Methods and systems for performing attenuation correction

1. A method for correcting a positron emission tomography (PET) image, said method comprising:

using one or more processors to obtain a magnetic resonance (MR) image dataset;

classifying, using the one or more processors, at least one object in the MR image as a bone;

identifying, using the one or more processors, a reference bone voxel in the MR image dataset;

counting, using the one or more processors, a number and distribution of neighbor bone voxels for the reference bone voxel;

generating, using the one or more processors, a MR-derived PET attenuation correction factor scaling for the reference bone voxel based on the number and distribution of neighbor bone voxels; and

attenuation correcting a plurality of PET emission data using the MR-derived PET attenuation correction factors.

2. The method of claim 1 , wherein classifying the object further comprises performing an intensity based segmentation of the MR image dataset to identify the bone.

3. The method of claim 1 , wherein classifying the object further comprises:

comparing the object to a plurality of images in an atlas; and

classifying the bone based on the comparison.

4. The method of claim 1 , wherein the MR image dataset comprises a plurality of voxels, said classifying the object further comprises:

assigning a first label map value to voxels representing bone; and

assigning a second label map value to voxels that are not bone.

5. The method of claim 1 , wherein generating MR-derived PET attenuation correction factors further comprises:

counting, using the one or more processors, a number and distribution of connecting bone voxels for the reference bone voxel; and

wherein the MR-derived PET attenuation correction factor scaling for the reference bone voxel is further based on the number and distribution of connecting bone voxels.

6. The method of claim 1 , wherein generating MR-derived PET attenuation correction factors further comprises:

assigning a label map value to voxels representing bone;

identifying neighbor voxels, and a relative distribution of the neighbor voxels, having the same label map value as a reference bone voxel; and

generating a MR-derived PET attenuation correction factor for the reference bone voxel based on the number and distribution of neighbor bone voxels.

7. The method of claim 1 , further comprising:

determining, using the one or more processors, a number and distribution of neighbor bone voxels around a reference bone voxel; and

generating, using the one or more processors, the MR-derived PET attenuation correction factors scaling based on the determined number and distribution of neighbor bone voxels.

8. A medical imaging system comprising:

a magnetic resonance imaging (MRI) system;

a positron emission tomography (PET) imaging system; and

a computer coupled to the MRI system and the PET system, said computer being programmed to:

obtain a MR image dataset;

classify at least one object in the MR image as a bone;

identify a reference bone voxel in the MR image dataset;

count a number and distribution of neighbor bone voxels for the reference bone voxel;

generate a MR-derived PET attenuation correction factor scaling for the reference bone voxel based on the number and distribution of neighbor bone voxels; and

attenuation correct a plurality of positron emission tomography (PET) emission data using the MR-derived PET attenuation correction factors.

9. The medical imaging system of claim 8 , wherein the computer is further programmed to perform an intensity based segmentation of the MR image dataset to identify the bone.

10. The medical imaging system of claim 8 , wherein the computer is further programmed to:

compare the object to a plurality of images in an atlas; and

classify the bone based on the comparison.

11. The medical imaging system of claim 8 , wherein the computer is further programmed to:

assign a first label map value to voxels representing bone; and

assign a second label map value to voxels that are not bone.

12. The medical imaging system of claim 8 , wherein the computer is further programmed to:

identify a reference bone voxel in the MR image dataset;

count a number and distribution of connecting bone voxels for the reference bone voxel; and

generate a MR-derived PET attenuation correction factor scaling for the reference bone voxel based on the number of connecting bone voxels and neighbor bone voxels.

13. The medical imaging system of claim 8 , wherein the computer is further programmed to:

assign a label map value to voxels representing bone;

identify neighbor voxels and a distribution of the neighbor voxels having the same label map value as a reference bone voxel; and

generate a MR-derived PET attenuation correction factor scaling for the reference bone voxel based on the number of neighbor bone voxels.

14. The medical imaging system of claim 8 , wherein the computer is further programmed to:

determine a number and distribution of connecting bone voxels around a reference bone voxel; and

generate the MR-derived PET attenuation correction factors based on the determined number and distribution of connecting bone voxels and neighbor bone voxels.

15. A non-transitory computer readable medium encoded with a program programmed to instruct a computer to:

obtain a magnetic resonance (MR) image dataset;

classify at least one object in the MR image as a bone;

identify a reference bone voxel in the MR image dataset;

count a number of neighbor bone voxels for the reference bone voxel;

generate a MR-derived PET attenuation correction factor for the reference bone voxel based on the number of neighbor bone voxels; and

attenuation correct a plurality of positron emission tomography (PET) emission data using the MR-derived PET attenuation correction factors.

16. The non-transitory computer readable medium of claim 15 , wherein the program is further programmed to instruct the computer to perform an intensity based segmentation of the MR image dataset to identify the bone.

17. The non-transitory computer readable medium of claim 15 , wherein the program is further programmed to instruct the computer to:

compare the object to a plurality of images in an atlas; and

classify the bone based on the comparison.

18. The non-transitory computer readable medium of claim 15 , wherein the program is further programmed to instruct the computer to:

assign a first label map value to voxels representing bone; and

assign a second label map value to voxels that are not bone.

19. The non-transitory computer readable medium of claim 15 , wherein the program is further programmed to instruct the computer to:

identify a reference bone voxel in the MR image dataset;

count a number and distribution of neighbor bone voxels for the reference bone voxel; and

generate a MR-derived PET attenuation correction factor scaling for the reference bone voxel based on the number of neighbor bone voxels.

20. The non-transitory computer readable medium of claim 15 , wherein the program is further programmed to instruct the computer to:

determine a number and distribution of neighbor bone voxels around a reference bone voxel; and

generate the MR-derived PET attenuation correction factors based on the determined number of neighbor bone voxels.