Automatic misregistration correction

• US 4,685,146 A
• Filed: 05/29/1984
• Issued: 08/04/1987
• Est. Priority Date: 07/26/1983
• Status: Expired due to Term

• Current Assignee: Elscint Limited (Elbit Imaging Limited)
• Original Assignee: Elscint Limited (Elbit Imaging Limited)
• Inventors: Fenster, Paul, Shimoni, Yair
• Primary Examiner(s): Boudreau, Leo H.
• Assistant Examiner(s): Mancuso, Joseph
• Application Number:
  US06/614,840
• Time in Patent Office: 1,162 Days
• Field of Search: 382/6, 382/30, 382/54, 382/41, 382/44, 364/414, 378/4, 378/21, 378/62, 378/901
• US Class Current: 382/294
• CPC Class Codes: G06T 5/80 Geometric correction

Abstract

A method and system for determining the movement of an object between different images of the same object; where the movement is defined by a vector value and the vector value is used to correct for misregistration between the different images.

Claims

1. A method for correcting for artifacts generated by object motion caused misregistration between different images of the same object,

said method comprising of:

obtaining data for first and second images of the same object, which may have normally changing and normally non-changing portions;

selecting a point on a normally non-changing portion;

locating said point in each of said images;

the step of locating the point including the steps of;

directing a Region-of-Interest (ROI) to encompass a normally non-changing portion in each of said first and second images;

determining the density in the ROI in each of said first and second images,

determining the density gradients in the ROI in each of said first and second images,

determining the amplitudes of the density gradients in the ROI in each of said first and second images,

transforming said amplitudes of the density gradients in the ROI in each of said first and second images to quantities that are positive, and

determining as the point in each of said first and second images the centers of gravity (COGs) of said quantities within the ROI in each of said first and second images,

subtracting coordinates of the COG of one of said images from coordinates of the COG of the other of said images to provide the vector extending between the point in each of said first and second images, and

using the vector to move said first and second images relative to each other a distance and direction as determined by said vector.

2. The method of claim 1 wherein said different images are temporally distinguished.

3. The method of claim 1 wherein said different images are energy distinguished.

4. The method of claim 1 wherein the step of directing an ROI includes the steps of:

(a) comparing said first image to said second image to obtain a third image comprising normally changing portions of the object and motion artifacts,

(b) locating an ROI on one of the motion artifacts and

(c) using the same ROI on said first and said second images.

5. The method of claim 1 including the step of smoothing said images.

6. The method of claim 5 wherein smoothing is accomplished by filtering the images to reduce noise.

7. The method of claim 1 wherein said normally changing portions of the object include vascular sections having contrast material entering therein.

8. The method of claim 1 wherein said normally non-changing portion comprises bone.

9. The method of claim 1 wherein said first image comprises a combination of images.

10. The method of claim 1 wherein the step of using the vector value includes using the vector value to re-register said first and second images to minimize artifacts generated by object motion in images obtained by adding said first and second images of the same object.

11. The method of claim 1 wherein the step of using the vector value includes using the vector value to minimize artifacts generated by object motion in images obtained by merging said first and second images of the same object.

12. The method of claim 1 wherein the step of using the vector value includes using the vector value to minimize artifacts generated by object motion in images obtained by subtracting said first and second images.

13. The method of claim 1 wherein one of said different images comprises a combination of images.

14. A system for correcting for artifacts generated by object motion caused misregistration between different images of the same object, said system comprising:

means for obtaining density data for first and second images of the same object, which may have normally changing and normally non-changing items,

means for selecting a point on a normally non-changing item,

means for locating said point in each of said images, said means for locating said point comprising

means for obtaining density gradients from the density data in a Region of Interest (ROI) similarily placed on both said first and second images

means for determining the amplitudes of the density gradients from the density gradients in the ROI similarily placed on both said first and second images,

means for transforming the said amplitudes of the density gradients to quantities that are positive,

means for determining coordinates of the COGs of said positive quantities within the ROI'"'"'s, and

means for subtracting said coordinates to provide the coordinate components of a vector extending between the points defined by the centers of gravity, and

means for using the vector to move the first and second images relative to each other an amount and direction determined by the vector.

