IP Library › Granted Patent US 10,346,993
Granted Patent B2
US 10,346,993 · App. 15/353,916 · Granted Jul 9, 2019

Systems and methods for image processing in magnetic resonance imaging

Inventors: Luan Jiang (Shanghai, CN); Qiang Li (Shanghai, CN); Shan Ling (Shanghai, CN)
Assignee: SHANGHAI UNITED IMAGING HEALTHCARE CO., LTD.
G06T7/33G06T7/0016G06T2207/10088G06T2207/30048
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 10,346,993
App. No.
15/353,916
Granted
Jul 9, 2019
Kind
B2
Abstract

A method and system for image processing are provided. An MR image including a plurality of slice images may be obtained. The plurality of slice images including a myocardium of a left ventricle. A reference image for each slice image of the plurality of slice images may be determined. An endocardial boundary of the myocardium for the each slice image of the plurality of slice images may be determined. The each slice image of the plurality of slice images may be registered according to a corresponding reference image. An epicardial boundary of the myocardium in the each slice image of the plurality of slice images may be determined according to the endocardial boundary of the myocardium in the registered each slice image of the plurality of slice images.

Claims (95)

1. A method comprising:

obtaining an MR image including a plurality of slice images, the plurality of slice images including a myocardium of a left ventricle, each of the plurality of slice images including one or more images obtained in one or more cardiac cycles;

for each slice image of the plurality of slice images, determining a reference image by selecting from the one or more images of the each slice image according to a blood pool area in at least one of the one or more images;

determining an endocardial boundary of the myocardium for the each slice image of the plurality of slice images;

registering the each slice image of the plurality of slice images according to the corresponding reference image; and

determining an epicardial boundary of the myocardium in the each registered slice image of the plurality of registered slice images according to the endocardial boundary of the myocardium in the each slice image of the plurality of slice images.

2. The method of claim 1 , wherein determining the reference image for the each slice image of the plurality of slice images and determining the endocardial boundary of the myocardium for the each slice image of the plurality of slice images includes:

determining the endocardial boundary of the myocardium in a starting slice image;

determining the reference image in the starting slice image;

determining the endocardial boundary of the myocardium in a non-starting slice image; and

determining the reference image in the non-starting slice image.

3. The method of claim 2 , wherein determining the endocardial boundary of the myocardium in the starting slice image and determining the reference image in the starting slice image further includes:

obtaining one or more candidate images in the starting slice image;

determining an initial region of interest in the one or more candidate images;

determining a blood pool area in the one or more candidate images according to the initial region of interest in the one or more candidate images;

determining a blood pool area in the starting slice image and the reference image in the starting slice image according to the blood pool areas in the one or more candidate images; and

determining a final region of interest in the starting slice image according to the blood pool area in the starting slice image.

4. The method of claim 2 , wherein determining the endocardial boundary of the myocardium in the non-starting slice image and determining the reference image in the non-starting slice image further includes:

obtaining one or more candidate images in the non-starting slice image;

determining an initial region of interest in the one or more candidate images according to the reference image of the previous slice of the non-starting slice image;

determining a blood pool area in the one or more candidate images according to the initial region of interest in the one or more candidate images and the reference image of the previous slice of the non-starting slice image;

determining a blood pool area in the non-starting slice image and the reference image in the non-starting slice image according to the blood pool area in the one or more candidate images;

determining a final region of interest in the non-starting slice image according to the blood pool area in the non-starting slice image.

5. The method of claim 1 , wherein registering each slice image of the plurality of slice images according to the corresponding reference image further includes:

performing a rigid registration on the each slice image of the plurality of slice images according to the corresponding reference image;

determining whether a non-rigid registration of the each slice image of the plurality of slice images after rigid registration is needed; and

performing a non-rigid registration on the each slice image of the plurality of slice images after the rigid registration if a non-rigid registration of the each slice image of the plurality of slice images after rigid registration is needed.

