Motion artifacts reduction algorithm for two-exposure dual-energy radiography

• US 6,661,873 B2
• Filed: 01/28/2002
• Issued: 12/09/2003
• Est. Priority Date: 01/28/2002
• Status: Expired due to Term

• Current Assignee: GE Medical Systems Global Technology Company LLC (GE Healthcare Technologies, Inc.)
• Original Assignee: GE Medical Systems Global Technology Company LLC (GE Healthcare Technologies, Inc.)
• Inventors: Jabri, Kadri N., Avinash, Gopal B.
• Primary Examiner(s): BRUCE, DAVID VERNON
• Application Number:
  US10/058,616
• Publication Number:
  US 20030142787A1
• Time in Patent Office: 680 Days
• Field of Search: 378/5, 378/62, 378/98.8, 378/98.9, 378/98.11, 378/98.12
• US Class Current: 378/98.11
• CPC Class Codes: A61B 6/4035 the source being combined with a filter or grating
• CPC Class Codes: A61B 6/4233 using matrix detectors
• CPC Class Codes: A61B 6/482 involving multiple energy imaging
• CPC Class Codes: A61B 6/505 for diagnosis of bone
• CPC Class Codes: A61B 6/5264 due to motion
• CPC Class Codes: G06T 2207/30004 Biomedical image processing
• CPC Class Codes: G06T 5/50 using two or more images, e.g. averaging or subtraction
• CPC Class Codes: G06T 7/33 using feature-based methods

Abstract

The present technique provides a variety of processing schemes for decomposing soft tissue and bone images more accurately from low and high-energy images acquired from an imaging system, such as a dual-energy digital radiography system using flat-panel technology. In particular, a pre-decomposition process is provided for spatially matching, or registering, low and high-energy images using warping registration prior to dual energy image decomposition, which creates the soft tissue and bone images. Accordingly, the pre-decomposition process reduces motion artifacts between the low and high-energy images, thereby improving image clarity of the decomposed soft tissue and bone images.

Claims

1. A method of improving image clarity of soft tissue and bone images decomposable from first and second energy images acquired by a digital radiography imaging system at different times, comprising the act of registering the first energy image to the second energy image using warping registration prior to decomposing the first and second energy images into the soft tissue and bone images.

2. The method of claim 1, wherein the act of registering the first energy image to the second energy image comprises the act of spatially matching anatomical features in the first and second energy images.

3. The method of claim 2, wherein the act of spatially matching anatomical features comprises the act of warping the first energy image relative to the second energy image.

4. The method of claim 1, wherein the act of registering the first energy image to the second energy image comprises the act of substantially reducing motion variations between the first and second energy images.

5. The method of claim 1, wherein the act of registering the first energy image to the second energy image comprises the act of computing shift vectors for transforming the first energy image relative to the second energy image.

6. The method of claim 5, wherein the act of computing shift vectors comprises the act of computing a pixel shift vector for each pixel of the first energy image.

7. The method of claim 5, wherein the act of computing shift vectors comprises the act of computing an area shift vector for each spatial area of a spatial matrix for the first energy image.

8. The method of claim 7, wherein the act of computing the area shift vectors comprises the act of defining the spatial matrix as a maximum number of non-overlapping spatial areas centered on the first energy image and retaining a border to allow for a desired search range of image shifting.

9. The method of claim 7, wherein the act of computing the area shift vector comprises the act of decomposing a soft tissue sub-image from the first and second energy images for each shift vector of a plurality of possible shift vectors for each spatial area of the spatial matrix.

10. The method of claim 9, wherein the act of computing the area shift vector comprises the act of selecting one shift factor of the plurality of possible shift vectors for the spatial area that provides a minimum total edge strength for the spatial area.

11. The method of claim 9, wherein the act of selecting one of the plurality of possible shift vectors comprises the acts of:

limiting the plurality of possible shift vectors to a spatial search range;

computing a total edge strength for each of the plurality of possible shift vectors within the spatial search range; and

identifying the shift vector corresponding to the minimum total edge strength.

12. The method of claim 7, wherein the act of computing shift vectors comprises the act of interpolating pixel shift vectors for each pixel of the first energy image from the area shift vectors of adjacent spatial areas of the spatial matrix.

13. The method of claim 1, comprising the act of acquiring the first and second energy images at different energy levels, which comprise low and high energy levels.

14. The method of claim 13, wherein the act of acquiring the first and second energy images comprises the act of acquiring the first and second energy images over a time interval.

