Highly constrained image reconstruction method

• US 7,519,412 B2
• Filed: 07/07/2006
• Issued: 04/14/2009
• Est. Priority Date: 07/08/2005
• Status: Active Grant

• Current Assignee: Wisconsin Alumni Research Foundation (University of Wisconsin System)
• Original Assignee: Wisconsin Alumni Research Foundation (University of Wisconsin System)
• Inventors: Mistretta, Charles A.
• Primary Examiner(s): Winakur; Eric F
• Assistant Examiner(s): Rozanski; Michael T
• Application Number:
  US11/482,372
• Publication Number:
  US 20070010731A1
• Time in Patent Office: 1,012 Days
• Field of Search: 600/407, 600/410, 324/314, 378/4, 382128-133, 382/294
• US Class Current: 600/407
• CPC Class Codes: A61B 5/055 involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging
• CPC Class Codes: A61B 5/7257 using Fourier transforms
• CPC Class Codes: A61B 6/032 Transmission computed tomography [CT]
• CPC Class Codes: A61B 6/037 Emission tomography
• CPC Class Codes: A61B 6/4441 the rigid structure being a C-arm or U-arm
• CPC Class Codes: A61B 6/507 for determination of haemodynamic parameters, e.g. perfusion CT
• CPC Class Codes: G01R 33/4824 using a non-Cartesian trajectory
• CPC Class Codes: G01R 33/561 by reduction of the scanning time, i.e. fast acquiring systems, e.g. using echo-planar pulse sequences
• CPC Class Codes: G01R 33/5619 by temporal sharing of data, e.g. keyhole, block regional interpolation scheme for k-Space [BRISK]
• CPC Class Codes: G01R 33/56308 Characterization of motion or flow; Dynamic imaging
• CPC Class Codes: G06T 11/006 Inverse problem, transformation from projection-space into object-space, e.g. transform methods, back-projection, algebraic methods

Abstract

An image reconstruction method includes reconstructing a composite image of a subject using a conventional reconstruction method. The composite image employs the best information available regarding the subject of the scan and this information is used to constrain the reconstruction of highly undersampled image frames. An image frame is reconstructed using limited acquired data and image quality is improved using a priori information of the subject by weighting pixel values therein by values of corresponding pixels in the composite image.

Claims

1. A method for producing an image of a subject positioned in a field of view (FOV) of an imaging system, the steps comprising:

a) acquiring with the imaging system a plurality of views of the subject positioned in the FOV;

b) producing an image data set from views acquired in step a);

c) producing a composite image by combining views acquired in step a), wherein the views combined to produce the composite image includes the views used to produce the image data set and additional acquired views; and

d) reconstructing an image of the subject by normalizing the image data set with information derived from the composite image and by multiplying the normalized result with the composite image.

2. The method as recited in claim 1 in which the imaging system is a medical imaging system, the views acquired in step a) are projection views and step d) includes:

d)i) backprojecting projection views into the FOV and weighting the value backprojected into each image pixel by the composite image; and

d)ii) summing the backprojected values for each image pixel.

3. The method as recited in claim 2 in which each image pixel backprojected value S_n is calculated in step d)i) as

4. The method as recited in claim 2 in which step c) includes editing the composite image to remove an object therein and to thereby substantially minimize the appearance of that object in the reconstructed image.

5. The method as recited in claim 2 in which the weighting in step d)i) includes normalizing each projection view using a corresponding projection view from the composite image and multiplying the backprojected value by the value of the corresponding pixel in the composite image.

6. The method as recited in claim 5 in which the composite image is produced using substantially all the projection views acquired in step a).

7. The method as recited in claim 2 in which the FOV is three-dimensional, a three-dimensional image is produced, and the image I_(x,y,z) reconstructed in step c) is:

8. The method as recited in claim 2 in which the projection views are acquired in step a) by accumulating counts over a period of time, the projection views used in step c) to produce the composite image include counts accumulated over substantially the entire period of time, and the projection views used in step b) include counts accumulated over a portion of said period of time.

9. The method as recited in claim 8 in which the medical imaging system is a PET scanner and the counts represent the number of detected positron electron annihilation events.

10. The method as recited in claim 8 in which the medical imaging system is a SPECT scanner and the courts represent the number of detected photons.

11. The method as recited in claim 3 which includes:

e) acquiring with the medical imaging system additional views of the subject positioned in the FOV;

f) reconstructing a mask image from the additional views; and

g) subtracting the mask image from the composite image prior to performing step d).

12. The method as recited in claim 1 in which the views acquired in step a) are radiograph views repeatedly acquired at the same view angle.

13. The method as recited in claim 11 in which the composite image is reconstructed using a tomosynthesis reconstruction.

14. The method as recited in claim 1 in which the imaging system is a medical imaging system, and step d) includes:

d)i) producing a constrained image of the subject in the FOV by multiplying values in the image data set at each image pixel location by the corresponding pixel value in the composite image; and

d)ii) producing the image of the subject by normalizing the constrained image.

