our method achieved average kappa indices kappa = 0.893 +/- 0.041 for GM and kappa = 0.928 +/- 0.039 for WM compared to the ground truth labeling on 12 volumes from the Montreal Neurological Institute's BrainWeb phantom. (C) 2001 Academic Press.
引用
收藏
页码:856 / 876
页数:21
相关论文
共 60 条
[21]   Local histogram correction of MRI spatially dependent image pixel intensity nonuniformity [J].
DeCarli, C ;
Murphy, DGM ;
Teichberg, D ;
Campbell, G ;
Sobering, GS .
JMRI-JOURNAL OF MAGNETIC RESONANCE IMAGING, 1996, 6 (03) :519-528
[22]   USING CANNY CRITERIA TO DERIVE A RECURSIVELY IMPLEMENTED OPTIMAL EDGE DETECTOR [J].
DERICHE, R .
INTERNATIONAL JOURNAL OF COMPUTER VISION, 1987, 1 (02) :167-187
[23]   MEASURES OF THE AMOUNT OF ECOLOGIC ASSOCIATION BETWEEN SPECIES [J].
DICE, LR .
ECOLOGY, 1945, 26 (03) :297-302
[24]   NONLINEAR ANISOTROPIC FILTERING OF MRI DATA [J].
GERIG, G ;
KUBLER, O ;
KIKINIS, R ;
JOLESZ, FA .
IEEE TRANSACTIONS ON MEDICAL IMAGING, 1992, 11 (02) :221-232
[25]   Estimating the bias field of MR images [J].
Guillemaud, R ;
Brady, M .
IEEE TRANSACTIONS ON MEDICAL IMAGING, 1997, 16 (03) :238-251
[26]   White matter changes with normal aging [J].
Guttmann, CRG ;
Jolesz, FA ;
Kikinis, R ;
Killiany, RJ ;
Moss, MB ;
Sandor, T ;
Albert, MS .
NEUROLOGY, 1998, 50 (04) :972-978
[27]  
Hart P.E., 1973, Pattern recognition and scene analysis
[28]   THE RELATIONSHIP OF QUANTITATIVE BRAIN MAGNETIC-RESONANCE-IMAGING MEASURES TO NEUROPATHOLOGIC INDEXES OF HUMAN-IMMUNODEFICIENCY-VIRUS INFECTION [J].
HEINDEL, WC ;
JERNIGAN, TL ;
ARCHIBALD, SL ;
ACHIM, CL ;
MASLIAH, E ;
WILEY, CA .
ARCHIVES OF NEUROLOGY, 1994, 51 (11) :1129-1135
[29]   Regional glucose metabolic abnormalities are not the result of atrophy in Alzheimer's disease [J].
Ibáñez, V ;
Pietrini, P ;
Alexander, GE ;
Furey, ML ;
Teichberg, D ;
Rajapakse, JC ;
Rapoport, SI ;
Schapiro, MB ;
Horwitz, B .
NEUROLOGY, 1998, 50 (06) :1585-1593
[30]  
John F., 1982, Partial differential equations