Method and system for automatically transforming CT studies to a common reference frame

1. A method for automatically transforming computed tomography (CT) studies to a common reference frame, the method comprising:

transforming selected CT studies to the common reference frame and generating a statistical atlas via a registration algorithm including a local refinement; and

establishing voxel-to-voxel correspondence between a CT study to be segmented and the statistical atlas via the registration algorithm including the local refinement.

2. The method of claim 1 , wherein the transforming selected CT studies to the common reference frame to generate the statistical atlas comprises:

selecting the CT studies;

manually segmenting the selected CT studies;

selecting a reference CT study;

selecting a volume of interest;

registering each of the selected CT studies against the reference CT study;

excluding misregistered studies;

transforming remaining selected CT studies to the common reference frame; and

generating the statistical atlas.

3. The method of claim 2 , wherein the selecting the reference CT study includes selecting one of the selected CT studies having:

quality and spatial resolution of a selected level relative to others of the selected CT studies; and

coverage of an entire region of the pelvis with no artificial objects in the entire region of the pelvis.

4. The method of claim 2 , wherein the selecting the volume of interest comprises manually selecting a local neighborhood of a pubic bone.

5. The method of claim 2 , wherein the excluding misregistered studies comprises:

identifying ones of the selected CT studies having large registration errors; and

excluding the ones of the selected CT studies having the large registration errors from further processing.

6. The method of claim 1 , wherein the establishing voxel-to-voxel correspondence between the CT study to be segmented and the statistical atlas comprises:

registering the CT study to be segmented against the reference CT study;

inverting a transfer function which registered the CT study to be segmented; and

transforming statistical data from the common reference frame to a coordinate system of the CT study to be segmented using the inverted transfer function.

7. The method of claim 1 . wherein the registration algorithm comprises:

performing global registration; and

performing the local refinement,

wherein both the performing global registration and the performing the local refinement include:

determining a similarity measure; and

optimizing the similarity measure such that a fit between the CT study to be segmented and a reference CT study is maximized.

8. The method of claim 7 , wherein the performing global registration comprises aligning bony structures between the CT study to be segmented and the reference CT study.

9. The method of claim 7 , wherein the performing local refinement comprises aligning bony structures and soft tissue structures between the CT study to be segmented and the reference CT study in a volume of interest.

10. The method of claim 7 , wherein the performing global registration comprises:

rotating an image around each of an x-axis, a y-axis that is substantially perpendicular to the x-axis, and a z-axis that is substantially perpendicular to an x-y plane;

translating the image along each of the x-axis, the y-axis, and the z-axis; and

scaling parameters of the image along each of the x-axis, the y-axis, and the z-axis.

11. The method of claim 7 , wherein the performing the local refinement comprises aligning rigid-body parameters using intensity information from a volume of interest.

12. The method of claim 7 , wherein the determining the similarity measure comprises evaluating a fit between the CT study to be segmented and the reference CT study by performing at least one of:

calculating a sum of squared differences;

calculating a cross correlation;

calculating metrics based on mutual information; and

calculating metrics based on normalized mutual information.

13. The method of claim 12 , wherein the optimizing the similarity measure comprises:

initializing parameters; and

employing a general nonlinear optimization method to systematically change the parameters to obtain an optimum similarity measure.

14. A system for automatically transforming computed tomography (CT) studies to a common reference frame, the system comprising:

an imaging system configured to generate a set of slices comprising image data of an object; and

a controller configured to:

transform selected CT studies to the common reference frame to generate a statistical atlas via a registration algorithm including a local refinement; and

establish voxel-to-voxel correspondence between a CT study to be segmented and the statistical atlas via the registration algorithm including the local refinement.

15. The system of claim 14 , wherein the registration algorithm further comprises a global registration, and wherein both the global registration and the local refinement include:

a similarity measure; and

an optimization of the similarity measure such that a fit between the CT study to be segmented and a reference CT study is maximized.

16. The system of claim 15 , wherein the optimization of the similarity measure comprises evaluation of the fit between the CT study to be segmented and the reference CT study using at least one of:

a sum of squared differences;

a cross correlation;

metrics based on mutual information; and

metrics based on normalized mutual information.

17. A computer program storage medium readable by a processing circuit, storing instructions for execution by the processing circuit for automatically segmenting image data of an object within a region of interest, and for storing instructions for execution by the processing circuit for:

transforming selected computed tomography (CT) studies to a common reference frame to generate a statistical atlas via a registration algorithm including a local refinement; and

establishing voxel-to-voxel correspondence between a CT study to be segmented and the statistical atlas via the registration algorithm including the local refinement.

18. The computer program storage medium of claim 17 , wherein the storage medium stores further instructions for:

selecting the CT studies;

manually segmenting the selected CT studies;

selecting a reference CT study;

selecting a volume of interest;

registering each of the selected CT studies against the reference CT study;

excluding misregistered studies;

transforming remaining selected CT studies to the common reference frame; and

generating the statistical atlas.

19. The computer program storage medium of claim 17 , wherein the storage medium stores further instructions for:

registering the CT study to be segmented against the reference CT study;

inverting a transfer function which registered the CT study to be segmented; and

transforming statistical data from the common reference frame to a coordinate.

20. The computer program storage medium of claim 17 , wherein the registration algorithm comprises:

performing global registration; and

performing the local refinement,

wherein both the performing global registration and the performing the local refinement include:

determining a similarity measure; and

optimizing the similarity measure such that a fit between the CT study to be segmented and a reference CT study is maximized.

21. The computer program storage medium of claim 20 , wherein the performing global registration comprises aligning bony structures between the CT study to be segmented and the reference CT study.

22. The computer program storage medium of claim 20 , wherein the performing local refinement comprises aligning bony structures and soft tissue structures between the CT study to be segmented and the reference CT study in a volume of interest.

23. The computer program storage medium of claim 17 , wherein the performing the local refinement comprises aligning rigid-body parameters using intensity information from a volume of interest.