Granted Patent B2

US 12,274,542 · App. 17/654,598 · Granted Apr 15, 2025

Inventors: Xiaodong Wang (Shanghai, CN); Wenjun Yu (Shanghai, CN); Yufei Mao (Shanghai, CN); Xu Wang (Shanghai, CN); Ke Wu (Shanghai, CN); Ce Wang (Shanghai, CN); Peng Zhao (Shanghai, CN); Chuanfeng Lv (Shanghai, CN)

Assignee: SHANGHAI UNITED IMAGING HEALTHCARE CO., LTD.

A61B5/055A61B5/0037A61B5/004A61B5/489A61B6/504A61B6/5294G06F18/214G06F18/2148G06F18/24G06T5/30G06T7/0014G06T7/11G06T7/136G06V10/34G06V10/755G06V10/7747A61B6/037A61B6/469A61B6/481G06T2200/04G06T2207/10081G06T2207/10088G06T2207/10104G06T2207/10108G06T2207/10116G06T2207/20021G06T2207/20081G06T2207/30101G06T2207/30172G06V2201/03

Abstract

Claims (80)

1. A method for determining a centerline of a blood vessel in an image associated with a subject implemented on at least one machine each of which has at least one processor and at least one storage, the method comprising:

obtaining a centerline model used for identifying a centerline of a blood vessel; and

identifying the centerline of the blood vessel based on the centerline model, wherein

the centerline model includes a centerline average model; and

the centerline model is generated by a process including:

registering a plurality of sample images;

extracting a reference centerline of a reference blood vessel in each of the plurality of registered sample images; and

determining the centerline average model based on a plurality of reference centerlines of the reference blood vessel in the plurality of registered sample images.

2. The method of claim 1 , wherein the determining the centerline average model based on a plurality of reference centerlines of the reference blood vessel in the plurality of registered sample images includes:

determining a plurality of centerline shape vectors corresponding the plurality of reference centerlines respectively;

performing an alignment operation on the plurality of centerline shape vectors; and

determining the centerline average model based on the plurality of aligned centerline shape vectors.

3. The method of claim 2 , wherein the determining the centerline average model based on a plurality of reference centerlines of the reference blood vessel in the plurality of registered sample images includes:

performing a principal component analysis (PCA) method after the alignment operation.

4. The method of claim 1 , wherein

identifying the centerline of the blood vessel based on the centerline model includes:

determining a candidate centerline of the blood vessel in the image by mapping the centerline average model to the image;

determining a distance field in the image; and

determining the centerline of the blood vessel by adjusting the candidate centerline of the blood vessel based on the distance field.

5. The method of claim 4 , wherein the determining the centerline of the blood vessel by adjusting the candidate centerline of the blood vessel based on the distance field includes:

for points on the candidate center line, assigning high weights to points satisfying a predetermined distance field value of the distance field and low weights to points not satisfying the predetermined distance field value; and

determining the centerline of the blood vessel by adjusting the candidate centerline of the blood vessel based on the weights of the points.

6. The method of claim 1 , wherein the registering a plurality of sample images includes:

determining one of the plurality of sample images as a template image;

registering residual other images of the plurality of sample images based on the template image.

7. The method of claim 1 , wherein the extracting a reference centerline of a reference blood vessel in each of the plurality of registered sample images includes:

sampling pixels on each reference centerline of the reference blood vessel in each of the plurality of registered sample images to obtain a plurality of centerline sample points.

8. The method of claim 7 , wherein the determining the centerline average model based on a plurality of reference centerlines of the reference blood vessel in the plurality of registered sample images includes:

extracting the centerline average model from the plurality of centerline sample points by using an active shape model (ASM) algorithm.

9. A system for determining a centerline of a blood vessel in an image associated with a subject, comprising:

at least one storage device including a set of instructions; and

at least one processor in communication with the at least one storage device, wherein when executing the set of instructions, the at least one processor is directed to cause the system to perform operations including:

obtaining a centerline model used for identifying a centerline of a blood vessel; and

identifying the centerline of the blood vessel based on the centerline model, wherein

the centerline model includes a centerline average model; and

the centerline model is generated by a process including:

registering a plurality of sample images;

extracting a reference centerline of a reference blood vessel in each of the plurality of registered sample images; and

determining the centerline average model based on a plurality of reference centerlines of the reference blood vessel in the plurality of registered sample images.

