Method for identification of anatomical landmarks

• US 8,160,677 B2
• Filed: 10/20/2006
• Issued: 04/17/2012
• Est. Priority Date: 09/08/2006
• Status: Active Grant

• Current Assignee: Medtronic Incorporated (Medtronic Plc)
• Original Assignee: Medtronic Incorporated (Medtronic Plc)
• Inventors: Gielen, Frans L. H., Hertel, Frank, Gemmar, Peter, Appenrodt, Peter
• Primary Examiner(s): Casler, Brian
• Assistant Examiner(s): Santos, Joseph M
• Application Number:
  US11/584,814
• Publication Number:
  US 20080123923A1
• Time in Patent Office: 2,006 Days
• Field of Search: 600/407, 600/426, 600/427, 600/429, 382/128, 382/154
• US Class Current: 600/427
• CPC Class Codes: G06T 2207/10088 Magnetic resonance imaging [MRI]
• CPC Class Codes: G06T 2207/30016 Brain
• CPC Class Codes: G06T 7/32 using correlation-based methods
• CPC Class Codes: G06T 7/73 using feature-based methods
• CPC Class Codes: G06V 10/245 by locating a pattern; Special marks for positioning

Abstract

A method for determining anatomical landmarks is provided. The anatomical landmarks can be used to determine anatomical targets of a patient. The method may be used for planning a procedure and performing a procedure, such as performing a procedure in the brain.

Claims

1. A method of determining an anatomical landmark in an anatomy of a patient, comprising:

obtaining image data of a brain of the patient, including at least one of obtaining a T1 weighted magnetic resonance image data, a T2 weighted magnetic resonance image data, X-ray image data, computed tomography image data, and Positron emission tomography image data;

pre-processing image data of the brain with a processor system configured to at least smooth the image data;

determining a seed portion in at least one of the obtained image data or the smoothed image data;

after pre-processing the image data, segmenting the image data based upon a region growing from the seed portion to identify a third ventricle of the brain as at least a portion of a selected region in the image data;

identifying the anatomical landmark as at least one of an anterior commissure and a posterior commissure after segmenting the third ventricle and based on the segmented third ventricle in the image data as a bright portion that includes an intensity change at an edge of the segmented third ventricle, wherein at least one of the anterior commissure and the posterior commissure is identified in the selected region; and

after identifying the anatomical landmark, determining a location of an anatomical target based on the identified anatomical landmark, wherein the anatomical target is substantially indistinguishable from surrounding structures in the image data of the brain.

2. The method of claim 1, further comprising:

executing instructions with the processor system to segment the third ventricle based upon the seed portion.

3. The method of claim 2, wherein determining a seed portion of the image data includes identifying a pixel, a voxel, or combinations thereof defining a portion of the third ventricle of a brain.

4. The method of claim 3, wherein determining a seed portion includes a user identifying in the image data a seed portion.

5. The method of claim 2, wherein region growing includes determining all of the portions of the image data that contact the seed portion of the image data that include characteristics similar to the seed portion at least by the execution of instructions with the processor system.

6. The method of claim 5, wherein determining a selected region based on the seed portion includes:

determining whether an intensity in a seed voxel reaches into a test voxel within a threshold distance when the test voxel contacts the seed voxel;

adding the test voxel to the seed voxel to create a homogeneity region at least within the selected region; and

finding each of the voxels in the image data that contact the homogeneity region that include an intensity that extends a distance into the test voxel.

7. The method of claim 6, further comprising:

providing the processor system;

providing a storage medium operable to store the image data including information about a plurality of voxels in the image data;

executing instructions with the processor system to determine whether the threshold distance has been exceeded in the test voxel; and

executing instructions with the processor system to add a test voxel to a homogeneous region if the threshold distance is reached.

8. The method of claim 7, further comprising:

determining a line that includes ends that are substantially equal distance from the edge of the homogeneity region in the image data; and

aligning a plurality of slices of image data based upon the determined line;

wherein aligning the image data includes substantially removing a slope from the line.

9. The method of claim 8, further comprising:

determining a brightest pixel in the image data of a brain that is located immediately inferior to a column of fornix in a mid-sagittal plane that is extracted from the image data based upon the line.

10. The method of claim 9 further comprising:

segmenting the image data to partition the image data into two classes including an object and a background to identify a cerebral aqueduct of the brain in the image data; and

identifying an area in the image data dorsal to the cerebral aqueduct as it interconnects with the determined third ventricle region.

11. The method of claim 10, further comprising:

determining a position of a posterior commissure from the area in the image data.

12. The method of claim 1, wherein pre-processing the image data includes non-linear anisotropic diffusion of the image data.

13. The method of claim 1, wherein pre-processing the image data further includes at least one of reducing noise of the image data, maintaining mass in the image data, defining an edge, removing unneeded data, or combinations thereof in the image data.

