Granted Patent B2

US 12,229,916 · App. 17/838,059 · Granted Feb 18, 2025

Inventors: Jingyi Yin (Nanjing, CN); Jue Zhang (Nanjing, CN)

Assignee: Nanjing Leapsonics Technology Co., Ltd.

G06T3/4053A61B8/481G06T5/30G06T5/70G06T7/136G06T7/30G06T7/50G06T7/68G06T2207/10016G06T2207/10136G06T2207/30004

Abstract

Claims (53)

1. A super-resolution reconstruction method for three-dimensional contrast-enhanced ultrasound images, applied to a first three-dimensional local image sequence comprising microbubbles, comprising:

performing at least one thinning operation on the first three-dimensional local image sequence, wherein the thinning operation is used to enhance motion trajectories of microbubbles; and

performing an image reconstruction operation based on the first three-dimensional local image sequence subjected to the at least one thinning operation, so as to generate a three-dimensional super-resolution image;

wherein the performing at least one thinning operation on the first three-dimensional local image sequence comprises:

performing a first thinning operation on the first three-dimensional local image sequence to generate a second three-dimensional local image sequence corresponding to the first three-dimensional local image sequence, wherein the first thinning operation is used to enhance motion trajectories of microbubbles for a first time; and

performing a second thinning operation on the second three-dimensional local image sequence to generate a third three-dimensional local image sequence corresponding to the second three-dimensional local image sequence, wherein the second thinning operation is used to enhance motion trajectories of microbubbles for a second time;

wherein the performing an image reconstruction operation based on the first three-dimensional local image sequence subjected to the at least one thinning operation, so as to generate a three-dimensional super-resolution image comprises:

performing an image reconstruction operation based on the third three-dimensional local image sequence, so as to generate three-dimensional super-resolution images corresponding to the third three-dimensional local image sequence.

2. The super-resolution reconstruction method for three-dimensional contrast-enhanced ultrasound images according to claim 1 , wherein the performing a first thinning operation on the first three-dimensional local image sequence to generate a second three-dimensional local image sequence corresponding to the first three-dimensional local image sequence comprises:

determining, for each frame of the first three-dimensional local image in the first three-dimensional local image sequence, distance information between a microbubble area in the first three-dimensional local image and a background area corresponding to the microbubble area;

performing a first weighting operation on the first three-dimensional local image based on the distance information to generate a first weighted image; and

generating the second three-dimensional local image sequence based on the first weighted images respectively corresponding to the first three-dimensional local images in the first three-dimensional local image sequence.

3. The super-resolution reconstruction method for three-dimensional contrast-enhanced ultrasound images according to claim 2 , wherein the determining distance information between a microbubble area in the first three-dimensional local image and a background area corresponding to the microbubble area comprises:

performing binarization processing on the first three-dimensional local images based on the microbubble area and the background area to generate a binarized image; and

determining the distance information between the microbubble area and the background area based on the binarized image.

4. The super-resolution reconstruction method for three-dimensional contrast-enhanced ultrasound images according to claim 2 , wherein a pixel value of a pixel coordinate corresponding to the microbubble area is set to 1, and a pixel value of a pixel coordinate corresponding to the background area is set to 0.

5. The super-resolution reconstruction method for three-dimensional contrast-enhanced ultrasound images according to claim 2 , wherein the microbubble area comprises a plurality of pixel blocks, and the determining distance information between a microbubble area in the first three-dimensional local image and a background area corresponding to the microbubble area comprises:

determining a shortest distance from each the pixel block in the plurality of pixel blocks to the background area respectively; and

determining the distance information based on the shortest distances respectively corresponding to the plurality of pixel blocks.

6. The super-resolution reconstruction method for three-dimensional contrast-enhanced ultrasound images according to claim 1 , wherein multiple frames of the second three-dimensional local images comprised in the second three-dimensional local image sequence are obtained by performing the first thinning operation on multiple frames of the first three-dimensional local images in the first three-dimensional local image sequence frame by frame.

7. The super-resolution reconstruction method for three-dimensional contrast-enhanced ultrasound images according to claim 1 , wherein the performing a second thinning operation on the second three-dimensional local image sequence to generate a third three-dimensional local image sequence corresponding to the second three-dimensional local image sequence comprises:

performing, for each frame of the second three-dimensional local image in the second three-dimensional local image sequence, a degree of radial symmetry estimation operation on each pixel unit in the second three-dimensional local images, so as to determine a weight value corresponding to the pixel unit;

performing a second weighting operation on the second three-dimensional local images based on the weight values respectively corresponding to the pixel units in the second three-dimensional local images to generate a second weighted image; and

generating the third three-dimensional local image sequence based on the second weighted images respectively corresponding to the second three-dimensional local images in the second three-dimensional local image sequence.