10. The system of claim 9 , wherein the determining the centerline average model based on a plurality of reference centerlines of the reference blood vessel in the plurality of registered sample images includes:

determining a plurality of centerline shape vectors corresponding the plurality of reference centerlines respectively;

performing an alignment operation on the plurality of centerline shape vectors; and

determining the centerline average model based on the plurality of aligned centerline shape vectors.

11. The system of claim 10 , wherein the determining the centerline average model based on a plurality of reference centerlines of the reference blood vessel in the plurality of registered sample images includes:

performing a principal component analysis (PCA) method after the alignment operation.

12. The system of claim 9 , wherein

identifying the centerline of the blood vessel based on the centerline model includes:

determining a candidate centerline of the blood vessel in the image by mapping the centerline average model to the image;

determining a distance field in the image; and

determining the centerline of the blood vessel by adjusting the candidate centerline of the blood vessel based on the distance field.

13. The system of claim 12 , wherein the determining the centerline of the blood vessel by adjusting the candidate centerline of the blood vessel based on the distance field includes:

for points on the candidate center line, assigning high weights to points satisfying a predetermined distance field value of the distance field and low weights to points not satisfying the predetermined distance field value; and

determining the centerline of the blood vessel by adjusting the candidate centerline of the blood vessel based on the weights of the points.

14. The system of claim 9 , wherein the registering a plurality of sample images includes:

determining one of the plurality of sample images as a template image;

registering residual other images of the plurality of sample images based on the template image.

15. The system of claim 9 , wherein the extracting a reference centerline of a reference blood vessel in each of the plurality of registered sample images includes:

sampling pixels on each reference centerline of the reference blood vessel in each of the plurality of registered sample images to obtain a plurality of centerline sample points.

16. The system of claim 15 , wherein the determining the centerline average model based on a plurality of reference centerlines of the reference blood vessel in the plurality of registered sample images includes:

extracting the centerline average model from the plurality of centerline sample points by using an active shape model (ASM) algorithm.

17. A non-transitory computer readable medium, comprising a set of instructions for determining a centerline of a blood vessel, wherein when executed by at least one processor, the set of instructions direct the at least one processor to effectuate a method, the method comprising:

obtaining a centerline model used for identifying a centerline of a blood vessel; and

identifying the centerline of the blood vessel based on the centerline model, wherein

the centerline model includes a centerline average model; and

the centerline model is generated by a process including:

registering a plurality of sample images;

extracting a reference centerline of a reference blood vessel in each of the plurality of registered sample images; and

determining the centerline average model based on a plurality of reference centerlines of the reference blood vessel in the plurality of registered sample images.

18. The non-transitory computer readable medium of claim 17 , wherein the determining the centerline average model based on a plurality of reference centerlines of the reference blood vessel in the plurality of registered sample images includes:

determining a plurality of centerline shape vectors corresponding the plurality of reference centerlines respectively;

performing an alignment operation on the plurality of centerline shape vectors; and

determining the centerline average model based on the plurality of aligned centerline shape vectors.

19. The non-transitory computer readable medium of claim 17 , wherein

the identifying the centerline of the blood vessel based on the centerline model includes:

determining a candidate centerline of the blood vessel in the image by mapping the centerline average model to the image;

determining a distance field in the image; and

determining the centerline of the blood vessel by adjusting the candidate centerline of the blood vessel based on the distance field.

20. The non-transitory computer readable medium of claim 19 , wherein the determining the centerline of the blood vessel by adjusting the candidate centerline of the blood vessel based on the distance field includes:

for points on the candidate center line, assigning high weights to points satisfying a predetermined distance field value of the distance field and low weights to points not satisfying the predetermined distance field value; and

determining the centerline of the blood vessel by adjusting the candidate centerline of the blood vessel based on the weights of the points.

Assignments (1)

Priority Claims (7)

Continuity (4)

References Cited (139)