Method for segmentation of lesions
View Patent ↗A method of segmenting a lesion ( 910 ) from normal anatomy in a 3-dimensional image comprising the steps of: receiving an initial set of voxels ( 520 ) that are contained within the lesion to be segmented; growing a region which includes the lesion from the initial set of voxels; identifying a second set of voxels ( 530 ) on a surface of the normal anatomy; determining a surface containing the second set of voxels which demarks a boundary ( 540 ) between the lesion and the normal anatomy; and classifying voxels which are part of the lesion.
1. A method of segmenting a lesion from normal anatomy in a 3-dimensional image comprising the steps of:
a) receiving an initial set of voxels that are contained within the lesion to be segmented;
b) growing an initial region which includes the lesion from the initial set of voxels;
c) identifying within the initial region a second set of voxels on a surface of the normal anatomy;
d) determining a surface containing the second set of voxels which demarks a boundary between the lesion and the normal anatomy within the initial region; and
e) classifying voxels which are part of the lesion.
2. A method as in claim 1 wherein said step of growing a region comprises a segmentation front analysis.
3. A method as in claim 2 wherein segmentation front analysis comprises a fast marching method.
4. A method as in claim 2 wherein said step of identifying a second set of voxels is derived from the segmentation front analysis.
5. A method as in claim 2 wherein the demarcation boundary is determined by a scattered data surface reconstruction method.
6. A method as in claim 5 wherein the scattered data surface reconstruction is by a low degree polynomial surface.
7. A method as in claim 5 wherein the scattered data surface reconstruction is achieved by interpolating radial basis function.
8. A method as in claim 7 wherein the interpolating function is a parametric function.
9. A method as in claim 7 wherein the interpolating function is an implicit function.
10. A method as in claim 5 wherein the scattered data surface reconstruction method utilizes a partition of unity.
11. A method as in claim 5 wherein the demarcation boundary is determined using a moving least squares method.
12. A method as in claim 5 wherein the demarcation boundary is a minimal surface.
13. A method as in claim 1 wherein the demarcation boundary is determined by a scattered data surface reconstruction method.
14. A method as in claim 13 wherein the scattered data surface reconstruction is achieved by interpolating radial basis function.
15. A method as in claim 14 wherein the interpolating function is a parametric function.
16. A method as in claim 14 wherein the interpolating function is an implicit function.
17. A method as in claim 13 wherein the scattered data surface reconstruction method utilizes a partition of unity.
18. A method as in claim 13 wherein the demarcation boundary is determined using a moving least squares method or wherein the demarcation boundary is a minimal surface.
19. A method as in claim 5 wherein the scattered data surface reconstruction method is determined by the segmentation front analysis.