6. The method of claim 5 , wherein performing the rigid registration on the each slice image of the plurality of slice images further includes a first step of image registration and a second step of image registration.

7. The method of claim 5 , wherein:

the rigid registration of the each slice image of the plurality of slice images is performed from the reference image of the each slice image of the plurality of slice images

to an image obtained in an earlier cardiac cycle than the reference image and

to an image obtained in a later cardiac cycle than the reference image, and

the rigid registration of the each slice image of the plurality of slice images includes:

deforming the each slice image of the plurality of slice images,

determining a first similarity between the deformed each slice image of the plurality of slice images and the reference image of the each slice image of the plurality of slice images,

determining a second similarity, the second similarity being determined

between the deformed each slice image and a registered image of a previous slice image, or

between the deformed each slice image and a registered image of a next slice image, and

determining whether a sum of the first similarity and the second similarity is largest.

8. The method of claim 1 , wherein determining the epicardial boundary of the myocardium in the each registered slice image of the plurality of registered slice images according to the endocardial boundary of the myocardium in the each slice image of the plurality of slice images further includes:

obtaining a maximum intensity projection image and the reference image of the each registered slice image of the plurality of registered slice images,

transforming the maximum intensity projection image and the reference image of the each registered slice image of the plurality of registered slice images to a maximum intensity projection image and a reference image in polar coordinates of the each registered slice image of the plurality of registered slice images according to the endocardial boundary of the myocardium in the each slice image of the plurality of slice images;

determining a right ventricle area and a constrained area in the reference image in the polar coordinates of the each registered slice image of the plurality of registered slice images; and

determining the epicardial boundary of the myocardium in the maximum intensity projection image in the polar coordinates of the each registered slice image of the plurality of registered slice images according to the right ventricle area and the constrained area in the reference image in the polar coordinates of the each registered slice image of the plurality of registered slice images.

9. The method of claim 8 , wherein determining the right ventricle area and the constrained area in the reference image in the polar coordinates of the each registered slice image of the plurality of registered slice images further includes:

clustering the reference image in the polar coordinates of the each registered slice image of the plurality of registered slice images;

determining the right ventricle area in the reference image in the polar coordinates in a starting slice image;

determining the right ventricle area in the reference image in the polar coordinates in a non-starting slice image;

determining the constrained area according to the right ventricle area and the endocardial boundary of the myocardium in the reference image in the polar coordinates of the each registered slice image of the plurality of registered slice images.

10. The method of claim 1 further comprising:

partitioning the myocardium of the left ventricle in the each slice image of the plurality of slice images.

11. The method of claim 10 , wherein partitioning the myocardium of the left ventricle in the each slice image of plurality of slice images is according to a Bull's eye method.

12. A device for image processing, the device comprising:

a memory device configured to store instructions; and

one or more processors configured to execute the instructions to perform a method comprising:

obtaining an MR image including a plurality of slice images, the plurality of slice images including a myocardium of a left ventricle, each of the plurality of slice images including one or more images obtained in one or more cardiac cycles;

for each slice image of the plurality of slice images, determining a reference image by selecting from the one or more images of the each slice image according to a blood pool area in at least one of the one or more images;

determining an endocardial boundary of the myocardium for the each slice image of the plurality of slice images;

registering the each slice image of the plurality of slice images according to the corresponding reference image; and

determining an epicardial boundary of the myocardium in the each registered slice image of the plurality of registered slice images according to the endocardial boundary of the myocardium in the each slice image of the plurality of slice images.

13. The device of claim 12 , wherein determining the reference image for the each slice image of the plurality of slice images and determining the endocardial boundary of the myocardium for the each slice image of the plurality of slice images includes:

determining the endocardial boundary of the myocardium in a starting slice image;

determining the reference image in the starting slice image;

determining the endocardial boundary of the myocardium in a non-starting slice image; and

determining the reference image in the non-starting slice image.