15. The method as recited in claim 14 in which the image data set is formed from views acquired in step a) that depict the subject in the FOV from a set of projection angles and the constrained image is normalized in step d)ii) by:

calculating a projection of the composite image at each of the projection angles; and

dividing each pixel value in the constrained image by a value in each of the composite image projections.

16. The method as recited in claim 15 in which the views acquired in step a) are projection views acquired at interleaved projection angles, the image data set produced in step b) includes producing the image data set from a set of said projection views, and the constrained image is normalized in step d)ii) by:

calculating a projection of the composite image at each projection angle of the set of said projection views; and

dividing each pixel value in the constrained image by a value in each of the composite image projections.

17. The method as recited in claim 14 in which the image data set produced in step b) includes:

selecting a set of views acquired in step a); and

transforming the selected views from acquisition space to real space.

18. The method as recited in claim 17 in which acquisition space is k-space and the transformation is an inverse Fourier transformation.

19. The method as recited in claim 18 in which the selected views are projection views and the transformation includes transforming the projection views to k-space.

20. The method as recited in claim 17 in which acquisition space is Radon space and the transformation is a backprojection of each view.

21. The method as recited in claim 15 in which the composite image is reconstructed from substantially all the views acquired in step a);

and the image data set is produced in step b) from substantially less than all the views acquired in step a).

22. The method as recited in claim 14 which includes:

d) producing additional images of the subject by repeating steps b) and d) using different ones of the views acquired in step a).

23. The method as recited in claim 22 in which the views acquired in step a) are projection views acquired at different view angles.

24. The method as recited in claim 14 in which the views acquired in step a) are acquired over a period of time during which changes occur in the subject;

step b) includes selecting a set of views acquired during a time interval during the performance of step a); and

step c) includes selecting a set of views acquired during a window of time that includes said time interval and is longer than said time interval.

25. The method as recited in claim 14 in which the medical imaging system is a magnetic resonance imaging system and each view samples a line in k-space.

26. The method as recited in claim 14 in which the medical imaging system is an x-ray CT system and each view is a projection in Radon space.

27. The method as recited in claim 14 in which the medical imaging system is an x-ray system and each view is a radiograph.

28. The method as recited in claim 1 in which the

composite image provides a priori knowledge of the subject to be imaged;

and

wherein the a priori knowledge of the subject in the composite image is used in the image reconstruction to improve the quality of the image.

29. The method as recited in claim 28 in which step d) includes:

d)i) backprojecting a selected view into the FOV; and

d)ii) weighting the value backprojected into each FOV image pixel based on the a priori knowledge of the subject at the image pixel.

30. The method as recited in claim 29 in which step d)ii) includes:

multiplying each backprojected image pixel value by a value in the composite image; and

normalizing the product of said multiplication.

31. The method as recited in claim 29 in which step d) includes Fourier transforming each selected view prior to backprojection in step d)i).

32. The method as recited in claim 28 in which step d) includes:

d)i) transforming the image data set to a real space image; and

d)ii) weighting the values of the real space image pixels based on the a priori knowledge of the subject at each image pixel.

33. The method as recited in claim 32 in which the transformation in step d)i) is an inverse Fourier transformation.

34. The method as recited in claim 1 in which the imaging system is a medical imaging system, and step d) includes:

d)i) normalizing the image data set by dividing values therein by values derived from the composite image;

d)ii) transforming the normalized image data set to form an unconstrained image; and

d)iii) multiplying values in the unconstrained image by corresponding values in the composite image.

35. The method as recited in claim 34 in which the values used in step d)i) are derived from the composite image by calculating a projection of the composite image for each view in the image data set.

36. The method as recited in claim 34 in which the views acquired in step a) are projection views acquired at interleaved projection angles, the image data set produced in step b) includes producing the image data set from a set of said projection views, and the image data set is normalized in step d)i) by:

calculating a projection of the composite image at each projection angle of the set of said projection views; and

dividing each value in the image data set by a value in each of the composite image projections.

37. The method as recited in claim 34 in which the composite image is reconstructed from substantially all the views acquired in step a);

and the image data set is produced in step b) from substantially less than all the views acquired in step a).

38. The method as recited in claim 35 which includes:

e) producing additional images of the subject by repeating steps b) and d) using different ones of the views acquired in step a).

39. The method as recited in claim 34 in which the views acquired in step a) are acquired over a period of time during which changes occur in the subject;

step b) includes selecting a set of views acquired during a time interval during the performance of step a); and

step c) includes selecting a set of views acquired during a window of time that includes said time interval and is longer than said time interval.

40. The method as recited in claim 34 in which the transformation in step d)ii) is a Fourier transformation.

41. The method as recited in claim 34 in which the transformation in step d)ii) includes:

Fourier transforming each normalized image data set view; and

reconstructing the unconstrained image from the Fourier transformed image data set views by performing one of either a two-dimensional or a three-dimensional inverse Fourier transformation thereon.

42. The method as recited in claim 35 in which step c) includes editing the composite image to remove selected structures.

43. The method as recited in claim 42 which includes reconstructing a mask image from selected views acquired in step a) and the editing of the composite image includes subtracting the mask image.