Correcting positron emission tomography image attenuation

1. A method of correcting Positron Emission Tomography (PET) image attenuation, comprising:

segmenting an MRI image of a subject to obtain respective MRI sub-images corresponding to attenuation units of the subject;

for each MRI sub-image, obtaining pseudo CT sub-images corresponding to the MRI sub-images based on an MRI-CT atlas, wherein,

the pseudo CT sub-image is obtained by selecting a plurality of CT template sub-images present in the MRI-CT atlas that corresponds to a plurality of MRI template sub-images in the MRI-CT atlas matched to the MRI sub-image, and

the plurality of MRI template sub-images is matched to the MRI sub-image by a matching relationship comprising shape factors and intensity weight factors;

splicing the respective pseudo CT sub-images to obtain a pseudo CT image; and

performing attenuation correction for a PET image of the subject according to the pseudo CT image, wherein the PET image and the MRI image are obtained through a same scan process for the subject.

2. The method of claim 1 , wherein for each MRI sub-image obtaining the pseudo CT sub-images corresponding to the MRI sub-images based on the MRI-CT atlas comprises:

obtaining the plurality of MRI template sub-images associated with the attenuation unit corresponding to the MRI sub-image from an MRI template set in the MRI-CT atlas;

obtaining the matching relationship between the plurality of MRI template sub-images and the MRI sub-image, wherein

the matching relationship comprises respective weighting coefficients for the MRI template sub-images based on image intensity and image shape such that a weight sum of the plurality of MRI template sub-images is matched with the MRI sub-image; and

the matching relationship is calculated with formula (1):

M

patient

i

=

∑

k

=

1

K

⁢

t

k

⁢

T

ik

⁡

(

M

template

ik

)

,

(

1

)

wherein

t k represents the intensity weight factor of the k-th MRI template sub-image,

T ik represents the shape factor of the k-th MRI template sub-image,

M patient i represents the MRI sub-image i,

M template ik template represents the k-th MRI template sub-image corresponding to the MRI sub-image i, and

K represents a number of the plurality of MRI template sub-image;

obtaining the plurality of CT template sub-images corresponding to the plurality of MRI template sub-images from a CT template set in the MRI-CT atlas; and

obtaining the pseudo CT sub-image corresponding to the MRI sub-image according to the plurality of CT template sub-images and the matching relationship.

3. The method of claim 1 , wherein splicing the respective pseudo CT sub-images comprises:

placing the respective pseudo CT sub-images into border images for the attenuation units corresponding to the respective pseudo CT sub-images.

4. The method of claim 1 , further comprising:

performing smoothing correction for respective adjacent borders of attenuation units in adjacent pseudo CT sub-images.

5. The method of claim 1 , wherein the attenuation unit is an organ tissue of the subject.

6. A PET/MRI device, comprising:

a processor, and

a non-transitory machine-readable storage medium storing machine executable instructions which are executable by the processor to:

segment an MRI image of a subject to obtain respective MRI sub-images corresponding to attenuation units of the subject;

for each MRI sub-image, obtain a pseudo CT sub-images corresponding to the MRI sub-images based on an MRI-CT atlas, wherein;

the pseudo CT sub-image is obtained by selecting a plurality of CT template sub-images present in the MRI-CT atlas that corresponds to a plurality of MRI template sub-images in the MRI-CT atlas matched to the MRI sub-image, and

the plurality of MRI template sub-images is matched to the MRI sub-image by a matching relationship comprising shape factors and intensity weight factors;

splice the respective pseudo CT sub-images to obtain a pseudo CT image; and

perform attenuation correction for a PET image of the subject according to the pseudo CT image, wherein the PET image and the MRI image are obtained through a same scan process for the subject.

7. The device of claim 6 , wherein the processor is caused by the machine-executable instructions to:

obtain, for each of the MRI sub-images, the plurality of MRI template sub-images associated with the attenuation unit corresponding to the MRI sub-image from an MRI template set in the MRI-CT atlas;

obtain the matching relationship between the plurality of MRI template sub-images and the MRI sub-image, wherein

the matching relationship comprises respective weighting coefficients for the MRI template sub-images based on image intensity and image shape in a way that a weight sum of the plurality of MRI template sub-images is matched with the MRI sub-image; and

the matching relationship is calculated with formula (1):

M

patient

i

=

∑

k

=

1

K

⁢

t

k

⁢

T

ik

⁡

(

M

template

ik

)

,

(

1

)

wherein

t k represents the intensity weight factor of the k-th MRI template sub-image,

T ik represents the shape factor of the k-th MRI template sub-image,

M patient i represents the MRI sub-image i,

M template ik represents the k-th MRI template sub-image corresponding to the MRI sub-image i, and

K represents a number of the plurality of MRI template sub-image;

obtain the plurality of CT template sub-images corresponding to the plurality of MRI template sub-images from a CT template set in the MRI-CT atlas; and

obtain the pseudo CT sub-image corresponding to the MRI sub-image according to the plurality of CT template sub-images and the matching relationship.

8. The device of claim 6 , wherein the processor is caused by the machine-executable instructions to:

place the respective pseudo CT sub-images into border images for the attenuation units corresponding to the respective pseudo CT sub-images.

9. The device of claim 6 , wherein the processor is caused by the machine-executable instructions to:

perform smoothing correction for respective adjacent borders of attenuation units in adjacent pseudo CT sub-images.

10. The device of claim 6 , wherein the attenuation unit is an organ tissue of the subject.