Segmenting colon wall via level set techniques

• US 20080304616A1
• Filed: 06/05/2007
• Published: 12/11/2008
• Est. Priority Date: 06/05/2007
• Status: Active Grant

• Current Assignee: United States Department Of Health And Human Services (Government of the United States of America)
• Original Assignee: The United States of America As Represented By The Secretary of Agriculture
• Inventors: Summers, Ronald M., Bitter, Ingmar, Van Uitert, Robert L. JR.
• Granted Patent: US 8,175,348 B2
• Time in Patent Office: Days
• US Class Current: 378/4
• CPC Class Codes: A61B 6/03 Computed tomography [CT] echo-tomography A61B8/14
• CPC Class Codes: G06T 2207/10081 Computed x-ray tomography [CT]
• CPC Class Codes: G06T 2207/20161 Level set
• CPC Class Codes: G06T 2207/30028 Colon; Small intestine
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06T 7/12 Edge-based segmentation
• CPC Class Codes: G06T 7/149 involving deformable models, e.g. active contour models

Abstract

Various level set techniques can be used to automatically segment the colon wall, including identifying the colon wall outer boundary. A speed image can be used during level set processing. For example, the speed image can be generated via inverting the gradient perpendicular to the segmented inner boundary of the colon wall. The techniques can be useful for determining wall thickness, which can be used to classify polyp candidates, diagnose diseases of the colon, and the like.

Claims

1. A computer-implemented method comprising:

receiving a digital representation for a colon, wherein the digital representation represents at least a portion of a colon wall for the colon;

identifying an outer boundary for the colon wall via a level set technique; and

outputting an indication of the outer boundary of the colon wall.

2. One or more computer-readable media comprising computer-executable instructions causing a computer to perform the method of claim 1.

3. The computer-implemented method of claim 1 wherein the indication of the outer boundary of the colon wall is of subvoxel accuracy.

4. The computer-implemented method of claim 1 wherein the level set technique comprises three-dimensional geodesic active contour level set segmentation.

5. The computer-implemented method of claim 1 wherein:

identifying the outer boundary for the colon wall via the level set technique comprises;

using a lumen segmentation of the digital representation as an initial level set boundary.

6. The computer-implemented method of claim 1 further comprising:

identifying an inner boundary of the colon wall; and

via the inner boundary of the colon wall and the outer boundary of the colon wall, calculating a thickness of the colon wall.

7. The computer-implemented method of claim 6 wherein:

the outer boundary of the colon wall is represented as a surface;

the inner boundary of the colon wall is represented as a surface; and

calculating thickness comprises determining distance between the surfaces.

8. The computer-implemented method of claim 6 further comprising:

submitting a set of characteristics for a polyp candidate to a polyp candidate classifier configured to determine whether the polyp candidate is a true positive;

wherein the set of characteristics comprises the thickness of the colon wall.

9. The computer-implemented method of claim 1 further comprising:

segmenting the colon wall.

10. The computer-implemented method of claim 9 further comprising:

determining a thickness of the colon wall.

11. The computer-implemented method of claim 9 further comprising:

detecting colonic diverticular disease via the thickness of the colon wall.

12. The computer-implemented method of claim 9 further comprising:

detecting colon spasm via the thickness of the colon wall.

13. The computer-implemented method of claim 9 further comprising:

detecting colon cancer via the thickness of the colon wall.

14. The computer-implemented method of claim 9 further comprising:

detecting presence of a polyp via the thickness of the colon wall.

15. The computer-implemented method of claim 1 wherein:

identifying the outer boundary for the colon wall comprises identifying at least a location of the outer boundary of the colon wall; and

the indication of the outer boundary of the colon wall indicates at least the location of the outer boundary of the colon wall.

16. The computer-implemented method of claim 1 wherein the level set technique comprises:

generating a speed image; and

evolving an isosurface based at least on the speed image.

17. The computer-implemented method of claim 16 wherein generating the speed image comprises:

calculating a directional derivative of the digital representation in a direction perpendicular to a colon wall inner boundary represented in a colon wall inner boundary segmentation.

18. The computer-implemented method of claim 17 further comprising:

suppressing local non-maximum gradients along a level set expansion direction via a sigmoid filter.

19. The computer-implemented method of claim 17 further comprising:

performing a lumen segmentation for the digital representation; and

using a lumen boundary in the lumen segmentation as the colon wall inner boundary segmentation.