14. The method of claim 1, wherein determining a selected region based on the seed portion includes:

determining whether a seed voxel that includes an intensity that is within a threshold distance from the seed voxel;

adding the test voxel to the seed voxel to create a homogeneity region; and

finding each of the voxels in the image data that contact the homogeneity region that meet the intensity of the seed voxel within the threshold distance.

15. The method of claim 1, wherein determining the location of the anatomical target is based upon both the identified anterior commissure and the identified posterior commissure.

16. The method of claim 15, wherein determining the anatomical target further includes moving a sensing device near a proposed anatomical target to sense an activity near the proposed anatomical target to refine an edge of the proposed anatomical target.

17. The method of claim 15, further comprising:

determining a location of the sub-thalamic nucleus as the anatomical target based upon a predetermined location of the sub-thalamic nucleus relative to the anterior commissure and the posterior commissure;

guiding an instrument to the sub-thalamic nucleus; and

stimulating the sub-thalamic nucleus.

18. A method of determining an anatomical landmark of a brain of a patient, comprising:

obtaining image data of the brain of the patient including a plurality of slices spaced apart and along an axis relative to the brain;

processing the image data of the brain of the patient with a processor system;

determining a third ventricle of the brain in the obtained image data of the brain of the patient by executing instructions with the processor system;

after determining the third ventricle, executing instructions with the processor system to determine a skeletonized line including ends that are equal distance from an edge of the determined third ventricle within the obtained image data of the brain of the patient, wherein determining the skeletonized line is based upon minimizing the determined third ventricle until a region of minimally connected pixels is determined;

after executing instructions with the processor system to determine a skeletonized line, aligning the plurality of slices of image data around the axis based upon the skeletonized line in each of the image data slices by substantially removing the slope of the skeletonized line between each of the image slices of the image data;

after aligning the plurality of slices of the image data, executing further instructions with the processor system to extract a mid-sagittal plane from the image data that is aligned with the skeletonized line;

after extracting the mid-sagittal plane, determine the anatomical landmark in the brain; and

after identifying the anatomical landmark, determining a location of the anatomical target based on the identified anatomical landmark, wherein the anatomical target is substantially indistinguishable from surrounding structures in the image data of the brain.

19. The method of claim 18 wherein obtaining the image data includes obtaining image data of the brain of the patient with a T1 weighted MRI, a T2 weighted MRI, a computed tomography scan, a positron emission tomography scan, image data showing spatial or time varying constructs, or combinations thereof.

20. The method of claim 18, wherein determining a third ventricle includes executing instructions with the processor system to:

determine the third ventricle of the brain in the image data by identifying test voxels in the image data that meet a threshold distance intensity when compared to a seed voxel; and

determine a homogeneous region in the image data;

wherein the skeletonized line is determined as a line that has ends that are equal distance from an edge of the determined third ventricle that are equal distance from selected edges of the homogeneous region of the third ventricle.

21. The method of claim 20, wherein determine the anatomical landmark in the brain includes determining a location in the obtained image data of a posterior commissure of the brain from the aligned plurality of slices of the image data including determining a brightest pixel in the image data that is located immediately inferior to the column of the fornix in the mid-sagittal plane, wherein the mid-sagittal plane is extracted from the aligned plurality of slices of the image data.

22. The method of claim 21, further comprising:

after extracting the mid-sagittal plane, determining a location of an anterior commissure in the obtained image data that includes the aligned plurality of slices based upon the determined third ventricle; and

wherein determining a location of the anatomical target based on the identified anatomical landmark is based upon the determined location of the anterior commissure and the posterior commissure.

23. The method of claim 22, wherein the anatomical target is indistinguishable in the image data and is the sub-thalamic nucleus and the determined location of the sub-thalamic nucleus is based upon a predetermined location of the sub-thalamic nucleus relative to the anterior commissure and the posterior commissure in the brain.

24. The method of claim 23, further comprising:

guiding an instrument relative to the sub-thalamic nucleus; and

performing stimulation of the sub-thalamic nucleus.

25. The method of claim 21, further comprising:

executing instructions with the processor system to determine a location of an anatomical target based on the determined posterior commissure of the brain, wherein the anatomical target is substantially indistinguishable in the image data of the brain.

26. A method of determining an anatomical landmark of a brain of a patient, comprising:

obtaining image data of the brain of the patient;

determining a third ventricle in a first region of the brain in the image data of the patient;

after determining the third ventricle, a processor system executing instructions configured to determine an anterior commissure as a portion of the image data that is a brightest pixel anterior to the determined third ventricle that is in the obtained image data and determining a posterior commissure relative to the determined third ventricle as a brightest pixel in the obtained image data just dorsal to a point where a cerebral aqueduct enters the determined third ventricle; and

after determining the anterior commissure and the posterior commissure, determining an anatomical target that is indistinguishable in the obtained image data from surrounding structures based upon the determined location of the anterior commissure and the posterior commissure.

