Re-planning CT image sketching method and system, terminal equipment and storage medium

Abstract

The invention discloses a replanning CT image sketching method and system, terminal equipment and a storage medium, and the method comprises the steps: constructing an image registration deformation field for a replanning CT image and a planning CT image through an image registration algorithm, carrying out the image registration of the replanning CT image through the image registration deformation field according to the planning CT image and a planning CT sketching image, and obtaining a registration image, a deep learning model is used to carry out sketching prediction according to the registration image and the replanning CT image to obtain a sketching prediction image of the replanning CT image, that is, the image registration method is used to sketch the high-precision tumor total volume, and the deep learning method is used to sketch the high-precision organ-endangering characteristic; according to the invention, high-precision sketching of the replanned CT image is realized, the treatment effect of radiotherapy is improved, and the problem of poor radiotherapy effect caused by low organ sketching precision due to the fact that the fine structure change or shape variation of the organ cannot be accurately captured by an image registration method is solved.

Description

Technical Field

[0001] The present invention relates to the field of image delineation, and in particular to a replanning CT image delineation method, system, terminal equipment and storage medium. Background Art

[0002] Radiotherapy is an important part of cancer treatment strategy. Among them, intensity modulated radiotherapy (IMRT) is more popular, and its treatment plan is based on CT simulation scans one to two weeks before treatment. The traditional plan assumes that there are no significant anatomical changes during treatment. But in fact, the gross tumor volume (GTV) and organs at risk (OAR) will change in size and shape during treatment. For example, the parotid gland will shrink during head and neck radiotherapy, which may lead to excessive organ dose or insufficient target area dose. Therefore, in the process of cancer radiotherapy, high-precision delineation results of the gross tumor volume (GTV) and organs at risk (OAR) are urgently needed.

[0003] The existing CT image delineation technology is mainly the image registration method. Image registration refers to the alignment of the position, angle, scale, etc. of different images or images so that they match each other in space. The purpose of medical image registration is to align images from different time points, different modalities, different patients or different imaging devices to achieve accurate comparison, analysis and diagnosis. Image registration methods can accurately identify the changes and boundaries of the gross tumor volume (GTV), making the gross tumor volume (GTV) delineation accurate. However, image registration methods may not accurately capture the subtle structural changes or shape variations of organs, resulting in low delineation accuracy of organs at risk (OAR), which in turn leads to poor radiotherapy effects. Summary of the invention

[0004] The present invention provides a replanning CT image delineation method, system, terminal device and storage medium, which can solve the problem that the image registration method in the prior art may not accurately capture the subtle structural changes or shape variations of the organs, resulting in low delineation accuracy of the organs at risk (OAR), thereby resulting in poor effect of radiotherapy.

[0005] In order to solve the above technical problems, an embodiment of the present invention provides a replanning CT image delineation method, comprising:

[0006] Acquire a CT image of a patient to be radiotherapy; wherein the CT image includes a planned CT image, a planned CT outline image, and a replanned CT image; the planned CT outline image includes a GTV outline image and an OAR outline image;

[0007] According to the planned CT image and the replanned CT image, an image registration deformation field is generated by an image registration algorithm;

[0008] Inputting the planned CT image, the planned CT outline image and the replanned CT image into the image registration deformation field, so that the image registration deformation field generates a registration image and an outline registration image of the replanned CT image according to the planned CT image, the planned CT outline image and the replanned CT image; wherein the outline registration image includes a GTV outline registration image and an OAR outline registration image;

[0009] The registration image, the outline registration image and the replanned CT image are input into the trained replanned CT image outline model, so that the trained replanned CT image outline model performs outline prediction according to the registration image, the outline registration image and the replanned CT image to obtain the outline prediction image of the replanned CT image; the outline prediction image includes the GTV outline prediction image and the OAR outline prediction image.

[0010] Further, after obtaining the CT image of the patient to be radiotherapy, the method further comprises:

[0011] The acquired CT image of the patient to be radiotherapy is subjected to image preprocessing to obtain a processed CT image of the patient to be radiotherapy; wherein the image preprocessing includes format conversion, denoising and smoothing, image scaling and normalization.