15. The method of claim 13, wherein the act of acquiring the first and second energy images is performed using flat-panel detector technology of the digital radiography imaging system.

16. The method of claim l,comprising the act of decomposing the soft tissue and bone images from the first and second energy images after being registered.

17. A method of reducing motion artifacts for two exposure dual energy radiography prior to decomposition of soft tissue and bone images, comprising:

obtaining first and second energy images from a digital radiography imaging system over a time interval;

defining a maximum pixel range that any point in the first energy image is allowed to shift relative to the second energy image;

defining a spatial matrix for analyzing the first energy image;

decomposing a sub-image from the first and second energy images for each shift vector of a plurality of possible shift vectors for each spatial area of the spatial matrix;

computing a total edge strength for each possible shift vector within the maximum pixel range for each spatial area of the spatial matrix;

defining an area shift vector for each spatial area to be the shift vector that minimizes the total edge strength for the spatial area;

computing a pixel shift vector for each pixel within the first energy image by interpolation based on adjacent area shift vectors; and

registering the first energy image to the second energy image by warping the first energy image pixel-by-pixel using the pixel shift vectors.

18. The method of claim 17, wherein the act of defining the spatial matrix comprises the acts of:

centering a maximum number of non-overlapping spatial areas on the first energy image; and

retaining a border around the spatial matrix to allow for the maximum pixel range.

19. The method of claim 17, wherein the act of registering the first energy image to the second energy image comprises the act of spatially matching anatomical features in the first and second energy images.

20. The method of claim 19, wherein the act of spatially matching anatomical features comprises the act of warping the first energy image relative to the second energy image.

21. The method of claim 19, wherein the act of registering the first energy image to the second energy image comprises the act of substantially reducing motion variations between the first and second energy images.

22. The method of claim 19, wherein the first and second energy images correspond to low-energy and high-energy image acquisition levels, respectively.

23. The method of claim 19, wherein the first and second energy images correspond to first and second acquisition times over the time interval.

24. The method of claim 19, wherein the first and second energy images are acquired from the digital radiography imaging system using flat-panel detector technology.

25. The method of claim 19, comprising the act of decomposing soft tissue and bone images from the first and second energy images after being registered.

26. A method of producing soft tissue and bone images of the desired anatomy of a patient, comprising the acts of:

acquiring first and second images of the desired anatomy at different times using a digital radiography imaging system having flat-panel detector technology, wherein one image of the first and second images is acquired at a low-energy level and a remaining image of the first and second images is acquired at a high-energy level;

registering the first image to the second image using warping registration; and

decomposing soft tissue and bone images of the desired anatomy from the warpingly registered first and second images.

27. The method of claim 26, wherein the act of acquiring the first and second images of the desired anatomy comprises the act of imaging chest anatomy over a time interval.

28. The method of claim 26, wherein the act of registering the first image to the second image comprises the act of spatially matching anatomical features in the first and second images.

29. The method of claim 28, wherein the act of spatially matching anatomical features comprises the act of warping the first image relative to the second image.

30. The method of claim 26, wherein the act of registering the first image to the second image comprises the act of substantially reducing motion variations between the first and second images corresponding to movement of the desired anatomy during acquisition of the first and second images.

31. The method of claim 26, wherein the act of registering the first image to the second image comprises the act of computing shift vectors for transforming the first image relative to the second image.

32. The method of claim 31, wherein the act of computing shift vectors comprises the act of computing a pixel shift vector for each pixel of the first image.

33. The method of claim 31, wherein the act of computing shift vectors comprises the act of computing an area shift vector for each spatial area of a spatial matrix for the first image.

34. The method of claim 33, wherein the act of computing the area shift vectors comprises the act of defining the spatial matrix as a maximum number of non-overlapping spatial areas centered on the first image and retaining a border to allow for a desired search range for pixel shifting.

35. The method of claim 33, wherein the act of computing the area shift vector comprises the acts of:

decomposing a sub-image from the first and second images for each shift vector of a plurality of possible shift vectors for each spatial area of the spatial matrix; and

selecting one shift vector of the plurality of possible shift vectors for the spatial area that provides a minimum total edge strength for the spatial area.

36. The method of claim 35,wherein the act of selecting one of the plurality of possible shift vectors comprises the acts of:

limiting the plurality of possible shift vectors to a spatial search range;

computing a total edge strength for each of the plurality of possible shift vectors within the spatial search range; and

identifying the shift vector corresponding to the minimum total edge strength for the spatial area.

