Systems, methods and apparatus for consistency-constrained filtered backprojection for out-of-focus artifacts in digital tomosythesis

• US 7,873,236 B2
• Filed: 08/28/2007
• Issued: 01/18/2011
• Est. Priority Date: 08/28/2007
• Status: Active Grant

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: Li, Baojun, Avinash, Gopal B.
• Primary Examiner(s): Kassa; Yosef
• Application Number:
  US11/846,009
• Publication Number:
  US 20090060310A1
• Time in Patent Office: 1,239 Days
• Field of Search: 382/260, 382/274, 382/275
• US Class Current: 382/275
• CPC Class Codes: G06T 11/006 Inverse problem, transformation from projection-space into object-space, e.g. transform methods, back-projection, algebraic methods
• CPC Class Codes: G06T 2211/421 Filtered back projection [FBP]

Abstract

Systems, methods and apparatus are provided through which in some embodiments during backprojection of filtered two-dimensional into three dimensional image each voxel is analyzed for inconsistent contributions based on at least two criteria and each of the inconsistent contributions is either replaced or excluded from the backprojection.

Claims

1. A method to suppress out-of-focus artifacts during backprojection of a plurality of two-dimensional projections, wherein the backprojecting further comprises for each voxel in a reconstructed three-dimensional image the method comprising:

storing a plurality of backprojected contributions in a memory;

sorting the plurality of backprojected contributions by value into a monotonic order in the memory;

identifying at least one contribution of the sorted contributions in the memory that is significantly different than other sorted contributions in reference to at least one criteria of significance;

replacing each of the at least one identified contribution in the memory with a value selected from at least one of a group of values; and

combining the replaced contributions in the memory by an operator.

2. The method of claim 1, wherein the criteria of significance includes at least two of:

an intensity of each of the at least one contribution being more than some multiple C of the standard deviation from a mean of the plurality of sorted contributions;

an intensity of each of the at least one contribution being within the top or the bottom P percentiles range of all of the plurality of sorted contributions;

or

an absolute distance between an intensity of each of the at least one contribution to the mean of the plurality of sorted contributions that is more than a value K.

3. The method of claim 2, wherein the multiple C of the standard deviation from a mean of the at least one contribution further comprise:

about more than 1.0 multiples of the standard deviation from a mean of the at least one contribution.

4. The method of claim 2, wherein the P percentiles range of all of the at least one contribution further comprise:

about a 5% percentile range of all of the at least one contribution.

5. The method of claim 2, wherein the K absolute distance between an intensity of each of the at least one contribution to the mean of the at least one contribution further comprise:

a range of about 25 to about 200 absolute distance between an intensity of each of the at least one contribution to the mean of the at least one contribution.

6. The method of claim 1, wherein the group of values consist of:

a mean of the at least one contribution other than the each of the at least one identified contribution;

a median of the at least one contribution other than the each of the at least one identified contribution;

another statistic of the at least one contribution other than the each of the at least one identified contribution; and

a mean of the at least one contribution other than the at least one identified contribution.

7. The method of claim 1 further comprising:

accessing the plurality of two-dimensional projections of an object being imaged from a plurality of angles; and

enhancing structures represented in the plurality of two-dimensional projections in the memory.

8. The method of claim 7, wherein enhancing the structures that are represented in the plurality of two-dimensional projections further comprises:

filtering each of the plurality of two-dimensional projections using a filter.

9. The method of claim 1, wherein the operator includes maximum, minimum, average, or median.

10. The method of claim 1, wherein the method is performed iteratively.

11. A computer-accessible medium having executable instructions that are capable of directing a processor to suppress out-of-focus artifacts during backprojection of a plurality of two-dimensional projections, wherein for each voxel in a reconstructed three-dimensional image, the executable instructions perform:

storing a plurality of backprojected contributions in a memory;

sorting the plurality of backprojected contributions by value into a monotonic order in the memory;

identifying at least one contribution of the sorted contributions in the memory that is significantly different than other sorted contributions in reference to at least one criteria of significance, wherein the criteria includes;

an intensity of each of the at least one contribution being more than some multiple C of the standard deviation from a mean of the plurality of sorted contributions;

an intensity of each of the at least one contribution being within the top or the bottom P percentiles range of all of the plurality of sorted contributions;

or

an absolute distance between an intensity of each of the at least one contribution to the mean of the plurality of sorted contributions that is more than a value K; and

replacing each of the at least one identified contribution in the memory with a value selected from at least one of a group of values consisting of;

a mean of the at least one contribution other than the each of the at least one identified contribution;

a median of the at least one contribution other than the each of the at least one identified contribution;

another statistic of the at least one contribution other than the each of the at least one identified contribution;

a mean of the at least one contribution other than the at least one identified contribution; and

combining the updated contributions in the memory by an operator, wherein the operator includes maximum, minimum, average, or median.