[0012] Furthermore, the step of inputting the planned CT image, the planned CT outline image and the re-planned CT image into the image registration deformation field so that the image registration deformation field generates a registration image and an outline registration image of the re-planned CT image according to the planned CT image, the planned CT outline image and the re-planned CT image, comprises:

[0013] Performing image registration on the planned CT image and the re-planned CT image through the image registration deformation field to generate a registration image of the re-planned CT image;

[0014] The planned CT delineation image and the re-planned CT image are registered by using the image registration deformation field to generate a delineation registration image of the re-planned CT image.

[0015] Furthermore, the model training of the replanned CT image delineation model includes:

[0016] Acquire historical CT images of different radiotherapy patients, and perform image preprocessing on the historical CT images of different radiotherapy patients to obtain processed historical CT images of different radiotherapy patients; wherein the historical CT images of different radiotherapy patients include historical planned CT images, historical planned CT delineation images, historical replanned CT images, and historical replanned CT delineation images; the historical planned CT delineation images and historical replanned CT delineation images include historical GTV delineation images and historical OAR delineation images;

[0017] According to the historical planned CT images, the historical planned CT outline images and the historical replanned CT images of each radiotherapy patient, image registration is performed by an image registration algorithm to obtain registered images of different radiotherapy patients; wherein the registered images include the registered images and the outline registered images of the historical replanned CT images;

[0018] Extracting features of the registered images and historical replanning CT images of different radiotherapy patients through the convolutional layer of the encoder in the replanning CT image delineation model to be trained, and obtaining features of the registered images and historical replanning CT images;

[0019] Upsampling the registration image features and the historical replanning CT image features through the upsampling layer of the decoder in the replanning CT image delineation model to be trained to obtain upsampled registration image features and historical replanning CT image features;

[0020] The convolutional layer of the decoder in the replanning CT image delineation model to be trained is used to fuse the upsampled registration image features and the historical replanning CT image features to obtain the replanning CT delineation prediction images of different radiotherapy patients;

[0021] The loss value is calculated by comparing the replanned CT outline prediction images and historical replanned CT outline images of different radiotherapy patients. The parameters of the replanned CT image outline model to be trained are optimized according to the loss value until the loss value converges to obtain the trained replanned CT image outline model.

[0022] Based on the above method embodiment, the present invention provides a corresponding system embodiment;

[0023] An embodiment of the present invention provides a replanning CT image delineation system, comprising: a data acquisition module, a registration deformation field construction module, an image registration module and a delineation prediction module;

[0024] The data acquisition module is used to obtain CT images of patients to be treated with radiotherapy; wherein the CT images include planned CT images, planned CT outline images, and replanned CT images; the planned CT outline images include GTV outline images and OAR outline images;

[0025] The registration deformation field construction module is used to generate an image registration deformation field through an image registration algorithm according to the planned CT image and the replanned CT image;

[0026] The image registration module is used to input the planned CT image, the planned CT outline image and the re-planned CT image into the image registration deformation field, so that the image registration deformation field generates a registration image and an outline registration image of the re-planned CT image according to the planned CT image, the planned CT outline image and the re-planned CT image; wherein the outline registration image includes a GTV outline registration image and an OAR outline registration image;

[0027] The delineation prediction module is used to input the registration image, the delineation registration image and the replanning CT image into the trained replanning CT image delineation model, so that the trained replanning CT image delineation model performs delineation prediction according to the registration image, the delineation registration image and the replanning CT image to obtain the delineation prediction image of the replanning CT image; the delineation prediction image includes the GTV delineation prediction image and the OAR delineation prediction image.

[0028] Furthermore, after the data acquisition module, it also includes: an image preprocessing module;

[0029] The image preprocessing module is used to perform image preprocessing on the acquired CT image of the patient to be radiotherapy, so as to obtain the processed CT image of the patient to be radiotherapy; wherein the image preprocessing includes format conversion, denoising and smoothing, image scaling and normalization.

