GEOMETRY CALIBRATION ALGORITHM FOR LARGE FLAT MODULE DETECTOR CT SCANNER

• US 20140211925A1
• Filed: 01/30/2014
• Published: 07/31/2014
• Est. Priority Date: 01/31/2013
• Status: Active Grant

• 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: DONG, Jiaqin, CAO, Ximiao, LI, Shuo, LI, Jun, WANG, Bin
• Granted Patent: US 9,398,890 B2
• Time in Patent Office: Days
• US Class Current: 378/207
• CPC Class Codes: A61B 6/032 Transmission computed tomography [CT]
• CPC Class Codes: A61B 6/583 using calibration phantoms
• CPC Class Codes: A61B 6/585 Calibration of detector units

Abstract

A method for geometric calibration of a CT scanner, including, for each row of at least one row of detector cells, establishing a complete geometric description of the CT scanner, including at least one unknown geometric parameter, establishing a description of a forward projection function using the complete geometric description, acquiring actual projection coordinates of a calibration phantom placed in a scanning field of view (SFOV) on a current row of detector cells and corresponding to a plurality of angles, acquiring calculated projection coordinates of the calibration phantom on the current row of detector cells and corresponding to the plurality of angles using the description of the forward projection function, and acquiring a calibrated value for the at least one unknown geometric parameter by evaluating the at least one unknown geometric parameter based on the acquired actual projection coordinates and calculated projection coordinates via a nonlinear least square fitting algorithm.

Claims

1. A method for geometric calibration of a CT scanner having at least one row of detector cells, the method comprising:

executing the following for each row of the at least one row of detector cells;

establishing a complete geometric description of the CT scanner, wherein the complete geometric description comprises at least one unknown geometric parameter;

establishing a description of a forward projection function using the complete geometric description;

acquiring actual projection coordinates of a calibration phantom placed in a scanning field of view (SFOV) on a current row of detector cells and corresponding to a plurality of angles;

acquiring calculated projection coordinates of the calibration phantom on the current row of detector cells and corresponding to the plurality of angles using the description of the forward projection function; and

acquiring a calibrated value for the at least one unknown geometric parameter by evaluating said at least one unknown geometric parameter based on the acquired actual projection coordinates and calculated projection coordinates using a nonlinear least square fitting algorithm.

2. The method according to claim 1, wherein the current row of detector cells has a structure of n×

q×

r, where n denotes the number of modules, q denotes the number of packs in each module, and r denotes the number of detector cells in each pack.

3. The method according to claim 2, wherein the at least one unknown geometric parameter comprises sizes of (n−

1) air gaps between adjacent modules, and an ISO channel offset.

4. The method according to claim 1, wherein the forward projection function is described as follows:

x_pβ

=P_β

(x,y,S),

wherein x_pβ

denotes a calculated projection coordinate of the calibration phantom on the current row of detector cells and corresponding to a forward projection angle β

, x and y are coordinates of the calibration phantom, S is a set comprising the at least one unknown geometric parameter, and P is a description of the forward projection function.

5. The method according to claim 4, wherein acquiring a calibrated value further comprises minimizing the following objective function by evaluating the at least one unknown geometric parameter using the nonlinear least square fitting algorithm:

F=Σ

_i=1^m(x′

_p(β

_i)−

P<sub>β

i</sub>(x,y,S))²,

wherein F denotes the objective function, β

_i denotes a current scan angle, m represents the number of scan angles, x′

_p(β

_i) represents an actual projection coordinate of the calibration phantom on the current row of detector cells and corresponding to the current scan angle, and P<sub>β

i</sub>(x, y, S) denotes a calculated projection coordinate of the calibration phantom on the current row of detector cells and corresponding to the current scan angle.

6. The method according to claim 5, wherein the actual projection coordinate of the calibration phantom on the current row of detector cells and corresponding to the current scan angle is the Gaussian center of projection data of the calibration phantom on the current row of detector cells and corresponding to the current scan angle.

7. The method according to claim 1, wherein the actual projection coordinates of the calibration phantom on the current row of detector cells and corresponding to the plurality of angles are obtained by an axial scan of the calibration phantom for the plurality of angles.

8. The method according to claim 1, wherein the nonlinear least square fitting algorithm includes the Powell optimization algorithm.

