A system and storage medium for determining a tumor target area contour set after drug treatment

Abstract

The present invention discloses a system and storage medium for determining a tumor target area contour set after drug treatment, and relates to the field of medical devices. The system comprises: a first acquisition module, used to acquire a tumor target area contour set before drug treatment and a first contour set related to the organ where the tumor is located; a second acquisition module, used to acquire a second contour set related to the organ where the tumor is located after drug treatment; a deformation field determination module, used to determine a deformation field capable of converting the first contour set into the second contour set based on the first contour set and the second contour set; a target area contour determination module, used to determine a tumor target area contour set after drug treatment based on the deformation field and the tumor target area contour set before drug treatment. The present invention can accurately determine and outline the tumor target area after drug treatment.

Description

Technical Field

[0001] The present invention relates to the field of medical devices, and in particular to a system and a storage medium for determining a tumor target area contour set after drug treatment. Background Art

[0002] Figure 1 It is a flowchart of tumor treatment for patients after drug treatment using related technologies, such as Figure 1 As shown, the steps include:

[0003] Step S101: Obtain a medical image set before drug treatment.

[0004] Step S102: Acquire a medical image set after drug treatment.

[0005] Step S103: Determine a tumor target area contour set according to clinical purposes, such as conventional radiotherapy, radioactive particle implantation therapy (referred to as particle implantation therapy) and other close-range radiotherapy, ablation therapy, or diagnostic puncture biopsy (hereinafter referred to as puncture).

[0006] For example, if the clinical purpose is conventional radiotherapy or radioactive seed implantation therapy, a positioning image is required, and the tumor target contour set is determined on the positioning image of the medical image set after drug treatment. For another example, if the clinical purpose is ablation therapy or puncture biopsy, a guiding image is required, and the tumor target contour set is determined on the guiding image of the medical image set after drug treatment.

[0007] The tumor target area refers to the range where clinical operations are performed, for example, it may refer to the range used for conventional radiotherapy, or it may refer to the range that requires close-range radiotherapy or ablation therapy based on experience, or the range where cancer cells may exist and puncture is performed (to confirm the pathological diagnosis by biopsy).

[0008] Step S104: Based on the tumor target area contour set, subsequent clinical operations are performed, such as conventional radiotherapy, brachytherapy, ablation therapy, or puncture.

[0009] Among them, the tumor target area can be automatically determined by the machine in the form of outlining, or it can be determined by the doctor based on experience. If the patient has received chemotherapy, endocrine therapy or other drug treatments before treatment (including conventional radiotherapy, particle implantation and other close-range radiotherapy or ablation therapy, etc.) or puncture, then the tumor target area after drug treatment will be diffuse on medical images such as magnetic resonance images, and the contour boundary of the tumor target area will be difficult to clearly and accurately identify. In order to overcome the problem that the contour of the tumor target area cannot be accurately identified due to the diffuse image of the tumor target area after drug treatment, the following methods are often used clinically:

[0010] Method 1: Based on the contour and position of the tumor in the image before drug treatment, the contour range of the tumor after treatment is determined on the image after treatment by manual methods and personal experience. This method will lead to inaccurate tumor target area irradiation range or puncture site;

[0011] Method 2: When the treatment (including conventional radiotherapy, brachytherapy such as particle implantation, or ablation therapy) is implemented, the intended irradiation range of the tumor target is expanded to the entire anatomical part (i.e., the organ where the tumor is located) containing the tumor target to prevent missed irradiation. For example, after drug treatment of prostate cancer, the entire prostate cancer contour is generally used as the intended irradiation range of the tumor target. In fact, the volume of the irradiation range is often much larger than the volume of the tumor target, and the distance between the irradiation range and the relevant organs at risk is closer or even overlaps. This will lead to two problems: First, when designing the prescription dose and fractionated dose, considering the dose limit of the relevant organs at risk, clinical practice often dare not increase the prescription dose and fractionated dose of the irradiation range, resulting in the inability to effectively achieve the purpose of radical radiotherapy; second, when designing the treatment plan, due to the large irradiation range, it is difficult to effectively reduce the irradiation dose of the relevant organs at risk, resulting in an increased probability of organ-threatening toxic and side effects during the treatment implementation. When performing a puncture biopsy, the puncture range is expanded to the entire organ for multiple punctures. This method will increase the risk of bleeding and infection in patients, etc. Summary of the invention

[0012] The embodiment of the present invention provides a system and storage medium for determining a tumor target area contour set after drug treatment, aiming to solve the problem in the prior art that the tumor target area is difficult to accurately identify and outline after drug treatment.