12. The computer-accessible medium of claim 11, wherein the three criteria further comprise:

an intensity of each of the at least one contribution being more than some multiple C of the standard deviation from a mean of the at least one contribution;

an intensity of each of the at least one contribution being within the top or the bottom P percentiles range of all of the at least one contribution; and

an absolute distance between an intensity of each of the at least one contribution to the mean of the at least one contribution that is more than a value K.

13. The computer-accessible medium of claim 12, wherein the multiple C of the standard deviation from a mean of the at least one contribution further comprise:

about 1.0 multiples of the standard deviation from a mean of the at least one contribution.

14. The computer-accessible medium of claim 12, wherein the P percentiles range of all of the at least one contribution further comprise:

about 5% percentile range of all of the at least one contribution.

15. The computer-accessible medium of claim 12, wherein the K absolute distance between an intensity of each of the at least one contribution to the mean of the at least one contribution further comprise:

a range of about 25 to about 200 absolute distance between an intensity of each of the at least one contribution to the mean of the at least one contribution.

16. The computer-accessible medium of claim 11, wherein the value in the replacing action is selected from the group consisting of:

a mean of the at least one contribution other than the each of the at least one identified contribution;

a median of the at least one contribution other than the each of the at least one identified contribution;

another statistic of the at least one contribution other than the each of the at least one identified contribution; and

a mean of the at least one contribution other than the at least one identified contribution.

17. The computer-accessible medium of claim 11, wherein the executable instructions further comprise executable instructions:

accessing a plurality of two-dimensional projections of an object being imaged from a plurality of angles.

18. A system comprising:

a processor;

a storage device coupled to the processor; and

apparatus operative on the processor to;

receive into the storage device, a plurality of two-dimensional projections of an object from a plurality of angles;

generate into the storage device, three-dimensional images by backprojecting an enhanced plurality of two-dimensional projections, wherein the backprojecting further comprises for each voxel in a reconstructed three-dimensional image;

storing all backprojected contributions from the plurality of two-dimensional projections in the storage device;

sorting the contributions in the storage device by their values into a monotonic order;

identifying outlier contributions in the storage device;

replacing the outlier contributions in the storage device with a value selected from at least one of a group of values; and

combining in the storage device the replaced contributions in the memory by an operator, wherein the operator includes maximum, minimum, average, or median.

19. The system of claim 18 wherein the apparatus is further operable to:

identify at least one contribution that is significantly different than other contributions in reference to at least one criteria of significance.

20. The system of claim 18 wherein the criteria includes:

an intensity of each of the at least one contribution being more than 1.0 multiples of the standard deviation from a mean of the at least one contribution;

an intensity of each of the at least one contribution being within the top 5% or the bottom 5% percentile range of all of the at least one contribution;

or

an absolute distance between an intensity of each of the at least one contribution to the mean of the at least one contribution that is more than a range of about 25 to about 200.

21. The system of claim 18 wherein the apparatus further comprises software apparatus.

22. The system of claim 18, further comprising apparatus to enhance structures represented in the plurality of two-dimensional projections further comprising:

a filter operable on each of the plurality of two-dimensional projections.

23. The system of claim 18, wherein the apparatus to replace outlier contributions further comprises apparatus to:

replace each of at least one identified contribution in the storage device with a value of that is significantly different than other contributions in reference to at least one criteria of significance, wherein the value is selected from the group consisting of;

a mean of the at least one contribution other than the each of the at least one identified contribution;

a median of the at least one contribution other than the each of the at least one identified contribution;

another statistic of the at least one contribution other than the each of the at least one identified contribution; and

a mean of the at least one contribution other than the at least one identified contribution.