[0030] Further, the image registration module includes: a registration image unit and a delineation registration image unit;

[0031] The registration image unit is used to perform image registration on the planned CT image and the re-planned CT image through the image registration deformation field to generate a registration image of the re-planned CT image;

[0032] The delineation and registration image unit is used to perform image registration on the planned CT delineation image and the re-planned CT image through the image registration deformation field to generate a delineation and registration image of the re-planned CT image.

[0033] Based on the above-mentioned method embodiment, the present invention provides a corresponding terminal device embodiment, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, a re-planning CT image delineation method as described in the present invention is implemented.

[0034] Based on the above method embodiment, the present invention provides a corresponding computer-readable storage medium embodiment, including: a stored computer program, which controls the device where the computer-readable storage medium is located to execute a re-planning CT image delineation method as described in the present invention when the computer program is running.

[0035] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0036] The present invention obtains a re-planned CT image, a planned CT image, a GTV delineation image of a planned CT delineation image, and an OAR delineation image of a planned CT delineation image of a patient to be radiotherapy, wherein the planned CT delineation image includes, and generates an image registration deformation field through an image registration algorithm according to the planned CT image and the re-planned CT image, and inputs the planned CT image, the planned CT delineation image, and the re-planned CT image into the image registration deformation field, so that the image registration deformation field generates a registration image of the re-planned CT image, a GTV delineation registration image, and an OAR delineation registration image according to the planned CT image, the planned CT delineation image, and the re-planned CT image, and then inputs the re-planned CT image and the registration image of the re-planned CT image, the GTV delineation registration image, and the OAR delineation registration image into a trained re-planned CT image delineation model, so that the trained re-planned CT image delineation model performs delineation prediction according to the registration image, the delineation registration image, and the re-planned CT image, to obtain a GTV delineation prediction image and an OAR delineation prediction image of the re-planned CT image. That is, the present invention utilizes the characteristics of the image registration method to delineate the high-precision gross tumor volume (GTV), and utilizes the characteristics of the deep learning method to delineate the high-precision organs at risk (OAR) because it can generalize and identify the contours of organs. The delineation prediction of GTV and OAR is performed on the registered image data output by the image registration method, which not only achieves high-precision delineation of the GTV of the replanned CT image, but also achieves high-precision delineation of the OAR of the replanned CT image, thereby improving the therapeutic effect of radiotherapy, and solving the problem that the image registration method in the prior art may not be able to accurately capture the subtle structural changes or shape variations of the organs, resulting in low delineation accuracy of the organs at risk (OAR), thereby resulting in poor effect of radiotherapy. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 : A flowchart of the steps of a replanning CT image delineation method provided by an embodiment of the present invention;

[0038] Figure 2 : A system structure diagram of a replanning CT image delineation system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] In the description of the present invention, it should be understood that the terms "first", "second" and "third" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features.

[0041] Embodiment 1:

[0042] Reference Figure 1 :A flow chart of the steps of a replanning CT image delineation method provided by an embodiment of the present invention, the method at least comprises the following steps:

[0043] Step S1: Acquire a CT image of a patient to be radiotherapy; wherein the CT image includes a planned CT image, a planned CT outline image, and a replanned CT image; the planned CT outline image includes a GTV outline image and an OAR outline image;

[0044] In this embodiment, since CT images are usually stored in DI COM format, in order to facilitate image processing and analysis, it is necessary to perform image preprocessing on the acquired CT images of the patient to be radiotherapy to obtain the processed CT images of the patient to be radiotherapy; wherein the image preprocessing includes but is not limited to format conversion, denoising and smoothing, image scaling and normalization; the image denoising and smoothing processing includes but is not limited to mean filtering, median filtering, Gaussian filtering and bilateral filtering; the image scaling includes but is not limited to nearest neighbor interpolation, bilinear interpolation and bicubic interpolation.

