Shape similarity measure for body tissue

Patent metadata

Field	Value
Publication number	US-10186031-B2
Application number	US-201615365537-A
Country	US
Kind code	B2
Filing date	Nov 30, 2016
Priority date	Jun 20, 2014
Publication date	Jan 22, 2019
Grant date	Jan 22, 2019

Abstract

A shape similarity metric can be provided that indicates how similar two or more shapes are. A difference between a union of the shapes and an intersection of the shapes can be used to determine the similarity metric. The shape similarity metric can provide an average distance between the shapes. Different processes for determining shapes can be evaluated for accuracy based on the shape similarity metric. New or alternative shape-determining processes can be compared for accuracy against other shape-determining processes including reference shape-determining processes. Shape similarity metrics can be determined for two-dimensional shapes and three-dimensional shapes.

First claim

What is claimed is:

1. A method of determining an accuracy of shape-determining processes for identifying shapes of a body tissue in one or more images of a patient, the method comprising:
receiving first data that defines a first boundary of a first shape of the body tissue in at least a portion of the one or more images, the first shape being determined by a first shape-determining process;
receiving second data that defines a second boundary of a second shape of the body tissue in at least a portion of the one or more images, the second shape being determined by a second shape-determining process;
determining, by a computer system, an intersection shape of the first shape and the second shape;
determining, by the computer system, a union shape of the first shape and the second shape;
calculating, by the computer system, a difference between the union shape and the intersection shape;
computing, by the computer system, a shape similarity metric based on the difference; and
providing the shape similarity metric for determining an accuracy of the second shape-determining process relative to the first shape-determining process.

2. The method of claim 1 , wherein the first data and the second data includes the one or more images of the patient, and wherein the method further comprises:
determining a reference point in the one or more images, wherein each of the one or more images include the reference point; and
defining coordinates of the first shape and the second shape with respect to the reference point.

3. The method of claim 2 , wherein calculating the difference between the union shape and the intersection shape includes:
for each of a plurality of points on the intersection shape:
identifying a corresponding point on the union shape that corresponds to the point on the intersection shape; and
calculating a distance between the point on the intersection shape and the corresponding point on the union shape; and

computing an average of the distances to obtain the difference.

4. The method of claim 1 , wherein calculating the difference between the union shape and the intersection shape includes:
computing an intersection size of the intersection shape;
computing a union size of the union shape; and
subtracting the intersection size from the union size to obtain the difference.

5. The method of claim 4 , wherein computing the shape similarity metric includes:
calculating a normalization factor; and
multiplying the difference and the normalization factor to obtain the shape similarity metric.

6. The method of claim 5 , wherein the normalization factor is determined based on a property of the first shape and the second shape.

7. The method of claim 5 , wherein computing the intersection size includes taking a square root of an intersection area of the intersection shape, wherein computing the union size includes taking the square root of a union area of the union shape, and wherein the normalization factor includes a sum of:
a first term including the square root of a first area of the first shape divided by a first circumference of the first shape; and
a second term including the square root of a second area of the second shape divided by a second circumference of the second shape.

8. The method of claim 5 , wherein the first shape is a first two-dimensional slice of the body tissue, wherein the second shape is a second two-dimensional slice of the body tissue, the method further comprising:
computing shape similarity metrics for other slices of the body tissue; and
combining the shape similarity metrics for the slices to obtain a total shape similarity metric.

9. The method of claim 5 , wherein the first shape and the second shape are three-dimensional, wherein computing the intersection size includes taking a cube root of an intersection volume of the intersection shape, wherein computing the union size includes taking the cube root of a union volume of the union shape, and wherein the normalization factor is

3
2

⁢

(

V
1

2
/
3

A
1

+

V
2

2
/
3

A

2
⁢

)

,

where V 1 is a first volume of the first shape, V 2 is a second volume of the second shape, A 1 is a first surface area of the first shape, and A 2 is a second surface area of the second shape.

10. A method of determining an accuracy of shape-determining processes for identifying shapes of a body tissue in one or more images of a patient, the method comprising:
receiving first data that defines a first boundary of a first shape of the body tissue in at least a portion of the one or more images, the first shape being determined by a first shape-determining process;
determining, by a computer system, that the first shape does not intersect with a second shape; and
computing, by the computer system, a shape similarity metric by:
determining a first size of the first shape, the first size determined using a first area or a first volume of the first shape and a first normalization factor for the first shape; and
using the first size to compute the shape similarity metric.

11. The method of claim 10 , further comprising:
receiving second data that defines a second boundary of the second shape of the body tissue, the second shape being determined by a second shape-determining process;
wherein computing the shape similarity metric further includes:
determining a second size of the second shape, the second size determined using a second area or a second volume of the second shape and a second normalization factor for the second shape; and

calculating a sum including the first size and the second size.

12. The method of claim 11 , wherein:
the first shape and the second shape are two-dimensional and the first normalization factor of the first shape includes an inverse of a circumference of the first shape, or
the first shape and the second shape are three-dimensional and the first normalization factor of the first shape includes the inverse of a surface area of the first shape.

13. The method of claim 10 , wherein a corresponding second shape does not exist, and wherein:
the first shape is two-dimensional and the shape similarity metric is

2
⁢
a

C

 where α is an area of the first shape and C is

Assignees

• Varian Medical Systems Int Ag

Inventors

• Coradi Thomas
• Haas Benjamin

Classifications

• G06T7/12 Primary

  Edge-based segmentation · CPC title

• G06V10/761

  Proximity, similarity or dissimilarity measures · CPC title

• G06T7/0012 Primary

  Biomedical image inspection · CPC title

• G06F18/22

  Matching criteria, e.g. proximity measures · CPC title

• G06V10/752

  Contour matching · CPC title