9. The method according to claim 1, wherein the calibration phantom is a pin having a length of 100 mm and a diameter of 2 mm.

10. The method according to claim 1, wherein the calibrated value for the at least one unknown geometric parameter is used for subsequent image reconstruction.

11. An apparatus for geometric calibration of a CT scanner having at least one row of detector cells, the apparatus comprising:

a geometric description establishing device configured to establish a complete geometric description of the CT scanner, the complete geometric description comprising at least one unknown geometric parameter;

a forward projection function establishing device configured to establish a description of a forward projection function using the complete geometric description;

an actual projection coordinate acquiring device configured to acquire actual projection coordinates of a calibration phantom placed in a scanning field of view (SFOV) on a current row of detector cells and corresponding to a plurality of angles;

a calculated projection coordinate acquiring device configured to acquire calculated projection coordinates of the calibration phantom on the current row of detector cells and corresponding to the plurality of angles using the description of the forward projection function; and

a calibrated value acquiring device configured for acquire a calibrated value for the at least one unknown geometric parameter by evaluating the at least one unknown geometric parameter based on the acquired actual projection coordinates and calculated projection coordinates via a nonlinear least square fitting algorithm.

12. The apparatus according to claim 11, wherein the current row of detector cells has a structure of n×

q×

r, where n denotes the number of modules, q denotes the number of packs in each module, and r denotes the number of detector cells in each pack.

13. The apparatus according to claim 12, wherein the at least one unknown geometric parameter includes sizes of (n−

1) air gaps between adjacent modules, and one ISO channel offset.

14. The apparatus according to claim 11, wherein the forward projection function is described as follows:

x_pβ

=P_β

(x,y,S),

wherein x_pβ

denotes a calculated projection coordinate of the calibration phantom on the current row of detector cells and corresponding to a forward projection angle β

, x and y are coordinates of the calibration phantom, S is a set comprising the at least one unknown geometric parameter, and P is a description of the forward projection function.

15. The apparatus according to claim 14, wherein the calibrated value acquiring device comprises an objective function minimizing part configured to minimize the following objective function by evaluating said at least one unknown geometric parameter via said nonlinear least square fitting algorithm:

F=Σ

_i=1^m(x′

_p(β

_i)−

P<sub>β

i</sub>(x,y,S))²,

wherein F denotes the objective function, β

_i denotes a current scan angle, m represents the number of scan angles, x′

_p(β

_i) represents an actual projection coordinate of the calibration phantom on the current row of detector cells and corresponding to the current scan angle, and P<sub>β

i</sub>(x, y, S) denotes a calculated projection coordinate of the calibration phantom on the current row of detector cells and corresponding to the current scan angle.

16. The apparatus according to claim 15, wherein the actual projection coordinate of the calibration phantom on the current row of detector cells and corresponding to the current scan angle is the Gaussian center of projection data of the calibration phantom on the current row of detector cells and corresponding to the current scan angle.

17. The apparatus according to claim 11, wherein the actual projection coordinates of the calibration phantom on the current row of detector cells and corresponding to the plurality of angles are obtained by an axial scan of the calibration phantom for the plurality of angles.

18. The apparatus according to claim 11, wherein the nonlinear least square fitting algorithm includes the Powell optimization algorithm.

19. The apparatus according to claim 11, wherein the calibrated value for the at least one unknown geometric parameter is used for subsequent image reconstruction.

20. A CT scan system, comprising:

an apparatus for geometric calibration of a CT scanner having at least one row of detector cells

the apparatus comprising;

a geometric description establishing device configured to establish a complete geometric description of the CT scanner, the complete geometric description comprising at least one unknown geometric parameter;

a forward projection function establishing device configured to establish a description of a forward projection function using the complete geometric description;

an actual projection coordinate acquiring device configured to acquire actual projection coordinates of a calibration phantom placed in a scanning field of view (SFOV) on a current row of detector cells and corresponding to a plurality of angles;

a calculated projection coordinate acquiring device configured to acquire calculated projection coordinates of the calibration phantom on the current row of detector cells and corresponding to the plurality of angles using the description of the forward projection function; and

a calibrated value acquiring device configured for acquire a calibrated value for the at least one unknown geometric parameter by evaluating the at least one unknown geometric parameter based on the acquired actual projection coordinates and calculated projection coordinates via a nonlinear least square fitting algorithm.