System and method for angle measurements using magnetic resonance images

Patent metadata

Field	Value
Publication number	US-10820837-B2
Application number	US-201916379122-A
Country	US
Kind code	B2
Filing date	Apr 9, 2019
Priority date	Apr 9, 2018
Publication date	Nov 3, 2020
Grant date	Nov 3, 2020

Abstract

In a method for alpha angle measurements of a bone joint based on magnetic resonance imaging (MRI) data of the bone joint, a three-dimensional (3D) bone surface image of the bone joint is generated based on two-dimensional (2D) segmentations on radial plane slices of the bone joint. Based on the 3D bone surface image, a head center of a head of the bone joint and a neck axis of a neck of the bone joint are estimated. An alpha angle model is constructed based on the estimated head center and neck axis. Further, the alpha angle measurement is determined for each of the radial plane slices of the bone joint based on the constructed alpha angle model.

First claim

What is claimed is:

1. A method for alpha angle measurements of a bone joint based on magnetic resonance imaging (MRI) data of the bone joint, comprising:
generating a three-dimensional (3D) bone surface image of the bone joint based on two-dimensional (2D) segmentations on radial plane slices of the bone joint;
estimating, based on the 3D bone surface image, a head center of a head of the bone joint and a neck axis of a neck of the bone joint;
constructing an alpha angle model based on the estimated head center and neck axis; and
determining the alpha angle measurement for each of the radial plane slices of the bone joint based on the constructed alpha angle model.

2. The method of claim 1 , further comprising:
generating a radial slice group including the radial plane slices of the bone joint based on pattern recognition of characteristics of the bone joint.

3. The method of claim 2 , further comprising:
performing 2D segmentation of the radial plane slices of the radial slice group to generate the 2D segmentations of the radial plane slices.

4. The method of claim 2 , wherein the 2D segmentation of the radial plane slices of the radial slice group comprises:
convoluting each of the radial plane slices using a convolutional network to obtain bone boundary information of the radial plane slices; and
regressing distance constraint information of the bone boundary information using a regression network to generate spatial information of the radial plane slices.

5. The method of claim 4 , wherein the generation of the 3D bone surface image is based on the spatial information of the radial plane slices.

6. The method of claim 1 , wherein the generation of the 3D bone surface image is based on a bone boundary for each of the radial plane slices of the bone joint.

7. The method of claim 2 , wherein the pattern recognition of the characteristics of the bone joint comprises performing an automatic landmark detection on the MRI data.

8. The method of claim 1 , wherein estimating the head center comprises:
roughly segmenting the 3D bone surface image to generate a bone head surface and a bone neck surface; and
performing a random sample consensus on the bone head surface to estimate sphere parameters of a sphere to fit the sphere on the head of the bone joint, the head center corresponding to a center of the sphere.

9. The method of claim 1 , wherein estimating the neck axis comprises:
determining a circle around the neck of the bone joint having a smallest diameter; and
fitting a plane for the circle, wherein a normal of the plane is the neck axis.

10. The method of claim 1 , wherein constructing the alpha angle model comprises constructing the alpha angle model for each radial plane slice of the radial plane slices based on the estimated head center and neck axis, each radial plane slice being defined by a radial azimuth angle.

11. The method of claim 1 , wherein constructing the alpha angle model comprises:
for each radial plane slice of the radial plane slices:
determining a bone boundary of the bone joint; and
determining an angle (γ) between a line defined from the head center to the determined bone boundary and a tangent line at the bone boundary, a projection of the line on the neck axis defining a center offset distance (x); and

determining, using a loss function, an optimized center offset distance based on the determined angles (γ) from each of the respective radial plane slices.

12. The method of claim 11 , wherein determining the alpha angle measurement comprises, for each radial plane slice of the radial plane slices:
determining an adjusted bone boundary of the bone joint based on the respective optimized center offset distance; and
determining an alpha angle between a line defined from the head center to the adjusted bone boundary and the neck axis.

13. The method of claim 1 , further comprising:
displaying an output image representative of the alpha angle measurement for each of the radial plane slices.

14. The method of claim 13 , wherein displaying an output image representative comprises:
overlaying the output image representative on the 3D bone surface image.

15. The method of claim 13 , wherein the bone joint is a femoral head-neck junction.

16. A non-transitory computer-readable storage medium with an executable program stored thereon, wherein, when executed, the program instructs a processor to perform the method of claim 1 .

17. An anatomical angle measurement system to determine alpha angle measurements of a bone joint based on magnetic resonance imaging (MRI) data of the bone joint, comprising:
a storage medium configured to store program instructions; and
processor circuitry configured execute the program instructions to:
generate a three-dimensional (3D) bone surface image of the bone joint based on two-dimensional (2D) segmentations on radial plane slices of the bone joint;
estimate, based on the 3D bone surface image, a head center of a head of the bone joint and a neck axis of a neck of the bone joint;
construct an alpha angle model based on the estimated head center and neck axis; and
determine the alpha angle measurement for each of the radial plane slices of the bone joint based on the constructed alpha angle model.

18. The system of claim 17 , wherein the 2D segmentation comprises:
convoluting each of the radial plane slices using a convolutional network to obtain bone boundary information of the radial plane slices; and
regressing distance constraint information of the bone boundary information using a regression network to generate spatial information of the radial plane slices.

19. The system of claim 17 , wherein:
estimating the head center comprises:
roughly segmenting the 3D bone surface image to generate a bone head surface and a bone neck surface; and
performing a random sample consensus on the bone head surface to estimate sphere parameters of a sphere to fit the sphere on the head of the bone joint, the head center corresponding to a center of the sphere; and

estimating the neck axis comprises:
determining a circle around the neck of the bone joint having a smallest diameter; and
fitting a plane for the circle, wherein a normal of the plane is the neck axis.

20. The system of claim 17 , wherein constructing the alpha angle model comprises:
for each radial plane slice of the radial plane slices:
determining a bone boundary of the bone joint; and
determining an angle (γ) between a line defined from the head center to the determined bone boundary and a tangent line at the bone boundary, a projection of the line on the neck axis defining a center offset distance (x); and

determining, using a loss function, an optimized center offset distance based on the determined angles (γ) from each of the respective radial plane slices, wherein alpha angle measurement is determined based on the optimized center offset distance.

Assignees

• Siemens Healthcare Gmbh

Inventors

• Zhao Liang
• Raithel Esther
• Zhan Yiqiang

Classifications

• A61B5/055

  involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging · CPC title

• G06T17/00 Primary

  Three-dimensional [3D] modelling for computer graphics · CPC title

• A61B5/4571

  Evaluating the hip · CPC title

• G06T2207/10088

  Magnetic resonance imaging [MRI] · CPC title

• G06T2207/30008

  Bone · CPC title