[0045] Step S2: generating an image registration deformation field through an image registration algorithm according to the planned CT image and the replanned CT image;

[0046] In this embodiment, the image registration algorithm includes but is not limited to a rigid registration algorithm, an affine registration algorithm, and a nonlinear transformation registration algorithm.

[0047] Step S3: inputting the planned CT image, the planned CT outline image and the re-planned CT image into the image registration deformation field, so that the image registration deformation field generates a registration image and an outline registration image of the re-planned CT image according to the planned CT image, the planned CT outline image and the re-planned CT image; wherein the outline registration image includes a GTV outline registration image and an OAR outline registration image;

[0048] In this embodiment, generating a registration image of the replanned CT image and outlining the registration image include:

[0049] Performing image registration on the planned CT image and the re-planned CT image through the image registration deformation field to generate a registration image of the re-planned CT image;

[0050] The planned CT delineation image and the re-planned CT image are registered by using the image registration deformation field to generate a delineation registration image of the re-planned CT image.

[0051] Step S4: input the registration image, the outline registration image and the replanned CT image into the trained replanned CT image outline model, so that the trained replanned CT image outline model performs outline prediction according to the registration image, the outline registration image and the replanned CT image to obtain the outline prediction image of the replanned CT image; the outline prediction image includes the GTV outline prediction image and the OAR outline prediction image.

[0052] In this embodiment, the re-planning CT image delineation model may adopt the NNUNetv2 neural network model.

[0053] In this embodiment, the model training of the replanned CT image delineation model includes:

[0054] Acquire historical CT images of different radiotherapy patients, and perform image preprocessing on the historical CT images of different radiotherapy patients to obtain processed historical CT images of different radiotherapy patients; wherein the historical CT images of different radiotherapy patients include historical planned CT images, historical planned CT delineation images, historical replanned CT images, and historical replanned CT delineation images; the historical planned CT delineation images and historical replanned CT delineation images include historical GTV delineation images and historical OAR delineation images;

[0055] In this embodiment, after acquiring historical CT images of different radiotherapy patients, image preprocessing is performed on the acquired historical CT images of different radiotherapy patients to obtain processed historical CT images of different radiotherapy patients; wherein the image preprocessing includes format conversion, denoising and smoothing, image scaling and normalization; the image denoising and smoothing processing includes but is not limited to mean filtering, median filtering, Gaussian filtering and bilateral filtering; the image scaling includes but is not limited to nearest neighbor interpolation, bilinear interpolation and bicubic interpolation;

[0056] According to the historical planned CT images, the historical planned CT outline images and the historical replanned CT images of each radiotherapy patient, image registration is performed by an image registration algorithm to obtain registered images of different radiotherapy patients; wherein the registered images include the registered images and the outline registered images of the historical replanned CT images;

[0057] Extracting features of the registered images and historical replanning CT images of different radiotherapy patients through the convolutional layer of the encoder in the replanning CT image delineation model to be trained, and obtaining features of the registered images and historical replanning CT images;

[0058] Upsampling the registration image features and the historical replanning CT image features through the upsampling layer of the decoder in the replanning CT image delineation model to be trained to obtain upsampled registration image features and historical replanning CT image features;

[0059] The convolutional layer of the decoder in the replanning CT image delineation model to be trained is used to fuse the upsampled registration image features and the historical replanning CT image features to obtain the replanning CT delineation prediction images of different radiotherapy patients;

[0060] The loss value is calculated by comparing the replanned CT outline prediction images and historical replanned CT outline images of different radiotherapy patients. The parameters of the replanned CT image outline model to be trained are optimized according to the loss value until the loss value converges to obtain the trained replanned CT image outline model.

[0061] In this embodiment, the loss value is calculated by comparing the re-planned CT outline prediction images and historical re-planned CT outline images of different radiotherapy patients through a loss function, and based on the loss value, the parameters of the re-planned CT image outline model to be trained are optimized by back propagation until the loss value converges to obtain a trained re-planned CT image outline model; wherein the loss function includes but is not limited to a cross entropy loss function and a Di ce loss function.

