Limits...
Multigrid Nonlocal Gaussian Mixture Model for Segmentation of Brain Tissues in Magnetic Resonance Images

View Article: PubMed Central - PubMed

ABSTRACT

We propose a novel segmentation method based on regional and nonlocal information to overcome the impact of image intensity inhomogeneities and noise in human brain magnetic resonance images. With the consideration of the spatial distribution of different tissues in brain images, our method does not need preestimation or precorrection procedures for intensity inhomogeneities and noise. A nonlocal information based Gaussian mixture model (NGMM) is proposed to reduce the effect of noise. To reduce the effect of intensity inhomogeneity, the multigrid nonlocal Gaussian mixture model (MNGMM) is proposed to segment brain MR images in each nonoverlapping multigrid generated by using a new multigrid generation method. Therefore the proposed model can simultaneously overcome the impact of noise and intensity inhomogeneity and automatically classify 2D and 3D MR data into tissues of white matter, gray matter, and cerebral spinal fluid. To maintain the statistical reliability and spatial continuity of the segmentation, a fusion strategy is adopted to integrate the clustering results from different grid. The experiments on synthetic and clinical brain MR images demonstrate the superior performance of the proposed model comparing with several state-of-the-art algorithms.

No MeSH data available.


Segmentation results for a clinical 3 T MR image: (a) original image; (b) segmentation result of GMM; (c) segmentation result of Wells; (d) segmentation result of MNGMM.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC5015012&req=5

fig3: Segmentation results for a clinical 3 T MR image: (a) original image; (b) segmentation result of GMM; (c) segmentation result of Wells; (d) segmentation result of MNGMM.

Mentions: In order to show the robustness of our method, we compared our method with GMM and Wells method on a clinical 3 T MR image, which has severe intensity inhomogeneity and noise. The acquisition parameters of the real data are as follows: slice thickness 1 mm, echo delay time (TE) 35 msec, repetition time (TR) 450 msec, and flip angle 90 deg. The size of the data is 256 × 256 × 170. From the image, we can find that most intensities of putamen are higher than those of the cortex, which also belong to the GM. GMM misclassifies putamen into WM since some intensities of putamen are closer to WM than to GM, which can be shown in Figure 3(b). The result of the Wells method [6] is shown in Figure 3(c). It can be seen from the result that the method can reduce the effect of the bias field; however, with the effect of the inhomogeneities in inner tissues, some pixels of putamen have been misclassified into WM and the method is sensitive to the noise. In contrast, MNGMM can yield satisfactory results, which are more compatible with human visual perception.


Multigrid Nonlocal Gaussian Mixture Model for Segmentation of Brain Tissues in Magnetic Resonance Images
Segmentation results for a clinical 3 T MR image: (a) original image; (b) segmentation result of GMM; (c) segmentation result of Wells; (d) segmentation result of MNGMM.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC5015012&req=5

fig3: Segmentation results for a clinical 3 T MR image: (a) original image; (b) segmentation result of GMM; (c) segmentation result of Wells; (d) segmentation result of MNGMM.
Mentions: In order to show the robustness of our method, we compared our method with GMM and Wells method on a clinical 3 T MR image, which has severe intensity inhomogeneity and noise. The acquisition parameters of the real data are as follows: slice thickness 1 mm, echo delay time (TE) 35 msec, repetition time (TR) 450 msec, and flip angle 90 deg. The size of the data is 256 × 256 × 170. From the image, we can find that most intensities of putamen are higher than those of the cortex, which also belong to the GM. GMM misclassifies putamen into WM since some intensities of putamen are closer to WM than to GM, which can be shown in Figure 3(b). The result of the Wells method [6] is shown in Figure 3(c). It can be seen from the result that the method can reduce the effect of the bias field; however, with the effect of the inhomogeneities in inner tissues, some pixels of putamen have been misclassified into WM and the method is sensitive to the noise. In contrast, MNGMM can yield satisfactory results, which are more compatible with human visual perception.

View Article: PubMed Central - PubMed

ABSTRACT

We propose a novel segmentation method based on regional and nonlocal information to overcome the impact of image intensity inhomogeneities and noise in human brain magnetic resonance images. With the consideration of the spatial distribution of different tissues in brain images, our method does not need preestimation or precorrection procedures for intensity inhomogeneities and noise. A nonlocal information based Gaussian mixture model (NGMM) is proposed to reduce the effect of noise. To reduce the effect of intensity inhomogeneity, the multigrid nonlocal Gaussian mixture model (MNGMM) is proposed to segment brain MR images in each nonoverlapping multigrid generated by using a new multigrid generation method. Therefore the proposed model can simultaneously overcome the impact of noise and intensity inhomogeneity and automatically classify 2D and 3D MR data into tissues of white matter, gray matter, and cerebral spinal fluid. To maintain the statistical reliability and spatial continuity of the segmentation, a fusion strategy is adopted to integrate the clustering results from different grid. The experiments on synthetic and clinical brain MR images demonstrate the superior performance of the proposed model comparing with several state-of-the-art algorithms.

No MeSH data available.