20. The computer-implemented method of claim 19 wherein the lumen segmentation comprises:

generating a threshold region growing segmentation; and

segmenting via a threshold level set technique, wherein the threshold level set technique uses the threshold region growing segmentation as an initial level set boundary.

21. The computer-implemented method of claim 16 wherein:

the speed image is determined via the digital representation for the colon.

22. The computer-implemented method of claim 1 further comprising:

segmenting an entire colon wall of the colon.

23. An apparatus comprising:

means for receiving a digital representation for a colon, wherein the digital representation represents at least a portion of a colon wall for the colon;

means for identifying an outer boundary for the colon wall via level set processing; and

means for outputting an indication of the outer boundary of the colon wall.

24. A computer-implemented method comprising:

receiving a digital representation for a colon, wherein the digital representation represents at least a portion of a colon wall for the colon;

segmenting the colon wall via a level set technique, whereby the segmenting results in a segmented colon wall for the digital representation; and

via the segmented colon wall for the digital representation, automatically calculating colon wall thickness for at least a portion of the colon wall.

25. A computer-implemented method comprising:

receiving a three-dimensional digital representation for a colon, wherein the digital representation comprises a computed tomography image representing at least a portion of a colon wall for the colon;

from the digital representation for the colon, generating a lumen segmentation indicating a boundary of an inner wall of the colon;

producing lumen segmentation level sets from the lumen segmentation;

from the lumen segmentation and the computed tomography image, generating a speed image via a three-dimensional derivative of the computed tomography image in a direction perpendicular to lumen segmentation level sets, wherein local non-maximum gradients along level set expansion direction are suppressed, and generating the speed image comprises applying a sigmoid filter emphasizing high directional derivatives and inverting the speed image;

generating a level set image via a level set segmentation of an outer wall of the colon via three-dimensional geodesic active contour level set segmentation with the speed image, wherein the lumen segmentation level sets are used as an initial level set boundary, an advection term attracts level set evolution to high gradient values, and a curvature term prevents evolution of the boundary from exceeding a maximum curvature;

determining a boundary of the outer wall of the colon via an isocontour in the level set image; and

outputting an indication of the boundary of the outer wall of the colon.

26. A computer-implemented method comprising:

receiving a digital representation for an anatomical structure, wherein the digital representation represents at least a portion of a wall for the anatomical structure;

identifying an outer boundary for the wall via a level set technique; and

outputting an indication of the outer boundary of the wall.

27. The computer-implemented method of claim 26 further comprising:

determining a thickness of the wall.

28. The computer-implemented method of claim 27 further comprising:

detecting atherosclerosis via the thickness of the wall.

29. The computer-implemented method of claim 27 further comprising:

detecting hyperplasia via the thickness of the wall.

30. A computer-implemented method comprising:

receiving a digital representation for a colon, wherein the digital representation represents at least a portion of a wall for the colon;

determining wall thickness for the wall; and

detecting presence of colonic diverticular disease via determined wall thickness for the wall.

31. The method of claim 30 wherein determining wall thickness for the wall comprises applying a level set technique to the digital representation.

32. The method of claim 30 wherein determining wall thickness for the wall comprises applying a binary space partitioning tree.

33. The method of claim 30 wherein detecting presence of colonic diverticular disease comprises:

clustering candidate detections within a threshold distance.

34. The method of claim 30 wherein detecting presence of colonic diverticular disease comprises:

responsive to determining that a location on the colon is under a threshold thickness, classifying the location as not having diverticular disease.

35. The method of claim 30 wherein detecting presence of colonic diverticular disease comprises:

responsive to determining that CT intensity of a location on an outer wall of the colon is outside of a normal range for colon wall, classifying the location as not being a detection of diverticular disease.

36. The method of claim 30 wherein detecting presence of colonic diverticular disease comprises:

computing an average and standard deviation of colon wall thickness for a cluster of locations.

37. The method of claim 30 wherein detecting presence of colonic diverticular disease comprises:

computing an average and standard deviation of intensity values for a cluster of locations.

38. The method of claim 30 wherein detecting presence of colonic diverticular disease comprises:

providing a plurality of features for a cluster of locations to a support-vector machine classifier configured to indicate whether the cluster of locations are indicative of colonic diverticular disease.

39. One or more computer-readable media comprising computer-executable instructions causing a computer to perform the method of claim 30.