15. The system of claim 14 wherein said second image comprises a combination of images.

16. The system of claim 14 wherein said first and second images are temporally distinguished.

17. The system of claim 14 wherein said first and second images are energy distinguished.

18. The system of claim 14 wherein said means for using the vector value comprises means to minimize the artifacts generated by the object motion when said first and second images are compared.

19. The system of claim 14, wherein the means for locating the points comprise:

means for directing an ROI to encompass a normally non-changing item in each of said images, and

means for determining the location of a point in the ROI in each of said images.

20. The system of claim 19 wherein the means for directing an ROI includes:

means for subtracting said first image from said second image to obtain a subtracted image, said subtracted image comprising normally changing items normally unchanging items and motion artifacts,

means for locating an ROI on one of said motion artifacts, and

means for using the same ROI on said first and second images.

21. The system of claim 14 including means for smoothing said images.

22. The system of claim 21 wherein said smoothing means comprises filter means for filtering out noise.

23. The system of claim 14 wherein said normally changing items include vascular sections having contrast material entering therein.

24. The system of claim 14 wherein said normally non-changing items comprise bone.

25. The system of claim 14 wherein said first image comprises a combination of images.

26. The system of claim 14 wherein said means for using said vector value includes:

means for minimizing artifacts generated by said object motion in images obtained by adding said first and second images.

27. The method of claim 14 wherein the step of transforming the amplitudes to positive quantities includes the step of squaring the amplitudes.

28. The system of claim 14 wherein said means for transforming said amplitudes to positive quantities include means for squaring the amplitudes.

29. The system of claim 14 including means for moving the images relative to each other using the determined vector value.

30. A method of correcting artifacts generated by object motion caused misregistration between first and second images of the same object, said artifacts appearing on a third image made by merging said first and second images, said method comprising the steps of:

obtaining data for first and second images of the same object, which may have normally changing and normally non-changing items;

selecting a point on a normally non-changing item,

locating said point in each of said images;

the step of locating said point on each of said images comprising the steps of;

locating an artifact on the third image,

selecting a portion of the third image which includes the located artifact,

selecting corresponding portions of said first and second images,

obtaining the density of said corresponding portions of said first and second image,

obtaining the gradients of the obtained density,

obtaining the amplitudes of density gradients of the corresponding

portions of the first and second images,

transforming the said amplitudes of the density gradients to quantitites that are all positive, and

and determining the centers of gravity of the positive amplitudes within the corresponding portions,

determining differences between the locations of the COGs in each of said first and second images to obtain a vector of the motion causing the misregistration, and using said vector to correct for the misregistration caused by object motion between the acquisition of said first and second images.

31. The method of claim 30 further comprising a noise reducing step.

32. The method of claim 31 wherein said noise reducing step includes setting relatively low values of the gradient to zero.

33. A system for correcting for artifacts generated by object motion caused misregistration between different images of the same object, said system comprising:

means for obtaining transmittance data for first and second images of the same object, which may have items normally changing in transmittance and items normally non-changing in transmittance, said artifacts appearing on a third image comprised of a merging of said first and second images,

means for selecting a point in a normally non-changing item,

means for locating said point in each of said images,

said means for locating said point in each of said images comprising;

means for selecting a portion of the third images, said portion containing an artifact,

means for obtaining gradients of said transmittance data in portions of said first and second images corresponding to said selected portion of said third image,

means for obtaining amplitude of the gradients of said transmittance data in the corresponding portions of the first and second images,

means for transforming the obtained amplitudes of the gradients to quantities that are all positive,

means for reducing the effect of image noise,

means for determining centers of gravity (COGs) of said positive amplitudes in regions around said normally non-changing items in the corresponding portions in each of said first and second images,

means for determining coordinate differences between the locations of the COGs in each of said images to obtain vector coordinate, of the vector of the motion causing the misregistration, and

means for utilizing the vector for moving said first image relative to the second image the amount of the vector in both amplitude and direction to correct for the misregistration.

34. The system of claim 33 wherein said noise reducing means includes a lower thresholding means.