[0062] In this embodiment, the present invention obtains a re-planned CT image of a patient to be radiotherapy, a planned CT image, a GTV outline image of the planned CT outline image, and an OAR outline image of the planned CT outline image, wherein the planned CT outline image includes, and generates an image registration deformation field through an image registration algorithm according to the planned CT image and the re-planned CT image, and inputs the planned CT image, the planned CT outline image, and the re-planned CT image into the image registration deformation field, so that the image registration deformation field generates a registration image of the re-planned CT image, a GTV outline registration image, and an OAR outline registration image according to the planned CT image, the planned CT outline image, and the re-planned CT image, and then registers the re-planned CT image and the registration image of the re-planned CT image, the GTV outline registration image, and the OAR outline registration image The image is input into the trained replanning CT image delineation model, so that the trained replanning CT image delineation model performs delineation prediction according to the registration image, the delineated registration image and the replanning CT image, and obtains the GTV delineation prediction image and the OAR delineation prediction image of the replanning CT image, that is, the present invention utilizes the image registration method to delineate the characteristics of the high-precision gross tumor volume (GTV), and utilizes the deep learning method to delineate the characteristics of the high-precision organs at risk (OAR) because it can generalize and identify the contours of organs, and performs GTV and OAR delineation prediction on the registered image data output by the image registration method, which not only realizes the high-precision delineation of the GTV of the replanning CT image, but also realizes the high-precision delineation of the OAR of the replanning CT image, thereby improving the therapeutic effect of radiotherapy.

[0063] Embodiment 2:

[0064] Reference Figure 2 : A system structure diagram of a replanning CT image delineation system provided by an embodiment of the present invention, the system comprises: a data acquisition module, a registration deformation field construction module, an image registration module and a delineation prediction module;

[0065] The data acquisition module is used to obtain CT images of patients to be treated with radiotherapy; wherein the CT images include planned CT images, planned CT outline images, and replanned CT images; the planned CT outline images include GTV outline images and OAR outline images;

[0066] The registration deformation field construction module is used to generate an image registration deformation field through an image registration algorithm according to the planned CT image and the replanned CT image;

[0067] The image registration module is used to input the planned CT image, the planned CT outline image and the re-planned CT image into the image registration deformation field, so that the image registration deformation field generates a registration image and an outline registration image of the re-planned CT image according to the planned CT image, the planned CT outline image and the re-planned CT image; wherein the outline registration image includes a GTV outline registration image and an OAR outline registration image;

[0068] The delineation prediction module is used to input the registration image, the delineation registration image and the replanning CT image into the trained replanning CT image delineation model, so that the trained replanning CT image delineation model performs delineation prediction according to the registration image, the delineation registration image and the replanning CT image to obtain the delineation prediction image of the replanning CT image; the delineation prediction image includes the GTV delineation prediction image and the OAR delineation prediction image.

[0069] In this embodiment, the image registration algorithm includes but is not limited to a rigid registration algorithm, an affine registration algorithm, and a nonlinear transformation registration algorithm.

[0070] In this embodiment, the re-planning CT image delineation model may adopt the NNUNetv2 neural network model.

[0071] In this embodiment, after the data acquisition module, it also includes: an image preprocessing module;

[0072] The image preprocessing module is used to perform image preprocessing on the CT image of the patient to be radiotherapy, so as to obtain the processed CT image of the patient to be radiotherapy; wherein the image preprocessing includes but is not limited to format conversion, denoising and smoothing, image scaling and normalization; the image denoising and smoothing processing includes but is not limited to mean filtering, median filtering, Gaussian filtering and bilateral filtering; the image scaling includes but is not limited to nearest neighbor interpolation, bilinear interpolation and bicubic interpolation.

[0073] In this embodiment, the image registration module includes: a registration image unit and a delineation registration image unit;

[0074] The registration image unit is used to perform image registration on the planned CT image and the re-planned CT image through the image registration deformation field to generate a registration image of the re-planned CT image;

[0075] The delineation and registration image unit is used to perform image registration on the planned CT delineation image and the re-planned CT image through the image registration deformation field to generate a delineation and registration image of the re-planned CT image.

