Target object tracking method, tracking system and electronic device for use in radiotherapy

Patent metadata

Field	Value
Publication number	US-12539433-B2
Application number	US-202318157929-A
Country	US
Kind code	B2
Filing date	Jan 23, 2023
Priority date	Jul 27, 2020
Publication date	Feb 3, 2026
Grant date	Feb 3, 2026

Abstract

A tracking method, a tracking system, and an electronic device are provided. The tracking method includes: acquiring an actual scattering image of a target object at time i, where i is an integer greater than 0, and the actual scattering image is generated according to rays scattered by body tissue where the target object is located; processing the actual scattering image or a reference image corresponding to the actual scattering image with a preset model, and determining a location offset of the target object at the time i according to the processing result; and tracking the target object according to the location offset of at least one time. The preset model is indicative of a location conversion relationship of corresponding pixels in images that are formed before and after the rays are scattered.

First claim

What is claimed is:

1 . A tracking method, comprising:
acquiring an actual scattering image of a target object at time i, wherein i is an integer greater than 0, and the actual scattering image is generated according to rays scattered by a body tissue where the target object is located; processing the actual scattering image or a reference image corresponding to the actual scattering image with a preset model, and determining a location offset of the target object at the time i according to the processing result, the preset model being indicative of a location conversion relationship of corresponding pixels in images that are formed before and after the rays are scattered; and tracking the target object according to the location offset of at least one time, the at least one time comprising the time i, wherein the preset model is a deconvolution model, and the processing the actual scattering image or a reference image corresponding to the actual scattering image with a preset model, and determining a location offset of the target object at the time i according to the processing result comprises:
performing deconvolution processing on the actual scattering image by using the deconvolution model to acquire a target image; and
comparing the target image with the reference image, and determining the location offset at the time i according to the comparison result; or
wherein the preset model is a convolution model, and the processing the actual scattering image or a reference image corresponding to the actual scattering image with a preset model, and determining a location offset of the target object at the time i according to the processing result comprises:
performing convolution processing on the reference image by using the convolution model to acquire a reference scattering image; and
comparing the reference scattering image with the actual scattering image, and determining the location offset at the time i according to the comparison result.

2 . The method according to claim 1 , wherein the processing the actual scattering image or a reference image corresponding to the actual scattering image with a preset model, and determining a location offset of the target object at the time i according to the processing result comprises:
acquiring at least one scattering sample point of the actual scattering image and at least one reference sample point of the reference image; and
performing convolution processing by using the convolution model according to the at least one reference sample point to convert the reference image into the target image comparable to the actual scattering image; or, performing deconvolution processing by using the deconvolution model according to the at least one scattering sample point to convert the actual scattering image into the reference scattering image comparable to the reference image.

3 . The method according to claim 1 , wherein the preset model is a probability model; and the method further comprises:
acquiring at least one scattering sample point of the actual scattering image and at least one reference sample point of the reference image; and calculating, according to the at least one scattering sample point and the at least one reference sample point, parameters in the probability model by using Monte Carlo simulation.

4 . The method according to claim 1 , wherein the acquiring an actual scattering image of a target object at time i comprises:
controlling the rays that are collimated to irradiate the body tissue where the target object is located at the time i; and generating the actual scattering image according to a signal converted from the rays scattered on the body tissue.

5 . The method according to claim 4 , wherein the generating the actual scattering image according to a signal converted from the rays scattered on the body tissue comprises:
controlling a detector to receive the rays scattered on the body tissue; receiving the signal sent by the detector, the signal being converted from the rays scattered; and generating the actual scattering image according to the signal.

6 . The method according to claim 5 , wherein the scattered rays are directly received by the detector without beam collimation, or the scattered rays are received by the detector after beam being collimated lower than a preset beam collimation requirement.

7 . The method according to claim 4 , wherein the controlling collimated rays that are collimated to irradiate the body tissue where the target object is located at the time i comprises:
controlling a ray source to emit the rays through a first beam collimator to the body tissue where the target object is located at the time i.

8 . The method according to claim 1 , wherein the acquiring an actual scattering image of a target object at time i comprises:
generating the actual scattering image according to the rays cumulatively scattered by the body tissue where the target object is located between time i-1 and the time i.

9 . An electronic device, comprising a processor and a memory, wherein the processor and the memory are in communication connection, and the memory stores a computer program; and
the processor is configured to execute a tracking method, the tracking method comprising:
acquiring an actual scattering image of a target object at time i, wherein i is an integer greater than 0, and the actual scattering image is generated according to rays scattered by a body tissue where the target object is located;
processing the actual scattering image or a reference image corresponding to the actual scattering image with a preset model, and determining a location offset of the target object at the time i according to the processing result, the preset model being indicative of a location conversion relationship of corresponding pixels in images that are formed before and after the rays are scattered; and
tracking the target object according to the location offset of at least one time, the at least one time comprising the time i, wherein the preset model is a deconvolution model, and the processing the actual scattering image or a reference image corresponding to the actual scattering image with a preset model, and determining a location offset of the target object at the time i according to the processing result comprises:
performing deconvolution processing on the actual scattering image by using the deconvolution model to acquire a target image; and
comparing the target image with the reference image, and determining the location offset at the time i according to the comparison result; or
wherein the preset model is a convolution model, and the processing the actual scattering image or a reference image corresponding to the actual scattering image with a preset model, and determining a location offset of the target object at the time i according to the processing result comprises:
performing convolution processing on the reference image by using the convolution model to acquire a reference scattering image; and
comparing the reference scattering image with the actual scattering image, and determining the location offset at the time i according to the comparison result.

10 . A tracking system, comprising a ray source, a detector and a processor, the detector being electrically connected with the processor, wherein:
the ray source is configured to emit rays to body tissue where a target object is located at time i, wherein i is an integer greater than 0; the detector is configured to receive the rays scattered on the body tissue, convert the rays into a scattering image generation signal, and send the scattering image generation signal to the processor; the processor is configured to generate

Assignees

• Our United Corp

Inventors

• Yan Hao
• Li Jinsheng

Classifications

• A61N2005/1061

  using an x-ray imaging system having a separate imaging source · CPC title

• A61N5/1045

  using a multi-leaf collimator, e.g. for intensity modulated radiation therapy or IMRT · CPC title

• A61N5/1049 Primary

  for verifying the position of the patient with respect to the radiation beam · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title