Hybrid segmentation of anatomical structure

• US 20080008367A1
• Filed: 07/06/2006
• Published: 01/10/2008
• Est. Priority Date: 07/06/2006
• Status: Active Grant

• Current Assignee: United States Department Of Health And Human Services
• Original Assignee: The United States of America As Represented By The Secretary of Agriculture
• Inventors: Summers, Ronald M., Franaszek, Marek
• Granted Patent: US 8,023,703 B2
• Time in Patent Office: Days
• US Class Current: 382/128
• CPC Class Codes: G06V 10/26 Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
• CPC Class Codes: G06V 2201/032 of protuberances, polyps nodules, etc.

Abstract

An image of an anatomical structure can be analyzed to determine an enclosing three-dimensional boundary when the anatomical structure is filled with two substances, such as air and a fluid. Various techniques can be used to determine the enclosing boundary including: analyzing the virtual structure to segment the structure into air and fluid pockets, determining if there are multiple fluid pockets whose surface touches a single air-fluid boundary, determining a separate threshold for respective fluid pockets, resegmenting the virtual anatomical structure using the separate threshold for different fluid pockets, forming a hierarchical pocket tree which represents the relationship between the fluid and air pockets, pruning the pocket tree based on various criteria which corresponds to deleting those pruned portions from the virtual anatomical structure, and resegmenting the remaining virtual anatomical structure using one or more of fuzzy connectedness, two-dimensional gap filling, and level set segmentation.

Claims

1. One or more computer readable media comprising computer-readable instructions for performing a method comprising:

receiving a digital representation of an anatomical structure;

segmenting the digital representation into at least trial air pockets and a plurality of trial fluid pockets, producing a trial representation; and

for at least a subset of the trial fluid pockets, determining at least one per-fluid-pocket tailored threshold for each trial fluid pocket in the subset.

2. The computer-readable media of claim 1 further comprising:

using the at least one per-fluid-pocket tailored threshold, resegmenting at least a portion of the trial representation, the resegmenting determining an enclosing boundary of at least a portion of the anatomical structure.

3. The one or more computer-readable media of claim 1 wherein the determining the at least one tailored threshold further comprises taking mean CT attenuation of a plurality of components.

4. The computer-readable media of claim 1 wherein segmenting the digital representation further comprises using a trial air threshold for the air pockets and a trial fluid threshold for the fluid pockets during the segmentation.

5. The one or more computer-readable media of claim 4 wherein the trial threshold for the trial fluid pocket comprises a threshold set at a value sufficient to segment at least 95% of the anatomical feature.

6. The one or more computer-readable media of claim 1 wherein the method further comprises determining at least one selected from the group consisting of:

a mean of CT attenuation for the trial fluid pocket, a standard deviation of CT attenuation for the trial fluid pocket, volume of the trial fluid pocket, and a bounding box of the trial fluid pocket.

7. The one or more computer-readable media of claim 1 wherein segmenting further comprises locating an air-fluid boundary between an air pocket and an adjacent trial fluid pocket.

8. The one or more computer-readable media of claim 1 wherein determining the at least one per-fluid-pocket tailored threshold comprises:

determining a component depth based on size of the air-fluid boundary; and

using the component depth and the air-fluid boundary to choose a plurality of threshold components.

9. The one or more computer readable media of claim 2 wherein the method further comprises:

determining a location of at least one trial fluid pocket, and

storing the location of the trial fluid pocket to determine if the trial fluid pocket comprises separate fluid pockets.

10. The one or more computer readable media of claim 2 wherein the method further comprises:

determining a location of at least one trial fluid pocket; and

storing the location of the trial fluid pocket to determine if the trial fluid pocket comprises separate fluid pockets.

11. The one or more computer-readable media of claim 9 wherein resegmenting at least a portion of the digital representation further comprises resegmenting at least one trial fluid pocket using the per-fluid-pocket tailored threshold to produce at least one tailored fluid pocket.

12. The one or more computer-readable media of claim 11 wherein the method further comprises:

organizing the tailored fluid pockets and trial air pockets into a hierarchical pocket tree wherein tailored fluid pockets touching the same air pocket are at the same level in the hierarchy of the pocket tree.

13. The one or more computer-readable media of claim 12 wherein air pockets touching the same fluid pocket are at the same level in the hierarchy of the pocket tree.

14. The one or more computer-readable media of claim 12 wherein the method further comprises:

analyzing tailored fluid pockets to determine whether a given fluid pocket belongs to a leaked region; and

responsive to determining that the given tailored fluid pocket belongs to a leaked region, deleting pockets below the given tailored fluid pocket in the hierarchical pocket tree.

