Segmentation and Spectrum Based Metal Artifact Reduction Method and System

• US 20170109904A1
• Filed: 10/28/2016
• Published: 04/20/2017
• Est. Priority Date: 04/29/2014
• Status: Active Grant

• Current Assignee: Carl Zeiss X-ray Microscopy, Inc. (Carl-Zeiss-Stiftung)
• Original Assignee: Carl Zeiss X-ray Microscopy, Inc. (Carl-Zeiss-Stiftung)
• Inventors: Huang, Zhifeng, Case, Thomas A.
• Granted Patent: US 10,297,048 B2
• Time in Patent Office: Days
• CPC Class Codes: G06T 11/008 Specific post-processing after tomographic reconstruction, e.g. voxelisation, metal artifact correction
• CPC Class Codes: G06T 2211/40 Computed tomography
• CPC Class Codes: G06T 2211/408 Dual energy
• CPC Class Codes: G06T 5/70 Denoising; Smoothing
• CPC Class Codes: H04N 5/32 Transforming X-rays cameras or camera modules for generating image signals from X-rays H04N23/30; circuitry of SSIS for transforming X-rays into image signals H04N25/30

Abstract

A segmentation-and-spectrum-based metal artifact reduction (MAR) system and method is applied in polychromatic X-ray CT system that uses a priori knowledge of high-Z metals in samples which contribute the primary artifacts at a known x-ray energy spectrum. Using a basis materials decomposition, the method solves the problem of reducing or eliminating metal artifacts associated with beam hardening using only a single scan of the sample performed at selected x-ray energy.

Claims

1. A data acquisition and image reconstruction method for an x-ray computed tomography system, the method comprising:

creating a set of baseline projections of a sample;

creating a baseline tomographic volume data set of the sample from the set of baseline projections;

segmenting high-Z structures from the baseline tomographic volume data set to create a segmented high-Z volume data set;

generating a set of high-Z projections from the segmented high-Z volume data set;

generating a set of equivalent low-Z projections using the set of high-Z projections taking into account an energy spectrum of an x-ray beam; and

generating a corrected tomographic volume data set using the set of equivalent low-Z projections.

2. A method as claimed in claim 1, wherein the set of baseline projections of the sample are generated by rotating the sample relative to the X-ray beam from an x-ray source.

3. A method as claimed in claim 2, wherein the x-ray beam is a polychromatic beam.

4. A method as claimed in claim 2, wherein the x-ray source is a laboratory x-ray source.

5. A method as claimed in claim 2, wherein the x-ray source is a white synchrotron beam or accelerator-based source.

6. A method as claimed in claim 1, wherein the equivalent low-Z projections in the set of equivalent low-Z projections provide information associated with a thickness of an equivalent low-Z structure.

7. A method as claimed in claim 1, further comprising generating the set of high-Z projections by performing forward projection of the segmented high-Z volume data set.

8. A method as claimed in claim 7, further comprising performing a blur of the set of high-Z projections to match the CT system resolution limit.

9. A method as claimed in claim 1, further comprising fusing an equivalent low-Z tomographic volume data set created from the set of equivalent low-Z projections with the segmented high-Z metal volume data set to produce the corrected tomographic volume data set.

10. A method as claimed in claim 1, further comprising generating the segmented high-Z volume data set by identifying a high-Z metal of the high-Z structures of the sample that contributes metal artifacts, and selecting thresholds for segmenting the high-Z metal from the baseline tomographic volume data set.

11. A method as claimed in claim 1, further comprising using N-order polynomial fitting beam-hardening corrected (BHC) projections as baseline projections and related BHC reconstructed tomographic volume data set as the baseline tomographic volume data set.

12. A method as claimed in claim 1, wherein the set of equivalent low-Z element projections is generated from the set of high-Z projections by a basis materials decomposition procedure according to a relationship among the baseline projections and basis materials.

13. A method as claimed in claim 12, wherein the basis materials decomposition procedure comprises:

identifying a high-Z metal of the sample contributing metal artifacts and an equivalent low-Z element representing other elements of the sample as the basis materials, the segmentation being generated from the high-Z metal;

identifying X-ray attenuations of the equivalent low-Z element and the high-Z metal at the energy spectrum.

14. A method as claimed in claim 13, wherein, in the basis materials decomposition procedure, the basis materials consist of one high-Z metal and one equivalent low-Z element, or consist of two high-Z metals and one equivalent low-Z element, or consist of more than 2 high-Z metals and one equivalent low-Z element.

15. A method as claimed in claim 12, further comprising using a Look Up Table (LUT) to determine the relationship among the basis materials.

16. A method as claimed in claim 12, wherein the relationship includes equivalent low-Z element(s) and the high-Z metal(s) and the X-ray attenuations of the equivalent low-Z element(s) and the high-Z metal(s) at the energy spectrum.

