Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2015 Feb 19:14:8.
doi: 10.1186/1475-925X-14-8.

Segmentation of MR image using local and global region based geodesic model

Xiuming Li  1   2   3 , Dongsheng Jiang  4   5 , Yonghong Shi  6   7 , Wensheng Li  8   9   10
Affiliations

Segmentation of MR image using local and global region based geodesic model

Xiuming Li et al. Biomed Eng Online. .

Abstract

Background: Segmentation of the magnetic resonance (MR) images is fundamentally important in medical image analysis. Intensity inhomogeneity due to the unknown noise and weak boundary makes it a difficult problem.

Method: The paper presents a novel level set geodesic model which integrates the local and the global intensity information in the signed pressure force (SPF) function to suppress the intensity inhomogeneity and implement the segmentation. First, a new local and global region based SPF function is proposed to extract the local and global image information in order to ensure a flexible initialization of the object contours. Second, the global SPF is adaptively balanced by the weight calculated by using the local image contrast. Third, two-phase level set formulation is extended to a multi-phase formulation to successfully segment brain MR images.

Results: Experimental results on the synthetic images and MR images demonstrate that the proposed method is very robust and efficient. Compared with the related methods, our method is much more computationally efficient and much less sensitive to the initial contour. Furthermore, the validation on 18 T1-weighted brain MR images (International Brain Segmentation Repository) shows that our method can produce very promising results.

Conclusions: A novel segmentation model by incorporating the local and global information into the original GAC model is proposed. The proposed model is suitable for the segmentation of the inhomogeneous MR images and allows flexible initialization.

PubMed Disclaimer

Figures

Figure 1
Figure 1
Illustration of the experimental results on the synthetic image with intensity inhomogeneity. (a) Original image and initial contour. (b) Result of CV model. (c) Result of GCV model (d) result of LBF model.
Figure 2
Figure 2
Graphical representation of one level set model. The green line denotes the contour curve, which divides the image into two regions, interior region C 1 and the exterior region C 2 (a). The local neighborhood of x, K σ(y - x) is represented by the black circle. The circle is spilt by the green curve into local interior (red) and local exterior (green) regions. The small yellow and blue dots represent the point x along the contour and point y in the local region of point x, respectively. f 1(x) and f 2(x) are computed in the local interior and local exterior region of the point x to fit the image intensities near the point x (b). f 1(y) and f 2(y) are computed in the local interior and local exterior region of the point x to fit the image intensities near the point y (c).
Figure 3
Figure 3
Graphical representation of two level set model. (a) Blue and red line are considered two zero level set function ϕ 1 and ϕ 2, which divides the image into four regions: C 1, C 2, C 3 and C 4 in (b). (c) The neighborhood of point x, K σ(y - x), is represented by the small black circle. The circle is spilt by the two level set function into local interior and local exterior regions. The small blue dot represents the point y in the local region of x. f 1(y), f 2(y), f 3(y), f 4(y) are computed in the local interior and local exterior region to fit the image intensities near the point y, respectively.
Figure 4
Figure 4
Masks of ω value. (a) Synthetic image; (b) ω value of (a); (c) MR image; (d) ω value of (c). In mask images, red to blue decreases gradually.
Figure 5
Figure 5
Segmenting a hand phantom using the GCV and the proposed model. Illustration of the performance of segmenting a hand phantom (downloaded from [13]) using the GCV and the proposed model: (a) initial contour, (b) segmentation result by the GCV model α = 20 (c) zoomed view of the narrow, green rectangle in (b), (d) segmentation result by our method, and (e) zoomed view of the narrow, green rectangle in (d). The parameter α = 20.
Figure 6
Figure 6
Segmentation results on the real blood vessel image. Yellow rectangle is the initial contours: Row 1 are the results by our model; Row 2 are the results by the LBF model; (a)-(e) are segmentation results in different initial contours. The parameter α = 10.
Figure 7
Figure 7
Comparison of our method with GCV and LBF model in the segmentation of the two MR images in (a1) and (b1), respectively. (a1),(b1) blue rectangles are drawn as initial contours, (a2), (b2) segmentation results from GCV, The parameter α = 10. (a3), (b3) segmentation results from LBF model, and (a4), (b4) segmentation results from our method, respectively. The parameter α = 10.
Figure 8
Figure 8
Application of our method in segmenting the 3T MR images. (a) original images and initial blue contours, (b) (c) (d) (e) segmentation results of ϵ is set to 0.1, 0.3, 0.5, 0.8, respectively. The parameter α = 30.
Figure 9
Figure 9
Comparison our method with Li’s method by using the data from IBSR. Colum 1: original images (row 1: Axial plane, row 2: sagittal plane, row 3: coronal plane); Colum 2: results of Li’s method; Colum 3: results of our method. The parameter α = 100.
Figure 10
Figure 10
Tissue Segmentation masks of Figure 8 . Colum 1: ground truth; Colum 2: Li’s method; Colum 3: our method.
Figure 11
Figure 11
Box plot of the DSC values of WM, GM and CSF by our method and Li’s method, respectively.
See this image and copyright information in PMC

References

    1. Kass M, Witkin A, Terzopoulos D. Snakes: active contour models. Int J Comput Vision. 1988;1(4):321–31. doi: 10.1007/BF00133570. - DOI
    1. Lankton S, Tannenbaum A. Localizing region-based active contours. IEEE Trans Image Process. 2008;17(11):2029–39. doi: 10.1109/TIP.2008.2004611. - DOI - PMC - PubMed
    1. Caselles V, Kimmel R, Sapiro G. Geodesic active contours. Int J Comput Vision. 1997;22(1):61–79. doi: 10.1023/A:1007979827043. - DOI
    1. Goldenberg R, Kimmel R, Rivlin E, Rudzsky M. Fast geodesic active contours. IEEE Trans Image Process. 2001;10(10):1467–75. doi: 10.1109/83.951533. - DOI - PubMed
    1. Zhang K, Zhang L, Song H, Zhou W. Active contours with selective local or global segmentation: a new formulation and level set method. Image Vision Comput. 2010;28(4):668–76. doi: 10.1016/j.imavis.2009.10.009. - DOI

Publication types

MeSH terms

LinkOut - more resources