[0013] An embodiment of the present invention provides a system for determining a tumor target area contour set after drug treatment, the system comprising: a first acquisition module, used to acquire the tumor target area contour set before drug treatment and a first contour set related to the organ where the tumor is located; a second acquisition module, used to acquire a second contour set related to the organ where the tumor is located after drug treatment; a deformation field determination module, used to determine a deformation field that can convert the first contour set into the second contour set based on the first contour set and the second contour set; and a target area contour determination module, used to determine the tumor target area contour set after drug treatment based on the deformation field and the tumor target area contour set before drug treatment.

[0014] Preferably, the first contour set related to the organ where the tumor is located before drug treatment acquired by the first acquisition module includes at least one of contour information of a subregion of the organ before drug treatment and contour information of a combination of adjacent subregions of the organ before drug treatment; the second contour set related to the organ where the tumor is located after drug treatment acquired by the second acquisition module includes at least one of contour information of a subregion of the organ after drug treatment and contour information of a combination of adjacent subregions of the organ after drug treatment.

[0015] Preferably, the deformation field determination module is specifically used to select at least the contour information of the sub-region where the tumor is located before drug treatment from the first contour set; correspondingly, to select at least the contour information of the sub-region where the tumor is located after drug treatment from the second contour set; and to use the deformation registration model to perform deformation registration on the selected contour information to obtain the deformation field.

[0016] Preferably, the target area contour determination module is specifically used to transform the tumor target area contour set before drug treatment by using the deformation field, and obtain the transformed tumor target area contour set as the tumor target area contour set after drug treatment.

[0017] Preferably, the system further comprises: an output module, configured to output the tumor target area contour set after drug treatment, so as to implement corresponding clinical operations based on the output tumor target area contour set after drug treatment.

[0018] An embodiment of the present invention provides a tumor treatment control system after drug treatment, and the system includes the aforementioned system for determining a tumor target area contour set after drug treatment.

[0019] An embodiment of the present invention provides a storage medium, on which a program is stored. When the program is executed by a processor, the following steps are implemented: obtaining a tumor target area contour set before drug treatment and a first contour set related to the organ where the tumor is located; obtaining a second contour set related to the organ where the tumor is located after drug treatment; determining a deformation field capable of converting the first contour set into the second contour set based on the first contour set and the second contour set; and determining the tumor target area contour set after drug treatment based on the deformation field and the tumor target area contour set before drug treatment.

[0020] Preferably, the first contour set related to the organ where the tumor is located before the drug treatment includes at least one of the contour information of a subregion of the organ before the drug treatment and the contour information of a combination of adjacent subregions of the organ before the drug treatment; the second contour set related to the organ where the tumor is located after the drug treatment includes at least one of the contour information of a subregion of the organ after the drug treatment and the contour information of a combination of adjacent subregions of the organ after the drug treatment.

[0021] Preferably, the step of determining a deformation field capable of converting the first contour set into the second contour set based on the first contour set and the second contour set implemented when the program is executed by the processor includes: selecting at least contour information of the sub-region where the tumor is located before drug treatment from the first contour set; correspondingly, selecting at least contour information of the sub-region where the tumor is located after drug treatment from the second contour set; and using a deformation registration model to perform deformation registration on the selected contour information to obtain the deformation field.

[0022] Preferably, the step of determining the tumor target contour set after drug treatment based on the deformation field and the tumor target contour set before drug treatment implemented when the program is executed by the processor includes: using the deformation field to transform the tumor target contour set before drug treatment, and obtaining the transformed tumor target contour set as the tumor target contour set after drug treatment.

