Fissure detection methods for lung lobe segmentation

• US 7,711,167 B2
• Filed: 12/04/2006
• Issued: 05/04/2010
• Est. Priority Date: 12/07/2005
• Status: Active Grant

• Current Assignee: Siemens Medical Solutions USA Incorporated (Siemens AG)
• Original Assignee: Siemens Medical Solutions USA Incorporated (Siemens AG)
• Inventors: Novak, Carol L., Kiraly, Atilla Peter
• Primary Examiner(s): AZARIAN, SEYED H
• Application Number:
  US11/566,323
• Publication Number:
  US 20070133894A1
• Time in Patent Office: 1,247 Days
• Field of Search: 382/100, 382/128, 382/129, 382/130, 382/131, 382/132, 382/133, 382/134, 382/154, 382/162, 382/168, 382/181, 382/199, 382/203, 382/219, 382/232, 382/254, 382260-266, 382/274, 382/276, 382/285, 382/291, 382/305, 382/312, 378/8, 128/207.14, 600/410, 312/100
• US Class Current: 382/128
• CPC Class Codes: G06V 10/443 by matching or filtering
• CPC Class Codes: G06V 2201/031 of internal organs

Abstract

A fissure detection method for lung lobe segmentation in 3D image data is disclosed. In this method, 3D lung image data is filtered using one or more filters based on at least one of planar structures coupled with vessel suppression, curvature computations, and local gradient magnitude and direction comparisons. Fissures are detected in the 3D lung image data based on the filtered 3D lung image data, and lung lobes are segmented from the 3D lung image data based on the detected fissures.

Claims

1. A method comprising:

filtering 3D lung image data, the 3D lung image data comprising a plurality of voxels and using at least one of a filter based on both planar structures and vessel suppression,

a filter based on curvature computations, and a filter based on local gradient magnitude and direction comparisons;

for each voxel of said plurality of voxels, calculating a score representing the likelihood of the voxel being a fissure using said at least one of a filter based on both planar structures and vessel suppression, a filter based on curvature computations,

and a filter based on local gradient magnitude and direction comparisons; and

identifying lobar fissures in the 3D lung image data based on the filtered 3D lung image data resulting from said filtering step.

2. The method of claim 1, wherein said step of identifying lobar fissures comprises:

performing region growing based on the filtered 3D lung image data resulting from said filtering step.

3. The method of claim 1, wherein said step of identifying lobar fissures comprises:

executing a level-set operation based on the filtered 3D lung image data resulting from said filtering step.

4. The method of claim 1, wherein said step of filtering 3D lung image data comprises:

filtering said 3D lung image data using vessel suppressing plane filter based on a Hessian matrix.

5. The method of claim 1, further comprising:

segmenting lung lobes from the 3D lung image data based on the identified lobar fissures.

6. The method of claim 5, further comprising:

receiving 3D image data; and

segmenting the 3D lung image data from the received 3D image data.

7. The method of claim 1, wherein said step of filtering 3D lung image data comprises:

filtering the 3D lung image data using two or more of a filter based on both planar structures and vessel suppression, a filter based on curvature computations, and a filter based on local gradient magnitude and direction comparisons.

8. The method of claim 1, wherein said step of identifying fissures in the 3D lung image data comprises:

performing hysteresis thresholding on the filtered 3D lung image data to determine fissure candidate voxels; and

identifying large groups of candidate voxels as fissures.

9. The method of claim 1, further comprising:

segmenting at least one of vessels and airways from the 3D lung image data.

10. The method of claim 9, wherein said step of detecting fissures in the 3D lung image data based on the filtered 3D lung image data comprises:

detecting fissures at locations in the 3D lung image data where said at least one of vessels and airways are not located.

