Contour guided deformable image registration

• US 9,245,336 B2
• Filed: 12/13/2011
• Issued: 01/26/2016
• Est. Priority Date: 12/15/2010
• Status: Active Grant

• Current Assignee: Koninklijke Philips N.V.
• Original Assignee: Koninklijke Philips N.V.
• Inventors: Mallya, Yogish, Poston, Timothy
• Primary Examiner(s): Le, Vu
• Assistant Examiner(s): Cese, Kenny
• Application Number:
  US13/993,142
• Publication Number:
  US 20130259335A1
• Time in Patent Office: 1,505 Days
• Field of Search: 382/128
• US Class Current: 1/1
• CPC Class Codes: G06T 2207/10072 Tomographic images
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06T 7/30 Determination of transform parameters for the alignment of images, i.e. image registration

Abstract

A method includes obtaining first volumetric image data, which is acquired at a first time, including a region of interest with a structural feature located at a first position. The method further includes obtaining second volumetric image data, which is acquired at a second different time, including the region of interest with the structural feature located at a second different position. The method further includes determining a registration transformation that registers the first and second volumetric image data such that the at least one structural feature in the first volumetric image data aligns with the at least one structural feature in the second volumetric image data. The registration transformation is based at least on a contour guided deformation registration. The method further includes generating a signal indicative of the registration transformation.

Claims

1. A method, comprising:

obtaining first volumetric image data, which is acquired at a first time, including a region of interest of a subject, wherein at least one structural feature in the region of interest is located at a first position in the first volumetric image data;

creating a contour for the at least one structural feature on the first volumetric image data, wherein the contour identifies the at least one structural feature;

obtaining second volumetric image data, which is acquired at a second time, including the region of interest of the subject, wherein the at least one structural feature in the region of interest is located at a second position in the second volumetric image data,

wherein the second time is subsequent to the first time, and the first and second positions are different positions;

adding a direction guideline to the first volumetric image data adjacent to the contour;

determining a registration transformation that registers the first and second volumetric image data such that the at least one structural feature in the first volumetric image data aligns with the at least one structural feature in the second volumetric image data, wherein the registration transformation propagates the contour from the first volumetric image data to the second volumetric image data based on a combination of an estimated direction of a maximum slope of the contour in gray levels and the direction guideline; and

generating a signal indicative of the registration transformation.

2. The method of claim 1, wherein determining the registration transformation includes determining a deformation field vector that extends along the direction guideline.

3. The method of claim 1, wherein determining the registration transformation includes determining a deformation field vector that extends along a line parallel to the direction guideline.

4. The method of claim 3, wherein the deformation field vector is pulled towards the direction guideline.

5. The method of claim 1, wherein determining the registration transformation includes determining a deformation field vector that extends perpendicularly from the direction guideline.

6. The method of claim 2, wherein the direction guideline is a curved line.

7. The method of claim 2, wherein direction guideline line is a straight line.

8. The method of claim 2, wherein the direction guideline is a user drawn line superimposed over the first image data.

9. The method of claim 1, wherein the combination weights the direction guideline for propagating a point on the contour to the second volumetric image data as a function of a distance between the point and the direction guideline.

10. The method of claim 9, further comprising:

applying a first weight to propagate a first point on the contour based on a first distance between the point and the direction guideline, and applying a second weight to propagate a second point on the contour based on a second distance between the point and the direction guideline, wherein the first distance is less than the second distance, and the first weight is greater than the second weight.

11. The method of claim 2, wherein the first and second volumetric image data include voxels represents gray level intensity values, and determining the registration transformation, comprising:

determining a difference in intensity value between corresponding voxels in the first and second volumetric image data;

determining an intensity gradient vector for the voxel in the second volumetric image data;

determining a normal vector that is perpendicular to the direction guideline;

determining a dot product of the intensity gradient vector and the normal vector;

determining a square of a magnitude of the intensity gradient vector;

determining a square of the difference; and

calculating a deformation field vector as the product of the difference, the dot product and the normal vector divided by a product of the square of the magnitude and the square of the difference, wherein the deformation field vector is the registration transformation.

12. The method claim 11, further comprising:

regularizing the deformation field vector; and

normalizing the square of the difference a normalization factor prior to calculating the deformation field vector and using the normalized square of the difference to calculate the deformation field vector.

13. The method of claim 1, further comprising:

presenting an interactive graphical user interface which includes contouring tools, wherein the interactive graphical user interface receives an input indicative of a user defined direction guideline, wherein the user-defined direction guideline is the generated direction guideline.

14. A system, comprising:

a processor that determines a registration transformation that registers two images, wherein the registration transformation is determined based at least on voxel intensity values of the two images, voxel intensity gradient values for one of the two images, and guide vector values that indicate a direction of the registration and that correspond to a line placed that is separate from the images and that is placed next to a contour of a structure of interest in one of the two images, and wherein the registration transformation propagates the contour to a second of the two images based on a combination of an estimated direction of a maximum slope of the contour in gray levels and the line.

15. The system of claim 14, wherein the guide vector values define on the line across which the registration is performed.

16. The system of claim 14, wherein the guide vector values define on the line along which the registration is performed.

17. The system of claim 14, wherein the line is separate and distinct from the contour.

18. A method, comprising:

determining a deformation field vector which registers two images based on a contour guided deformable image registration; and

registering the two images using the deformation field vector, wherein the registering propagates a contour of a structure of interest in a first image of the two images to a second image of the two images using a weighted registration which applies a first weight to an estimated direction of a maximum slope of the contour in gray levels and applies a second weight to a direction guideline placed adjacent to the contour.

19. The method of claim 18, wherein the first image is acquired at a first time and includes the structure of interest at a first location and the contour, which identifies the structure of interest, and the second image is acquired at a second different time and includes the structure of interest at a second different location.

20. The method of claim 19, wherein the contour is created by a user via a graphical user interface.

21. The method of claim 18, wherein the first image is a treatment planning image and the second image is a follow-up image.

22. The method of claim 21, wherein the second image is acquired during a treatment procedure.

23. The method of claim 22, wherein the treatment procedure includes at least one of radiation therapy, chemotherapy, particle therapy, high intensity focused ultrasound, ablation, interferential therapy, transcutaneous electrical nerve stimulation, pulsed shortwave therapy, laser therapy, or a combination thereof.

24. The method of claim 18, wherein the direction guideline is superimposed over the first image and extends along or across the direction guideline.

25. The method of claim 18, wherein the direction guideline is a user drawn line superimposed over the first image.