[0076] In this embodiment, the delineation prediction module includes: a delineation prediction model training submodule; wherein the delineation prediction model training submodule includes a training data acquisition unit, a training data registration unit and a delineation prediction model training unit; the delineation prediction model training unit includes a first training subunit, a second training subunit and a third training subunit;

[0077] The training data acquisition unit is used to obtain historical CT images of different radiotherapy patients, and perform image preprocessing on the historical CT images of different radiotherapy patients to obtain processed historical CT images of different radiotherapy patients; wherein the historical CT images of different radiotherapy patients include historical planned CT images, historical planned CT delineation images, historical replanned CT images and historical replanned CT delineation images; the historical planned CT delineation images and historical replanned CT delineation images include historical GTV delineation images and historical OAR delineation images;

[0078] The training data registration unit is used to perform image registration based on the historical planned CT image, the historical planned CT outline image and the historical replanned CT image of each radiotherapy patient through an image registration algorithm to obtain registered images of different radiotherapy patients; wherein the registered images include the registered images and the outline registered images of the historical replanned CT images;

[0079] The first training subunit is used to extract features from the registered images and historical replanning CT images of different radiotherapy patients through the convolutional layer of the encoder in the replanning CT image delineation model to be trained, so as to obtain the registered image features and the historical replanning CT image features;

[0080] The second training subunit is used to upsample the registration image features and the historical replanning CT image features through the upsampling layer of the decoder in the replanning CT image delineation model to be trained to obtain the upsampled registration image features and the historical replanning CT image features, and perform feature fusion on the upsampled registration image features and the historical replanning CT image features through the convolution layer of the decoder in the replanning CT image delineation model to be trained to obtain the replanning CT delineation prediction images of different radiotherapy patients;

[0081] The third training subunit is used to calculate the loss value by comparing the replanned CT outline prediction images and historical replanned CT outline images of different radiotherapy patients, and optimize the parameters of the replanned CT image outline model to be trained according to the loss value until the loss value converges to obtain the trained replanned CT image outline model.

[0082] In this embodiment, a loss value is calculated by comparing the re-planned CT outline prediction images and historical re-planned CT outline images of different radiotherapy patients through a loss function, and based on the loss value, the parameters of the re-planned CT image outline model to be trained are optimized by back propagation until the loss value converges to obtain a trained re-planned CT image outline model; wherein the loss function includes but is not limited to a cross entropy loss function and a Dice loss function.

[0083] Based on the above method embodiment, another embodiment is provided;

[0084] Another embodiment of the present invention provides a re-planning CT image delineation terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, a re-planning CT image delineation method described in any one of the above-mentioned method embodiments of the present invention is implemented.

[0085] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program in the replanning CT image delineation terminal device.

[0086] The re-planning CT image delineation terminal device may be a computing device such as a desktop computer, a notebook, a PDA, and a cloud server. The re-planning CT image delineation terminal device may include, but is not limited to, a processor and a memory. Those skilled in the art may understand that, for example, the re-planning CT image delineation terminal device may also include an input and output device, a network access device, a bus, etc.

[0087] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the replanning CT image delineation terminal device, and uses various interfaces and lines to connect various parts of the entire replanning CT image delineation terminal device.

[0088] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the re-planning CT image delineation terminal device by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Med ia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0089] Based on the above method embodiment, another embodiment is provided;

[0090] Another embodiment of the present invention provides a storage medium, comprising a stored computer program, wherein when the computer program is executed, the device where the storage medium is located is controlled to execute a replanning CT image delineation method described in any one of the above method embodiments of the present invention.

[0091] Wherein, the above-mentioned storage medium is a computer-readable storage medium. The module / unit integrated in the replanning CT image delineation system / terminal device, if implemented in the form of a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on such an understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned method embodiments can be implemented. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0092] It should be noted that the above-mentioned terminal device may include, but is not limited to, a processor and a memory. Those skilled in the art will understand that the above-mentioned terminal device is merely an example and does not constitute a limitation on the terminal device. It may include more or fewer components, or a combination of certain components, or different components.