8. The super-resolution reconstruction method for three-dimensional contrast-enhanced ultrasound images according to claim 7 , wherein the degree of radial symmetry estimation operation is implemented based on a degree of radial symmetry calculation method in super-resolution radial fluctuations.

9. The super-resolution reconstruction method for three-dimensional contrast-enhanced ultrasound images according to claim 7 , wherein a number of sampling points is 12 and a sampling radius is 1 in the degree of radial symmetry estimation operation.

10. The super-resolution reconstruction method for three-dimensional contrast-enhanced ultrasound images according to claim 1 , wherein before the performing at least one thinning operation on the first three-dimensional local image sequence, the method further comprises:

performing a registration operation on the first three-dimensional local image sequence, wherein the registration operation comprises rigid registration operations and/or flexible registration operations.

11. The super-resolution reconstruction method for three-dimensional contrast-enhanced ultrasound images according to claim 10 , wherein the performing a registration operation on the first three-dimensional local image sequence comprises:

performing, by using a three-dimensional Morphon multi-scale registration method, the registration operation on the first three-dimensional local image sequence.

12. The super-resolution reconstruction method for three-dimensional contrast-enhanced ultrasound images according to claim 1 , wherein the performing an image reconstruction operation based on the first three-dimensional local image sequence subjected to the at least one thinning operation, so as to generate three-dimensional super-resolution images comprises:

performing an accumulation operation on the first three-dimensional local image sequence subjected to the at least one thinning operation based on image sequence information corresponding to the first three-dimensional local image sequence, so as to generate the three-dimensional super-resolution image.

13. A non-transitory computer readable storage medium, storing a computer program for executing the super-resolution reconstruction method for three-dimensional contrast-enhanced ultrasound images according to claim 1 .

14. An electronic device, comprising:

a processor; and

a memory for storing executable instructions of the processor;

wherein the processor is configured to execute a super-resolution reconstruction method for three-dimensional contrast-enhanced ultrasound images, applied to a first three-dimensional local image sequence comprising microbubbles, comprising:

performing at least one thinning operation on the first three-dimensional local image sequence, wherein the thinning operation is used to enhance motion trajectories of microbubbles; and

performing an image reconstruction operation based on the first three-dimensional local image sequence subjected to the at least one thinning operation, so as to generate a three-dimensional super-resolution image;

wherein the performing at least one thinning operation on the first three-dimensional local image sequence comprises:

performing a first thinning operation on the first three-dimensional local image sequence to generate a second three-dimensional local image sequence corresponding to the first three-dimensional local image sequence, wherein the first thinning operation is used to enhance motion trajectories of microbubbles for a first time; and

performing a second thinning operation on the second three-dimensional local image sequence to generate a third three-dimensional local image sequence corresponding to the second three-dimensional local image sequence, wherein the second thinning operation is used to enhance motion trajectories of microbubbles for a second time;

wherein the performing an image reconstruction operation based on the first three-dimensional local image sequence subjected to the at least one thinning operation, so as to generate a three-dimensional super-resolution image comprises:

performing an image reconstruction operation based on the third three-dimensional local image sequence, so as to generate three-dimensional super-resolution images corresponding to the third three-dimensional local image sequence.

15. The electronic device according to claim 14 , wherein the performing a first thinning operation on the first three-dimensional local image sequence to generate a second three-dimensional local image sequence corresponding to the first three-dimensional local image sequence comprises:

determining, for each frame of the first three-dimensional local image in the first three-dimensional local image sequence, distance information between a microbubble area in the first three-dimensional local image and a background area corresponding to the microbubble area;

performing a first weighting operation on the first three-dimensional local image based on the distance information to generate a first weighted image; and

generating the second three-dimensional local image sequence based on the first weighted images respectively corresponding to the first three-dimensional local images in the first three-dimensional local image sequence.

16. The electronic device according to claim 15 , wherein the determining distance information between a microbubble area in the first three-dimensional local image and a background area corresponding to the microbubble area comprises:

performing binarization processing on the first three-dimensional local images based on the microbubble area and the background area to generate a binarized image; and

determining the distance information between the microbubble area and the background area based on the binarized image.

17. The electronic device according to claim 15 , wherein a pixel value of a pixel coordinate corresponding to the microbubble area is set to 1, and a pixel value of a pixel coordinate corresponding to the background area is set to 0.

18. The electronic device according to claim 14 , wherein multiple frames of the second three-dimensional local images comprised in the second three-dimensional local image sequence are obtained by performing the first thinning operation on multiple frames of the first three-dimensional local images in the first three-dimensional local image sequence frame by frame.

Assignments (2)

Priority Claims (1)

Continuity (2)

References Cited (20)