14. The device of claim 13 , wherein determining the endocardial boundary of the myocardium in the starting slice image and determining the reference image in the starting slice image further includes:

obtaining one or more candidate images in the starting slice image;

determining an initial region of interest in the one or more candidate images;

determining a blood pool area in the one or more candidate images according to the initial region of interest in the one or more candidate images;

determining a blood pool area in the starting slice image and the reference image in the starting slice image according to the blood pool areas in the one or more candidate images; and

determining a final region of interest in the starting slice image according to the blood pool area in the starting slice image.

15. The device of claim 13 , wherein determining the endocardial boundary of the myocardium in the non-starting slice image and determining the reference image in the non-starting slice image further includes:

obtaining one or more candidate images in the non-starting slice image;

determining an initial region of interest in the one or more candidate images according to the reference image of the previous slice of the non-starting slice image;

determining a blood pool area in the one or more candidate images according to the initial region of interest in the one or more candidate images and the reference image of the previous slice of the non-starting slice image;

determining a blood pool area in the non-starting slice image and the reference image in the non-starting slice image according to the blood pool area in the one or more candidate images;

determining a final region of interest in the non-starting slice image according to the blood pool area in the non-starting slice image.

16. The device of claim 12 , wherein registering each slice image of the plurality of slice images according to the corresponding reference image further includes:

performing a rigid registration on the each slice image of the plurality of slice images according to the corresponding reference image;

determining whether a non-rigid registration of the each slice image of the plurality of slice images after rigid registration is needed; and

performing a non-rigid registration on the each slice image of the plurality of slice images after the rigid registration if a non-rigid registration of the each slice image of the plurality of slice images after rigid registration is needed.

17. The device of claim 16 , wherein performing the rigid registration on the each slice image of the plurality of slice images further includes a first step of image registration and a second step of image registration.

18. The device of claim 12 , wherein determining the epicardial boundary of the myocardium in the each registered slice image of the plurality of registered slice images according to the endocardial boundary of the myocardium in the each slice image of the plurality of slice images further includes:

obtaining a maximum intensity projection image and the reference image of the each registered slice image of the plurality of registered slice images,

transforming the maximum intensity projection image and the reference image of the each registered slice image of the plurality of registered slice images to a maximum intensity projection image and a reference image in polar coordinates of the each registered slice image of the plurality of registered slice images according to the endocardial boundary of the myocardium in the each slice image of the plurality of slice images;

determining a right ventricle area and a constrained area in the reference image in the polar coordinates of the each registered slice image of the plurality of registered slice images; and

determining the epicardial boundary of the myocardium in the maximum intensity projection image in the polar coordinates of the each registered slice image of the plurality of registered slice images according to the right ventricle area and the constrained area in the reference image in the polar coordinates of the each registered slice image of the plurality of registered slice images.

19. The method of claim 18 , wherein determining the right ventricle area and the constrained area in the reference image in the polar coordinates of the each registered slice image of the plurality of registered slice images further includes:

clustering the reference image in the polar coordinates of the each registered slice image of the plurality of slice images;

determining the right ventricle area in the reference image in the polar coordinates in a starting slice image;

determining the right ventricle area in the reference image in the polar coordinates in a non-starting slice image;

determining the constrained area according to the right ventricle area and the endocardial boundary of the myocardium in the reference image in the polar coordinates of the each registered slice image of the plurality of registered slice images.

20. The device of claim 12 , wherein the instructions method further comprising:

partitioning the myocardium of the left ventricle in the each slice image of the plurality of slice images.

21. The device of claim 20 , wherein partitioning the myocardium of the left ventricle in the each slice image of plurality of slice images is according to a Bull's eye method.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Aug 5, 2017
From: JIANG, LUAN; LI, QIANG; LING, SHAN
To: SHANGHAI UNITED IMAGING HEALTHCARE CO., LTD.
Reel/Frame 043209/0883 →
Priority Claims (2)
CN 2015 1 0788341 · Nov 17, 2015 · national
CN 2015 1 0788374 · Nov 17, 2015 · national
Continuity (1)
Related Publication 20170140530A1 · May 18, 2017