[0093] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A replanning CT image delineation method, characterized in that: include: Acquire a CT image of a patient to be radiotherapy; wherein the CT image includes a planned CT image, a planned CT outline image, and a replanned CT image; the planned CT outline image includes a GTV outline image and an OAR outline image; According to the planned CT image and the replanned CT image, an image registration deformation field is generated by an image registration algorithm; Inputting the planned CT image, the planned CT outline image and the replanned CT image into the image registration deformation field, so that the image registration deformation field generates a registration image and an outline registration image of the replanned CT image according to the planned CT image, the planned CT outline image and the replanned CT image; wherein the outline registration image includes a GTV outline registration image and an OAR outline registration image; The registration image, the outline registration image and the replanned CT image are input into the trained replanned CT image outline model, so that the trained replanned CT image outline model performs outline prediction according to the registration image, the outline registration image and the replanned CT image to obtain the outline prediction image of the replanned CT image; the outline prediction image includes the GTV outline prediction image and the OAR outline prediction image.

2. A replanning CT image delineation method according to claim 1, characterized in that: After obtaining the CT image of the patient to be radiotherapy, the method includes: The acquired CT image of the patient to be radiotherapy is subjected to image preprocessing to obtain a processed CT image of the patient to be radiotherapy; wherein the image preprocessing includes format conversion, denoising and smoothing, image scaling and normalization.

3. A replanning CT image delineation method according to claim 2, characterized in that: The step of inputting the planned CT image, the planned CT outline image and the re-planned CT image into the image registration deformation field so that the image registration deformation field generates a registration image and an outline registration image of the re-planned CT image according to the planned CT image, the planned CT outline image and the re-planned CT image, comprises: Performing image registration on the planned CT image and the re-planned CT image through the image registration deformation field to generate a registration image of the re-planned CT image; The planned CT delineation image and the re-planned CT image are registered by using the image registration deformation field to generate a delineation registration image of the re-planned CT image.

4. A replanning CT image delineation method according to claim 3, characterized in that: The model training of the replanned CT image delineation model comprises: Acquire historical CT images of different radiotherapy patients, and perform image preprocessing on the historical CT images of different radiotherapy patients to obtain processed historical CT images of different radiotherapy patients; wherein the historical CT images of different radiotherapy patients include historical planned CT images, historical planned CT delineation images, historical replanned CT images, and historical replanned CT delineation images; the historical planned CT delineation images and historical replanned CT delineation images include historical GTV delineation images and historical OAR delineation images; According to the historical planned CT images, the historical planned CT outline images and the historical replanned CT images of each radiotherapy patient, image registration is performed by an image registration algorithm to obtain registered images of different radiotherapy patients; wherein the registered images include the registered images and the outline registered images of the historical replanned CT images; Extracting features of the registered images and historical replanning CT images of different radiotherapy patients through the convolutional layer of the encoder in the replanning CT image delineation model to be trained, and obtaining features of the registered images and historical replanning CT images; Upsampling the registration image features and the historical replanning CT image features through the upsampling layer of the decoder in the replanning CT image delineation model to be trained to obtain upsampled registration image features and historical replanning CT image features; The convolutional layer of the decoder in the replanning CT image delineation model to be trained is used to fuse the upsampled registration image features and the historical replanning CT image features to obtain the replanning CT delineation prediction images of different radiotherapy patients; The loss value is calculated by comparing the replanned CT outline prediction images and historical replanned CT outline images of different radiotherapy patients. The parameters of the replanned CT image outline model to be trained are optimized according to the loss value until the loss value converges to obtain the trained replanned CT image outline model.

