Biomedical Engineering Reference
In-Depth Information
role in the shape-based method. Therefore, processing speed is a big drawback of
this method. Besides, for image segmentation, the initial position of the average
shape is of great importance to the final segmentation. For complicated medical
images, if the surrounding context around the object to be segmented has occupied
a very large portion of the whole image, and similar structures are present in
the images, the gradient descent method is not able to move the initial average
shape to the place where the object is located, because the energy minimization
method scheme makes the segmentation become stuck in a non-object position
due to the local minimum. So, preprocessing has to be performed to restrict
the segmentation to a smaller region. Or, statistic parameters can be assigned to
the pose factors. For example, the largest size of the object (scaling), the largest
tilting of the object (rotation), etc. However, this extra preprocessing will make the
systemmore complex, with an unwieldy number of new parameters for controlling
segmentation. Therefore, there is yet much work to be done to make the shape-
based segmentation method more efficient in level set frameworks.
7.
ACKNOWLEDGMENTS
The research presented in this chapter was supported, in part, by the follow-
ing NIH grants: R01-HL63373, R01-HL071809, R01-HL075446, R33-CA94801,
and R01-HL64368.
8.
REFERENCES
1. Suri JS, Liu K, Singh S, Laxminarayan S, Zeng X, Reden L. 2002. Shape recovery algorithms
using level sets in 2-d/3-d medical imagery: a state-of-the-art review.
IEEE Trans Inform
Technol Biomed
6
:8-28.
2. Angenent S, Chopp D, Ilmanen T. 1995. On the singularities of cones evolving by mean cur-
vature.
Commun Partial Differ Eq
20
:1937-1958.
3. Chopp DL. 1993. Computing minimal surfaces via level set curvature flow.
J Comput Phys
106
:77-91.
4. Sethian JA. 1990. Numerical algorithms for propagating interfaces: Hamilton-Jacobi equations
and conservation laws.
J Differ Geom
31
:131-161.
5. Sethian JA, Malladi R. 1996. An
O
(
N
log
N
)
algorithm for shape modeling.
Proc Natl Acad
Sci USA
93
:9389-9392.
6. Sethian JA. 1995. Algorithms for tracking interfaces in CFD and material science.
Annu Rev
Comput Fluid Mech
12
:125-138.
7. Sussman M, Smereka P, Osher SJ. 1994. A level set method for computing solutions to incom-
pressible two-phase flow.
J Comput Phys
114
:146-159.
8. Rhee C, Talbot L, Sethian JA. 1995. Dynamical study of a premixed V-flame.
J Fluid Mech
300
:87-115.
9. Sethian JA, Strain JD. 1992. Crystal growth and dentritic solidification.
J Comput Phys
98
:231-
253.
10. Adalsteinsson D, Sethian JA. 1995. A unified level set approach to etching, deposition and
lithography, I: algorithms and two-dimensional simulations.
J Comput Phys
120
:128-144.
