Systems and Methods for Processing Medical Image Data to Facilitate Comparisons Among Groups of Subjects

• US 20080298659A1
• Filed: 05/31/2007
• Published: 12/04/2008
• Est. Priority Date: 05/31/2007
• Status: Active Grant

• Current Assignee: Southern Methodist University, Board of Regents of the University of Texas System (University of Texas System)
• Original Assignee: Board of Regents of the University of Texas System (University of Texas System)
• Inventors: Lin, Qihua, Carmack, Patrick S., Spence, Jeffrey S., Schucany, William R., Haley, Robert W., Gunst, Richard F.
• Granted Patent: US 7,961,922 B2
• Time in Patent Office: Days
• US Class Current: 382/131
• CPC Class Codes: G06T 2207/30016 Brain
• CPC Class Codes: G06T 7/0012 Biomedical image inspection
• CPC Class Codes: G06V 10/507 Summing image-intensity values; Histogram projection analysis
• CPC Class Codes: G06V 2201/031 of internal organs

Abstract

Systems and methods for processing medical image data with increased sensitivity to facilitate comparisons among groups of subjects are disclosed herein. In one embodiment, a method comprises receiving a first three-dimensional image comprising a plurality of voxels, reducing a regional trend within the three-dimensional image, computing a semivariogram for a region of interest, defining at least one block of spatially correlated voxels, calculating voxel weights for each voxel within the at least one block, and determining a block average count and variance for the at least one block.

Claims

1. A method for processing medical image data, the method comprising:

receiving a first three-dimensional image comprising a plurality of voxels;

reducing a regional trend within the three-dimensional image;

computing a semivariogram for a region of interest;

defining at least one block of spatially correlated voxels;

calculating voxel weights for each voxel within the at least one block; and

determining a block average count and variance for the at least one block.

2. The method of claim 1, further comprising repeating the receiving, reducing, computing, defining, calculating, and determining steps for a second three-dimensional image.

3. The method of claim 2, further comprising performing a statistical analysis upon the first and second three-dimensional images to detect a difference between the first and second three-dimensional images.

4. The method of claim 1, where removing regional trends from the three-dimensional image comprises fitting a regional mean spatially across a region.

5. The method of claim 1, where computing a semivariogram for a region of interest comprises determining an extent of correlation among at least a subset of the plurality of voxels.

6. The method of claim 5 where determining the extent of correlation comprises determining range parameter.

7. The method of claim 6, where defining the at least one block comprises defining at least one block of spatially correlated voxels based, at least in part, upon the extent of correlation

8. The method of claim 7, where defining the at least one block comprises choosing the at least one block so that centers of adjacent blocks are separated by a distance approximately equal to the range parameter.

9. The method of claim 7, where defining the at least one block comprises choosing the at least one block so that centers of adjacent blocks are separated by a distance larger than the range parameter.

10. The method of claim 1, where the at least one block corresponds to a known structure.

11. The method of claim 1, where the at least one block is chosen based upon the range parameter and an anatomic shapes of a region of interest.

12. The method of claim 1, where calculating voxel weights comprises solving a kriging system of equations.

13. The method of claim 1, where determining a block average count and variance comprises taking a weighted average of voxels within a block.

14. The method of claim 1, where the first three-dimensional image is a brain image.

15. A method for processing brain image data with increased sensitivity to facilitate comparisons among groups of subjects, the method comprising:

selecting a plurality of groups of subjects;

receiving at least one three-dimensional brain image from each of the subjects, each of three-dimensional brain images comprising a plurality of voxels; and

modeling the at least one region of interest of the plurality of three-dimensional brain images, in part, by selecting blocks of voxels so that centers of adjacent blocks are separated by a distance no closer than a range at which voxel counts are uncorrelated.

16. The method of claim 15, further comprising pre-processing the plurality of three-dimensional brain images to reduce a regional trend.

17. The method of claim 15, further comprising averaging a plurality of blocks of for each of the plurality of three-dimensional brain images.

18. The method of claim 17, further comprising using statistical analysis to detect a medical condition for at least one of the plurality of groups of subjects.

19. A computer readable medium having a computer program recorded thereon that causes a computer to perform the steps of:

receiving a plurality of three-dimensional images, each image comprising a plurality of voxels and each voxel indicating an intensity measurement;

reducing a regional trend by spatially fitting a regional mean across a region;

computing semivariograms using detrended voxel measurements for each region of interest (ROI) to determine an extent of intervoxel correlation;

defining blocks of spatially correlated voxels so that centers of adjacent blocks are no closer than a range parameter computed from the semivariograms;

calculating voxels weights within each defined block by solving a kriging system of equations; and

determining block average counts and variances for each block by taking the weighted average of voxels within a block.

20. The computer readable medium of claim 19, the computer program recorded thereon further causing a computer to perform a statistical analysis upon the plurality of three-dimensional images to detect a difference among subsets of the plurality of three-dimensional images, each subset belonging to a different subject group.