5. A replanning CT image delineation system, characterized in that: include: Data acquisition module, registration deformation field construction module, image registration module and delineation prediction module; The data acquisition module is used to obtain CT images of patients to be treated with radiotherapy; wherein the CT images include planned CT images, planned CT outline images, and replanned CT images; the planned CT outline images include GTV outline images and OAR outline images; The registration deformation field construction module is used to generate an image registration deformation field through an image registration algorithm according to the planned CT image and the replanned CT image; The image registration module is used to input the planned CT image, the planned CT outline image and the re-planned CT image into the image registration deformation field, so that the image registration deformation field generates a registration image and an outline registration image of the re-planned CT image according to the planned CT image, the planned CT outline image and the re-planned CT image; wherein the outline registration image includes a GTV outline registration image and an OAR outline registration image; The delineation prediction module is used to input the registration image, the delineation registration image and the replanning CT image into the trained replanning CT image delineation model, so that the trained replanning CT image delineation model performs delineation prediction according to the registration image, the delineation registration image and the replanning CT image to obtain the delineation prediction image of the replanning CT image; the delineation prediction image includes the GTV delineation prediction image and the OAR delineation prediction image.

6. A replanning CT image delineation system according to claim 5, characterized in that: After the data acquisition module, it also includes: an image preprocessing module; The image preprocessing module is used to perform image preprocessing on the acquired CT image of the patient to be radiotherapy, so as to obtain the processed CT image of the patient to be radiotherapy; wherein the image preprocessing includes format conversion, denoising and smoothing, image scaling and normalization.

7. A replanning CT image delineation system according to claim 6, characterized in that: The image registration module comprises: a registration image unit and a delineation registration image unit; The registration image unit is used to perform image registration on the planned CT image and the re-planned CT image through the image registration deformation field to generate a registration image of the re-planned CT image; The delineation and registration image unit is used to perform image registration on the planned CT delineation image and the re-planned CT image through the image registration deformation field to generate a delineation and registration image of the re-planned CT image.

8. A replanning CT image delineation system according to claim 7, characterized in that: The delineation prediction module comprises: a delineation prediction model training submodule; wherein the delineation prediction model training submodule comprises a training data acquisition unit, a training data registration unit and a delineation prediction model training unit; the delineation prediction model training unit comprises a first training subunit, a second training subunit and a third training subunit; The training data acquisition unit is used to obtain historical CT images of different radiotherapy patients, and perform image preprocessing on the historical CT images of different radiotherapy patients to obtain processed historical CT images of different radiotherapy patients; wherein the historical CT images of different radiotherapy patients include historical planned CT images, historical planned CT delineation images, historical replanned CT images and historical replanned CT delineation images; the historical planned CT delineation images and historical replanned CT delineation images include historical GTV delineation images and historical OAR delineation images; The training data registration unit is used to perform image registration based on the historical planned CT image, the historical planned CT outline image and the historical replanned CT image of each radiotherapy patient through an image registration algorithm to obtain registered images of different radiotherapy patients; wherein the registered images include the registered images and the outline registered images of the historical replanned CT images; The first training subunit is used to extract features from the registered images and historical replanning CT images of different radiotherapy patients through the convolutional layer of the encoder in the replanning CT image delineation model to be trained, so as to obtain the registered image features and the historical replanning CT image features; The second training subunit is used to upsample the registration image features and the historical replanning CT image features through the upsampling layer of the decoder in the replanning CT image delineation model to be trained to obtain the upsampled registration image features and the historical replanning CT image features, and perform feature fusion on the upsampled registration image features and the historical replanning CT image features through the convolution layer of the decoder in the replanning CT image delineation model to be trained to obtain the replanning CT delineation prediction images of different radiotherapy patients; The third training subunit is used to calculate the loss value by comparing the replanned CT outline prediction images and historical replanned CT outline images of different radiotherapy patients, and optimize the parameters of the replanned CT image outline model to be trained according to the loss value until the loss value converges to obtain the trained replanned CT image outline model.

9. A re-planning CT image delineation terminal device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, a re-planning CT image delineation method as described in any one of claims 1 to 4 is implemented.

10. A storage medium, characterized in that: The storage medium includes a stored computer program, wherein when the computer program is executed, the device where the storage medium is located is controlled to execute a replanning CT image delineation method as claimed in any one of claims 1 to 4.

Images