17. A method as claimed in claim 12, further comprising using the relationship among the basis materials and X-ray attenuations of the equivalent low-Z element and high-Z metal at the energy spectrum, and using the baseline projections and forward-projected high-Z projections for calculating thickness of the equivalent low-Z element to be the equivalent low-Z element projections.

18. A method as claimed in claim 1, further comprising identifying new artifacts introduced during the generation of the set of equivalent low-Z projections, and reducing the new artifacts in the set of equivalent low-Z projections.

19. A method as claimed in claim 1, further comprising:

filtering the equivalent low-Z projections by removing sharp edges and abnormal negative values from the equivalent low-Z projections; and

reconstructing an equivalent low-Z tomographic volume data set from the filtered set of equivalent low-Z projections.

20. A method as claimed in claim 1, further comprising improving image quality in the corrected tomographic volume data set by an iterative operation, comprising:

creating a new forward-projected set of equivalent low-Z projections by performing forward projection based on a corrected equivalent low-Z tomographic volume data set generated from the set of equivalent low-Z projections;

identifying portions of the sample shadowed by high-Z structures of the sample;

creating a set of integrated low-Z projections by performing an integration on the new forward-projected set of equivalent low-Z element projections and the filtered set of equivalent low-Z projections for the identified portions of the sample;

reconstructing a new equivalent low-Z projections tomographic volume data set; and

creating a new corrected tomographic volume data set by fusing the new equivalent low-Z tomographic volume data set and the segmented high-Z volume data set.

21. A method as claimed in claim 1, further comprising improving image quality in unexpected artifacts in upper/lower slices of reconstructed metal objects due to insufficient pixels of the metal in basis materials decomposition by identifying locations of metal edges in the upper/lower slices and creating a new metal-artifact-corrected tomographic volume data set by blending fused metal-artifact-corrected tomographic volume data set with the baseline tomographic volume data set in these identified slices.

30. A method as claimed in claim 1, further comprising:

identifying locations of high-Z edges in upper and lower slices within the baseline tomographic volume data set;

creating an equivalent low-Z volume data set from the equivalent low-Z projections;

fusing the equivalent low-Z volume data set with the segmented high-Z volume data set to create a fused tomographic volume data set; and

blending the fused tomographic volume data set and the baseline tomographic volume data set at the locations of the high-Z edges in the upper and lower slices within the baseline tomographic volume data set to create a blended tomographic volume data set as the corrected tomographic volume data set.

31. A method as claimed in claim 1, further comprising:

generating a set of high-Z thickness projections based on a thickness of the high-Z structures of the segmented high-Z volume data set;

generating a set of x-ray attenuated projections based on x-ray attenuation of the high-Z structures of the segmented high-Z volume data set;

creating a set of normalized high-Z projections by normalizing the set of x-ray attenuated projections with the set of high-Z thickness projections;

fusing the equivalent low-Z projections with the normalized high-Z projections, based upon a selected monochromatic x-ray energy of the energy spectrum of the x-ray beam, to create a set of monochromatic fused projections at the selected x-ray energy; and

reconstructing the corrected tomographic volume data set from the monochromatic fused projections.

22. An x-ray CT system, comprising:

an x-ray imaging system that generates a set of baseline projections by rotating a sample relative in an x-ray beam; and

a computer system that generates a baseline tomographic volume data set of the sample from the set of baseline projections, segments high-Z structures from the baseline tomographic volume data set to create a segmented high-Z volume data set, generates a set of high-Z projections from the segmented high-Z volume data set, generates a set of equivalent low-Z projections using the set of high-Z projections taking into account an energy spectrum of an x-ray beam, and generates a corrected tomographic volume data set using the equivalent low-Z element projections.

23. A system as claimed in claim 22, wherein the x-ray imaging system comprises an x-ray source that generates the x-ray beam, a rotation stage that rotates the sample in the x-ray beam and a detector system for detecting the x-ray beam after interaction with the sample.

24. A system as claimed in claim 23, wherein the x-ray beam is a polychromatic beam.

25. A system as claimed in claim 23, wherein the x-ray source is a laboratory x-ray source.

26. A system as claimed in claim 23, wherein the x-ray source is a white synchrotron beams or accelerator-based source.

27. A system as claimed in claim 22, wherein the equivalent low-Z projections in the set of equivalent low-Z projections provide information associated with a thickness of the equivalent low-Z structure at each projection angle formed when rotating the sample in the x-ray beam.

28. A system as claimed in claim 22, wherein the computer system further creates a segmented high-Z volume data set of the sample from the baseline tomographic volume data set.

29. A system as claimed in claim 28, wherein the computer system generates a set of high-Z projections by performing forward projection of the segmented high-Z metal volume data set.