15. The one or more computer-readable media of claim 14 wherein analyzing tailored fluid pockets to determine whether a given fluid pocket belongs to a leaked region comprises determining if a given tailored fluid pocket is a child of two or more pockets in the hierarchical tree.

16. The one or more computer-readable media of claim 2 wherein the method further comprises:

using at least a portion of results of the resegmenting to perform a third segmentation on at least portions of the digital representation to determine an enclosing boundary of the anatomical structure.

17. The one or more computer-readable media of claim 1 wherein the method further comprises:

performing equalization on at least a portion of the digital representation using the at least one tailored threshold for at least a portion of the fluid pockets to determine an equalized image of the at least a portion of the anatomical structure.

18. The one or more computer-readable media of claim 17 wherein performing equalization on at least a portion of the digital representation comprises performing fuzzy connectedness segmentation on at least a portion of the digital representation.

19. The one or more computer-readable media of claim 17 wherein the method further comprises:

determining a gap between an air region and a fluid region in the fuzzy connectedness image; and

filling the gap using a two-dimensional region-growing procedure.

20. The one or more computer-readable media of claim 19 wherein the method further comprises:

using level set segmentation to determine an enclosing boundary of the at least a portion of the anatomical structure.

21-29. -29. (canceled)

30. A computer-implemented method for generating a three-dimensional image of an anatomical structure which has sections with a contrast agent and sections without the contrast agent comprising:

accessing image data of the anatomical structure;

generating a first three-dimensional image from the image data using a first parameter to produce a digital representation of sections with the contrast agent and a second parameter to produce a digital representation of sections without the contrast agent;

for sections with the contrast agent, generating a section-specific parameter; and

generating at least a portion of a second three-dimensional image from the image data using the section-specific parameter for respective sections with the contrast agent.

31. The computer-implemented method of claim 30 wherein the parameter for at least one section with the contrast agent is a threshold based on a level of contrast agent in the section.

32. The computer-implemented method of claim 30 wherein:

the generating a first three-dimensional image comprises;

determining a first volume for respective sections with the contrast agent; and

wherein generating at least a portion of a second three-dimensional image from the image data comprises;

determining a second volume for respective sections with the contrast agent using the respective section-specific parameter.

33. The computer-implemented method of claim 32 further comprising:

generating a section tree comprising hierarchical relationships between the sections;

for at least one section with the contrast agent, determining if the at least one section should be deleted based at least in part on a relationship between the first volume of the section and the second volume of the section.

34. The computer-implemented method of claim 33 comprising determining that a section with the contrast agent should be deleted;

deleting the section detected with the contrast agent; and

deleting descendents of the section detected with the contrast agent as represented on a section tree.

35. One or more computer-readable media comprising computer-readable instructions for performing:

obtaining at least one initial digital representation of an anatomical feature comprising pockets filled with a first substance and pockets filled with a second substance;

generating an image of the anatomical feature from the initial digital representation;

for pockets filled with the first substance, calculating at least a first pocket-specific characteristic; and

generating at least a portion of a second image of the anatomical feature from the initial digital representation using at least the first pocket-specific characteristic to generate the portion of the second image comprising pockets filled with the first substance.

36. The one or more computer-readable media of claim 35 wherein the generating an image further comprises using a first parameter for the first substance and a second parameter for the second substance.

37. The one or more computer-readable media of claim 35 wherein the generating a second image further comprises calculating for respective pockets filled with the first substance a first value of a second pocket-specific characteristic and a second value of the second pocket-specific characteristic.

38. The one or more computer-readable media of claim 37 further comprising:

deleting at least one pocket filled with the first substance if a ratio which incorporates the first value of the second characteristic and the second value of the second characteristic falls outside at least one predetermined value.

39. The one or more computer-readable media of claim 37 wherein the first pocket-specific characteristic is a threshold based on an amount of a contrast agent present within a pocket and wherein the second pocket-specific characteristic is volume of the pocket.

40. A system for determining an enclosing three-dimensional colon boundary comprising:

means for receiving a digital representation of a colon;

means for segmenting the digital representation into at least air pockets and fluid pockets using an initial threshold for the air pockets and an initial threshold for the fluid pockets;

for at least a portion of the fluid pockets, means for extracting the fluid pockets, wherein extracting comprises determining at least one tailored threshold per fluid pocket; and

means for using the at least one tailored threshold for at least a portion of the fluid pockets to resegment at least a portion of the digital representation to determine an enclosing boundary of the at least a portion of the colon.

41-44. -44. (canceled)