METHOD FOR INTERACTIVELY DETERMINING A BOUNDING SURFACE FOR SEGMENTING A LESION IN A MEDICAL IMAGE

• US 20090279754A1
• Filed: 12/22/2008
• Published: 11/12/2009
• Est. Priority Date: 05/06/2008
• Status: Active Grant

• Current Assignee: CareStream Health Incorporated
• Original Assignee: CareStream Health Incorporated
• Inventors: Ray, Lawrence A., Simon, Richard A., Gindele, Edward B.
• Granted Patent: US 8,150,120 B2
• Time in Patent Office: Days
• US Class Current: 382/128
• CPC Class Codes: G06T 2207/20104 Interactive definition of region of interest [ROI]
• CPC Class Codes: G06T 2207/20156 Automatic seed setting
• CPC Class Codes: G06T 2207/30004 Biomedical image processing
• CPC Class Codes: G06T 7/11 Region-based segmentation

Abstract

A method of bounding an anatomical object of interest in a 3-dimensional volume image includes displaying an image of at least a portion of the object, selecting a plurality of points in the displayed image, at least a first and second point of the plurality of points spanning the object, forming a non-rectilinear surface bounding the plurality of points, identifying a seed point within the surface and extracting a plurality of statistical values corresponding to image voxels disposed proximate the seed point, and classifying image voxels within the surface into a first class and a second class based on the plurality of statistical values.

Claims

1. A method of bounding an anatomical object of interest in a 3-dimensional volume image, comprising:

displaying an image of at least a portion of the object;

selecting a plurality of points in the displayed image, at least a first and second point of the plurality of points spanning the object;

forming a non-rectilinear surface bounding the plurality of points;

identifying a seed point within the surface and extracting a plurality of statistical values corresponding to image voxels disposed proximate the seed point; and

classifying image voxels within the surface into a first class and a second class based on the plurality of statistical values.

2. The method of claim 1, wherein the surface is one of spherical and elliptical.

3. The method of claim 1, wherein the seed point is located at a midpoint between the first and second points.

4. The method of claim 1, wherein the surface is centered about the seed point.

5. The method of claim 1, wherein the plurality of points further includes a third point and a fourth point, the third and fourth points spanning the object.

6. The method of claim 5, wherein at least three of the first, second, third, and fourth points lie on the surface.

7. The method of claim 5, wherein at least three of the first, second, third, and fourth points are coplanar.

8. The method of claim 5, wherein a line segment connecting the first and second points approximates a maximum length of the object, and a line segment connecting the third and fourth points approximates a maximum width of the object.

9. The method of claim 5, wherein the plurality of points further includes a fifth point and a sixth point, the fifth and sixth points spanning the object.

10. The method of claim 1, further including calculating a set of noise-filtered voxel values corresponding to the image voxels within the surface.

11. The method of claim 10, further including classifying the image voxels within the surface into the first and second classes based on the set of noise-filtered voxel values.

12. A method of bounding an anatomical object of interest in a 3-dimensional volume image, comprising:

displaying an image of at least a portion of the object;

selecting a first and second pair of points in the displayed image, the first pair and the second pair spanning the object;

forming an ellipsoid bounding the first pair and the second pair;

identifying a seed point within the ellipsoid and extracting a plurality of statistical values corresponding to image voxels disposed proximate the seed point; and

classifying image voxels within the ellipsoid into a first class and a second class based on the plurality of statistical values.

13. The method of claim 12, wherein the ellipsoid is an ellipsoid of revolution.

14. The method of claim 12, further including selecting a third pair of points in the displayed image, the third pair spanning the object.

15. The method of claim 14, wherein the ellipsoid is a minimal volume ellipsoid containing the first, second, and third pairs of points.

16. The method of claim 14, wherein at least one pair of the first, second, and third pairs of points lies in a different plane than the remaining pairs of points.

17. The method of claim 14, wherein a line segment connecting the first pair of points approximates a maximum length of the object, a line segment connecting the second pair of points approximates a maximum width of the object, and a line segment connecting the third pair of points approximates a maximum height of the object.

18. The method of claim 12, wherein classifying the image voxels within the ellipsoid comprises comparing image voxel values corresponding to the image voxels within the ellipsoid to a mean statistical value.

19. A method of bounding an anatomical object of interest in a 3-dimensional volume image, comprising:

displaying an image of at least a portion of the object;

selecting a first and second pair of points in the displayed image, the first pair and the second pair spanning the object;

forming a minimum area ellipse bounding the first pair and the second pair in response to the selection of the first pair and the second pair;

identifying a seed point within the ellipse and extracting a plurality of statistical values corresponding to voxels disposed proximate the seed point; and

classifying voxels within the ellipse into a first class and a second class based on the plurality of statistical values.

20. The method of claim 19, wherein the object comprises a lesion.