Cone-beam CT imaging scheme

• US 8,144,829 B2
• Filed: 02/27/2009
• Issued: 03/27/2012
• Est. Priority Date: 02/27/2008
• Status: Active Grant

• Current Assignee: Board of Trustees of the Leland Stanford Junior University (Stanford Management Co.)
• Original Assignee: Board of Trustees of the Leland Stanford Junior University (Stanford Management Co.)
• Inventors: Zhu, Lei, Xing, Lei
• Primary Examiner(s): Glick, Edward
• Assistant Examiner(s): CORBETT, JOHN M
• Application Number:
  US12/380,585
• Publication Number:
  US 20090225932A1
• Time in Patent Office: 1,124 Days
• Field of Search: 378/7, 382/131
• US Class Current: 378/7
• CPC Class Codes: A61B 6/032 Transmission computed tomography [CT]
• CPC Class Codes: A61B 6/466 adapted to display 3D data
• CPC Class Codes: A61B 6/542 involving control of exposure
• CPC Class Codes: A61N 5/103 Treatment planning systems
• CPC Class Codes: G06T 11/005 Specific pre-processing for tomographic reconstruction, e.g. calibration, source positioning, rebinning, scatter correction, retrospective gating
• CPC Class Codes: G21K 1/02 using diaphragms, collimators

Abstract

A general imaging scheme is proposed for applications of CBCT. The approach provides a superior CBCT image quality by effective scatter correction and noise reduction. Specifically, in its implementation of CBCT imaging for radiation therapy, the proposed approach achieves an accurate patient setup using a partially blocked CBCT with a significantly reduced radiation dose. The image quality improvement due to the proposed scatter correction and noise reduction also makes CBCT-based dose calculation a viable solution to adaptive treatment planning.

Claims

1. A method for medical imaging comprising:

a) performing a multi-detector computed tomography (MDCT) scan of a patient to produce MDCT scan data;

b) reconstructing MDCT image slices from the MDCT scan data;

c) performing a cone-beam computed tomography (CBCT) scan of the patient using a CBCT beam source to produce CBCT scan data, wherein performing the CBCT scan comprises positioning a patterned lead sheet between the CBCT beam source and the patient to partially block x-rays from the CBCT beam source, such that the CBCT scan data is produced from a partially blocked CBCT beam;

d) reconstructing CBCT image slices from the CBCT scan data;

e) performing 3D rigid registration of the MDCT image slices and the CBCT image slices to obtain fused reconstructed images of the patient;

wherein the method further comprises;

storing a scatter distribution for the patient; and

correcting for scatter by using scatter estimates derived from the scatter distribution together with a kernel-based scatter model.

2. The method of claim 1 wherein the patterned lead sheet reduces x-rays reaching the patient from the CBCT beam source by a factor of at least six.

3. The method of claim 2 wherein the patterned lead sheet comprises rectangular lead strips arranged in parallel to produce a pattern of parallel slits through which x-rays may pass to produce a sequence of parallel illuminated slices corresponding to gaps between the strips.

4. The method of claim 3 wherein the rectangular lead strips are sufficiently thick to completely block x-rays from passing through the strips.

5. The method of claim 1 wherein the patterned lead sheet comprises an array of holes.

6. The method of claim 1 wherein the patterned lead sheet comprises an array of blocking elements.

7. The method of claim 1 further comprising deriving the scatter distribution for the patient from the CBCT scan data produced from the partially blocked CBCT beam.

8. The method of claim 7 wherein the scatter distribution for the patient is derived from measured x-ray signals in shadow regions of the CBCT scan data produced from the partially blocked CBCT beam.

9. The method of claim 1 further comprising deriving the scatter distribution for the patient from the MDCT scan data.

10. The method of claim 9 wherein the scatter distribution for the patient is derived by converting Hounsfield unit CT values of the MDCT scan data into electron density, and then performing a Monte Carlo (MC) simulation of the MDCT scan data to obtain the scatter distribution.

11. The method of claim 1 further comprising correcting for scatter in an unblocked CBCT scan of the patient by registering the MDCT scan data with an unblocked CBCT reconstruction, modifying the scatter distribution according to the registration, and subtracting the modified scatter distribution from unblocked CBCT scan data.

12. The method of claim 1 further comprising correcting for scatter in an unblocked CBCT scan of the patient by registering the reconstructed CBCT image slices from the CBCT scan data produced from the partially blocked CBCT beam with an unblocked CBCT reconstruction, modifying the scatter distribution according to the registration, and subtracting the modified scatter distribution from unblocked CBCT scan data.

13. The method of claim 1 wherein reconstructing CBCT image slices comprises correcting for scatter by reconstructing the CBCT image slices without scatter correction, registering the reconstructed CBCT image slices with the MDCT image slices, and performing a Monte Carlo simulation on the MDCT image slices to estimate and correct for scatter in the reconstructed CBCT image slices.

14. The method of claim 1 wherein reconstructing CBCT image slices comprises correcting for scatter, and reducing noise introduced by the correcting for scatter.

15. The method of claim 14 wherein reducing the noise comprises applying a penalized weighted least-squares (PWLS) technique to CBCT image projection data.

16. The method of claim 15 wherein the PWLS technique uses a cost function involving a variance of scatter-corrected line integral data q_c given by var(q_c)=p_m/(p_m−

s_e)², where p_m is measured image projection data and s_e is a scatter estimate.