37. The method of claim 33, wherein the act of computing shift vectors comprises the act of interpolating pixel shift vectors for each pixel of the first image from the area shift vectors of adjacent spatial areas of the spatial matrix.

38. A computer program for processing image data acquired from a digital radiography imaging system, comprising:

a tangible medium configured to support machine-readable code; and

machine-readable code supported on the medium and comprising a pre-decomposition warping-image-registration routine adapted to register first and second images prior to decomposition into soft tissue and bone images, wherein the first and second images are obtained from the digital radiography imaging system at different energy levels and different times.

39. The computer program of claim 38, wherein the machine-readable code comprises an acquisition routine adapted to acquire the first and second images at the different energy levels, which comprise low and high energy levels.

40. The computer program of claim 39, wherein the acquisition routine comprises a timing routine adapted to acquire the first and second images over a time interval.

41. The computer program of claim 38, wherein the machine-readable code comprises a decomposition routine adapted to decompose the soft tissue and bone images from the first and second images after being registered.

42. The computer program of claim 38, wherein the pre-decomposition image registration routine comprises a routine for spatially matching anatomical features in the first and second images.

43. The computer program of claim 42, wherein the routine for spatially matching anatomical features is adapted to reduce motion variations between the first and second images.

44. The computer program of claim 38, wherein the pre-decomposition image registration routine comprises a routine adapted to compute shift vectors to warp the first image relative to the second image.

45. The computer program of claim 44, wherein the routine for computing shift vectors comprises a routine adapted to compute a pixel shift vector for each pixel of the first image.

46. The computer program of claim 44, wherein the routine for computing shift vectors comprises a routine adapted to compute an area shift vector for each spatial area of a spatial matrix for the first image.

47. The computer program of claim 46, wherein the routine adapted to compute the area shift vector comprises:

a sub-image decomposition routine adapted to decompose a soft tissue sub-image from the first and second images for each shift vector of a plurality of possible shift vectors for each spatial area of the spatial matrix; and

a selection routine adapted to select one shift vector of the plurality of possible shift vectors in a maximum shift search range for the spatial area that provides a minimum total edge strength for the spatial area.

48. The computer program of claim 46, wherein the routine adapted to compute shift vectors comprises an interpolation routine adapted to interpolate pixel shift vectors for each pixel of the first image from the area shift vectors of adjacent spatial areas of the spatial matrix.

49. A medical imaging system, comprising:

a digital radiographic imaging system, comprising;

an x-ray device adapted to generate x-rays;

a collimator adapted to filter the x-rays in a desired anatomical region of a patient;

a flat-panel digital x-ray detector adapted to detect x-rays passing through the patient; and

dual-energy control circuitry adapted to acquire first and second images of the desired anatomical region at different energy levels over a time interval; and

an image processing system, comprising;

a pre-decomposition processing module adapted to register the first image to the second image using warping registration prior to decomposition; and

a dual-energy image decomposition module adapted to decompose soft tissue and bone images from the warpingly registered first and second images.

50. The medical imaging system of claim 49, wherein the pre-decomposition processing module comprises machine-readable code for spatially matching anatomical features in the first and second images.

51. The medical imaging system of claim 50, wherein the machine-readable code comprises an image warping routine operable to reduce motion variations between the first and second images.

52. The medical imaging system of claim 51, wherein the image warping routine comprises an edge-strength-based routine adapted to compute shift vectors to warp areas of the first image relative to the second image.

53. The medical imaging system of claim 52, wherein the edge-strength-based routine comprises a region of interest analysis routine adapted to decompose a sub-image from the first and second images for each shift vector of a plurality of possible shift vectors for each area of a spatial matrix for the first image, and adapted to compute an area shift vector for each area of the spatial matrix for the first image.

54. The medical imaging system of claim 53, wherein the edge-strength-based routine comprises an area vector selection routine adapted to select one shift vector of the plurality of possible shift vectors in a maximum shift search range for the area that provides a minimum total edge strength for the area.

55. A system for decomposing soft tissue and bone images from first and second energy images acquired by a digital radiography imaging system, comprising:

means for warpingly registering the first energy image to the second energy image prior to image decomposition; and

means for decomposing the soft tissue and bone images from the first and second energy images.

56. The system of claim 55, comprising means for acquiring the first and second energy images from the digital radiography imaging system.