Limits...
Multiple sclerosis lesion detection using constrained GMM and curve evolution.

Freifeld O, Greenspan H, Goldberger J - Int J Biomed Imaging (2009)

Bottom Line: The proposed CGMM-CE algorithm is used to segment 3D MRI brain images with an arbitrary number of channels.The CGMM-CE algorithm is automated and does not require an atlas for initialization or parameter learning.Experimental results on both standard brain MRI simulation data and real data indicate that the proposed method outperforms previously suggested approaches, especially for highly noisy data.

View Article: PubMed Central - PubMed

Affiliation: Department of Biomedical Engineering, Tel-Aviv University, Tel Aviv 69978, Israel. orenf@eng.tau.ac.il

ABSTRACT
This paper focuses on the detection and segmentation of Multiple Sclerosis (MS) lesions in magnetic resonance (MRI) brain images. To capture the complex tissue spatial layout, a probabilistic model termed Constrained Gaussian Mixture Model (CGMM) is proposed based on a mixture of multiple spatially oriented Gaussians per tissue. The intensity of a tissue is considered a global parameter and is constrained, by a parameter-tying scheme, to be the same value for the entire set of Gaussians that are related to the same tissue. MS lesions are identified as outlier Gaussian components and are grouped to form a new class in addition to the healthy tissue classes. A probability-based curve evolution technique is used to refine the delineation of lesion boundaries. The proposed CGMM-CE algorithm is used to segment 3D MRI brain images with an arbitrary number of channels. The CGMM-CE algorithm is automated and does not require an atlas for initialization or parameter learning. Experimental results on both standard brain MRI simulation data and real data indicate that the proposed method outperforms previously suggested approaches, especially for highly noisy data.

No MeSH data available.


Related in: MedlinePlus

Illustration of the influence of the EM iterations. As the parameter estimation improves with the iterations, the segmentation induced from the model becomes more and more robust to noise. Blue-CSF; Green-GM; Yellow-WM. (a) Initial segmentation. (b) Segmentation after convergence (five EM iterations).
© Copyright Policy - open-access
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2742654&req=5

fig2: Illustration of the influence of the EM iterations. As the parameter estimation improves with the iterations, the segmentation induced from the model becomes more and more robust to noise. Blue-CSF; Green-GM; Yellow-WM. (a) Initial segmentation. (b) Segmentation after convergence (five EM iterations).

Mentions: Once the optimal set of parameters is obtained using the EM algorithm, we can compute soft tissue segmentation maps. Moving from soft segmentation to hard segmentation is straight forward, using the Maximum-A-Posteriori (MAP) criterion. Tissue segmentation is achieved by the affiliation of each voxel to the tissue that maximizes the a posteriori probability: (9)tissue-label(voxelt)=arg max j  p(tissue=j ∣ xt,I(xt))=arg max  j ∑i ∣ π(i)=jαifi(xt,I(xt) ∣ μi,Σi).Figure 2 illustrates the segmentation induced from the CGMM model and shows how the EM iterations improve the segmentation quality. Since a lot of the algorithmic effort is spent on finding a good initialization (see Section 2.2), the EM needs only few iterations to converge.


Multiple sclerosis lesion detection using constrained GMM and curve evolution.

Freifeld O, Greenspan H, Goldberger J - Int J Biomed Imaging (2009)

Illustration of the influence of the EM iterations. As the parameter estimation improves with the iterations, the segmentation induced from the model becomes more and more robust to noise. Blue-CSF; Green-GM; Yellow-WM. (a) Initial segmentation. (b) Segmentation after convergence (five EM iterations).
© Copyright Policy - open-access
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2742654&req=5

fig2: Illustration of the influence of the EM iterations. As the parameter estimation improves with the iterations, the segmentation induced from the model becomes more and more robust to noise. Blue-CSF; Green-GM; Yellow-WM. (a) Initial segmentation. (b) Segmentation after convergence (five EM iterations).
Mentions: Once the optimal set of parameters is obtained using the EM algorithm, we can compute soft tissue segmentation maps. Moving from soft segmentation to hard segmentation is straight forward, using the Maximum-A-Posteriori (MAP) criterion. Tissue segmentation is achieved by the affiliation of each voxel to the tissue that maximizes the a posteriori probability: (9)tissue-label(voxelt)=arg max j  p(tissue=j ∣ xt,I(xt))=arg max  j ∑i ∣ π(i)=jαifi(xt,I(xt) ∣ μi,Σi).Figure 2 illustrates the segmentation induced from the CGMM model and shows how the EM iterations improve the segmentation quality. Since a lot of the algorithmic effort is spent on finding a good initialization (see Section 2.2), the EM needs only few iterations to converge.

Bottom Line: The proposed CGMM-CE algorithm is used to segment 3D MRI brain images with an arbitrary number of channels.The CGMM-CE algorithm is automated and does not require an atlas for initialization or parameter learning.Experimental results on both standard brain MRI simulation data and real data indicate that the proposed method outperforms previously suggested approaches, especially for highly noisy data.

View Article: PubMed Central - PubMed

Affiliation: Department of Biomedical Engineering, Tel-Aviv University, Tel Aviv 69978, Israel. orenf@eng.tau.ac.il

ABSTRACT
This paper focuses on the detection and segmentation of Multiple Sclerosis (MS) lesions in magnetic resonance (MRI) brain images. To capture the complex tissue spatial layout, a probabilistic model termed Constrained Gaussian Mixture Model (CGMM) is proposed based on a mixture of multiple spatially oriented Gaussians per tissue. The intensity of a tissue is considered a global parameter and is constrained, by a parameter-tying scheme, to be the same value for the entire set of Gaussians that are related to the same tissue. MS lesions are identified as outlier Gaussian components and are grouped to form a new class in addition to the healthy tissue classes. A probability-based curve evolution technique is used to refine the delineation of lesion boundaries. The proposed CGMM-CE algorithm is used to segment 3D MRI brain images with an arbitrary number of channels. The CGMM-CE algorithm is automated and does not require an atlas for initialization or parameter learning. Experimental results on both standard brain MRI simulation data and real data indicate that the proposed method outperforms previously suggested approaches, especially for highly noisy data.

No MeSH data available.


Related in: MedlinePlus