11. An apparatus comprising:

means for filtering 3D lung image data, the 3D lung image data comprising a plurality of voxels and using at least one of a filter based on both planar structures and vessel suppression, a filter based on curvature computations, and a filter based on local gradient magnitude and direction comparisons;

mean for calculating a score for each voxel of said plurality of voxels, the score representing the likelihood of the voxel being a fissure using said at least one of a filter based on both planar structures and vessel suppression, a filter based on curvature computations, and a filter based on local gradient magnitude and direction comparisons;

and means for identifying lobar fissures in the 3D lung image data based on the filtered 3D lung image data.

12. The apparatus of claim 11, wherein said means for identifying lobar fissures comprises:

means for performing region growing based on the filtered 3D lung image data.

13. The apparatus of claim 11, wherein said means for identifying lobar fissures comprises:

means for executing a level-set operation based on the filtered 3D lung image data.

14. The apparatus of claim 11, wherein said filter based on both planar structures and vessel suppression filters the 3D lung image data using a Hessian matrix.

15. The apparatus of claim 11, further comprising:

means for segmenting lung lobes from the 3D lung image data based on the identified lobar fissures.

16. The apparatus of claim 11, wherein said means for filtering 3D lung image data comprises:

means for filtering the 3D lung image data using two or more of a filter based on both planar structures and vessel suppression, a filter based on curvature computations, and a filter based on local gradient magnitude and direction comparisons.

17. The apparatus of claim 11, wherein said means for identifying fissures in the 3D lung image data comprises:

means for performing hysteresis thresholding on the filtered 3D lung image data to determine fissure candidate voxels; and

means for identifying large groups of candidate voxels as fissures.

18. The apparatus of claim 11, further comprising:

means for segmenting at least one of vessels and airways from the 3D lung image data.

19. A computer readable medium storing computer program instructions, said computer program instructions defining the steps comprising:

filtering 3D lung image data, the 3D lung image data comprising a plurality of voxels, based on at least one of planar structures coupled with vessel suppression, curvature computations, and local gradient magnitude and direction comparisons;

for each voxel of said plurality of voxels, calculating a score representing the likelihood of the voxel being a fissure using said at least one of a filter based on both planar structures and vessel suppression, a filter based on curvature computations, and a filter based on local gradient magnitude and direction comparisons; and

identifying lobar fissures in the 3D lung image data based on the filtered 3D lung image data resulting from said filtering step.

20. The computer readable medium of claim 19, wherein the computer program instructions defining the step of identifying lobar fissures comprise computer program instructions defining the step of:

performing region growing based on the filtered 3D lung image data resulting from said filtering step.

21. The computer readable medium of claim 19, wherein the computer program instructions defining the step of identifying lobar fissures comprise computer program instructions defining the step of:

executing a level-set operation based on the filtered 3D lung image data resulting from said filtering step.

22. The computer readable medium of claim 21, wherein the computer program instructions defining the step of filtering 3D lung image data comprise computer program instructions defining the step of:

filtering 3D lung image data based on planar structures coupled with vessel suppression using a Hessian matrix.

23. The computer readable medium of claim 19, further comprising computer program instructions defining the step of:

segmenting lung lobes from the 3D lung image data based on the identified lobar fissures.

24. The computer readable medium of claim 19, wherein the computer program instructions defining the step of filtering 3D lung image data comprise computer program instructions defining the step of:

filtering the 3D lung image data based on two or more of planar structures coupled with vessel suppression, curvature computations, and local gradient magnitude and direction comparisons.

25. The computer readable medium of claim 19, wherein the computer program instructions defining the step of identifying fissures in the 3D lung image data comprise computer program instructions defining the steps of:

performing hysteresis thresholding on the filtered 3D lung image data to determine fissure candidate voxels; and

identifying large groups of candidate voxels as fissures.

26. The computer readable medium of claim 19, further comprising computer program instructions defining the step of:

segmenting at least one of vessels and airways from the 3D lung image data.

27. The computer readable medium of claim 26, wherein the computer program instructions defining the step of detecting fissures in the 3D lung image data based on the filtered 3D lung image data comprise computer program instructions defining the step of:

detecting fissures at locations in the 3D lung image data where said at least one of vessels and airways are not located.