27. The method of claim 26, wherein the anatomical target is the sub-thalamic nucleus and a location of the sub-thalamic nucleus is based upon a predetermined location of the sub-thalamic nucleus relative to the anterior commissure and the posterior commissure.

28. The method of claim 27, further comprising:

guiding an instrument relative to the sub-thalamic nucleus; and

performing stimulation of the sub-thalamic nucleus.

29. The method of claim 26, wherein determining a posterior commissure includes:

segmenting a cerebral aqueduct in the obtained image data by determining whether an object pixel has an intensity greater than a threshold brightness when compared to a background pixel.

30. The method of claim 26, wherein determining an anterior commissure as a portion of the image data anterior to the determined third ventricle that is the brightest pixel in the obtained image data at least by executing instructions with the processor system, includes:

determining an edge of the anterior commissure in the obtained image data as a bright data adjacent to the determined third ventricle that is dark, wherein the edge of the anterior commissure is where the anterior commissure and the determined ventricle meet and where the intensity changes from the determined third ventricle.

31. A method of determining an anatomical landmark of a brain of a patient, comprising:

obtaining image data of the brain including a plurality of slices of the brain of the patient, wherein the plurality of slices are spaced apart along an axis in the brain;

executing instructions with a processor system configured to include;

processing the image data to at least one of smooth the image data, remove unneeded data from the image data, define edges in the image data, or combinations thereof;

after the processing the image data, determining a third ventricle in the image data of the brain at least in part by segmenting the image data including comparing a test voxel intensity to a threshold intensity, adding the test voxel to a seed voxel if the test voxel intensity is within the threshold intensity to begin growing a homogeneous region, and identifying every test voxel that touches the growing homogeneous region until no further test voxels have the test voxel intensity within the threshold intensity;

after determining the third ventricle, determining a skeletonized line in the determined third ventricle in each of the plurality of slices by successively thinning the edges of the homogeneous region and stopping when a line of portions that are minimally connected includes ends that are substantially equal distant from the boundaries of the homogeneous region;

after determining the skeletonized line, aligning each of the slices of the image data around the axis based upon the first region of the image data by removing a slope between the plurality of slices of the brain of the patient based upon the skeletonized line;

after aligning each of the slices of the image data to generate aligned image data, extracting a mid-sagittal plane based upon a sagittal plane formed by the plurality of skeleton lines in the plurality of slices in the aligned image data; and

after extracting the mid-sagittal plane, determining a landmark region in the image data of the brain.

32. The method of claim 31, wherein obtaining the image data includes obtaining a plurality of T1 weighted magnetic resonance image data slices, a T2 weighted MRI, a computed tomography scan, a positron emission tomography scan, image data showing spatial or time varying constructs, or combinations thereof of the brain.

33. The method of claim 32, wherein processing the image data includes at least one of applying a linear Gaussian filter, non-linear median filter, non-linear anisotropic diffusion, or combinations thereof to the obtained image data.

34. The method of claim 32, wherein segmenting the image data includes at least one of region growing, geometric active contour models, or combinations thereof.

35. The method of claim 32, wherein determining a landmark region in the image data of the brain, includes executing instructions with the processor system including:

determining an area in the generated aligned image data located immediately inferior to the column of the fornix in the image data of the brain;

determining a brightest pixel in the area located immediately inferior to the column of the fornix in the same level as the mid-sagittal plane; and

determining the brightest pixel to be an anterior commissure of the brain in the image data.

36. The method of claim 35, wherein determining a landmark region in the image data of the brain, includes executing instructions with the processor system including determining a location of a posterior commissure in the image data;

and

executing an instruction with the processor system to determine a location of an anatomical target that is indistinguishable in the obtained image data of the brain based upon a predetermined location of the anatomical target to the anterior commissure and the posterior commissure.

37. The method of claim 36, further comprising;

guiding an instrument to the anatomical target; and

at least one of stimulating the anatomical target, providing a gene therapy to the target, providing a chemical therapy to the target;

providing a protein therapy to the target, or combinations thereof.

38. The method of claim 32, wherein determining the landmark region of the generated aligned image data includes:

executing the instructions by the processor system to, in order;

segment the image data at the posterior boundary of the determined mid-sagittal plane to determine an object portion from a background portion;

determine that the object portion is a cerebral aqueduct;

determine the area where the cerebral aqueduct opens into the defined first region in the image data;

determine the brightest portion just dorsal to the point where the cerebral aqueduct opens into the first region; and

determine that the brightest portion defines a point of the posterior commissure in the brain.

39. The method of claim 38,

wherein executing instructions with a processor system further includes:

determine a location of an anterior commissure in the image data; and

determine a location of an anatomical target that is indistinguishable in the obtained image data of the brain based upon a predetermined location of the anatomical target to the anterior commissure and the posterior commissure.

40. The method of claim 39, further comprising;

guiding a deep brain stimulation probe to the anatomical target; and

stimulating the anatomical target with the deep brain stimulation probe.