[0023] Preferably, when the program is executed by the processor, the following steps are further implemented: outputting the tumor target area contour set after drug treatment, so as to implement corresponding clinical operations based on the output tumor target area contour set after drug treatment.

[0024] An embodiment of the present invention provides a system and storage medium for determining a tumor target area contour set after drug treatment. The system determines a deformation field capable of converting a first contour set related to an organ where the tumor is located before drug treatment and a second contour set related to the organ where the tumor is located after drug treatment, and determines a tumor target area contour set after drug treatment based on the deformation field and the tumor target area contour set before drug treatment. In this way, the tumor target area after drug treatment determined by the tumor target area contour set after drug treatment is more accurate, and can not only effectively and accurately locate the tumor range clinically during conventional radiotherapy, close-range radiotherapy, ablation therapy, or puncture, but also reduce the probability of various side effects during implementation. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is a flowchart of the related technology for treating tumors in patients after drug treatment;

[0026] Figure 2a is a structural block diagram of a system for determining a tumor target area contour set after drug treatment provided by an embodiment of the present invention;

[0027] Figure 2b is based on Figure 2a A flowchart of the system for treating tumors in patients after drug treatment;

[0028] Figure 2c and 2dis the MRI tumor target contour before treatment, using Figure 2a The first and second comparison images of the tumor target area contour after drug treatment determined by the system and the postoperative pathological section contour (gold standard);

[0029] Figure 3 is a conventional radiotherapy flow chart for a patient after drug treatment provided by an embodiment of the present invention;

[0030] Figure 4 is a flow chart of generating a tumor target area contour set after drug treatment provided by an embodiment of the present invention;

[0031] Figure 5a are an original image and a target image for deformable registration provided by an embodiment of the present invention;

[0032] Figure 5b is a deformation field established between an original image and a target image provided by an embodiment of the present invention;

[0033] Figure 6a , 6b , 6c are small field high-definition T2WI sequence images, DWI sequence images, and ADC sequence images before endocrine therapy, respectively;

[0034] Figure 7a , 7b , 7c are small field high-definition T2WI sequence images, DWI sequence images, and ADC sequence images after endocrine therapy, respectively;

[0035] Figure 8a , 8b 8c, 8d are the sub-region sets and tumor volume maps before endocrine therapy, the sub-region set volume maps after endocrine therapy, the sub-region sets and deformed tumor contour maps after endocrine therapy, and the sub-region sets and deformed tumor volume maps after endocrine therapy, respectively;

[0036] Figure 9a , 9b , 9c are small field high-definition T2WI sequence images with Tb contour set after endocrine therapy, DWI sequence images, and ADC sequence images, respectively;

[0037] Fig.10a , 10b 10c and 10d are respectively a CT image dedicated to radiotherapy planning, an original MRI image after color adjustment before fusion with CT, a fusion image of CT and MRI, and an MRI image after registration with CT. DETAILED DESCRIPTION

[0038] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0039] In the subsequent description, the suffixes such as "module", "component" or "unit" used to represent elements are only used to facilitate the description of the present invention, and have no special meanings. Therefore, "module", "component" or "unit" can be used in a mixed manner.

[0040] The present invention proposes a system for generating a set of tumor target area contours after drug treatment. The system can use the contours of the organ where the tumor is located before and after drug treatment, the contours of each sub-region of the organ, and the contours of the combination of each sub-region to establish a conversion relationship between the contours before and after drug treatment, and then use the conversion relationship to process the set of tumor target area contours before drug treatment to obtain the set of tumor target area contours after drug treatment, thereby solving the problem that the tumor target area is difficult to accurately identify and outline after drug treatment.

[0041] Figure 2a is a structural block diagram of a system for determining a tumor target area contour set after drug treatment provided by an embodiment of the present invention, such as Figure 2a As shown, the system may include: a first acquisition module 10 , a second acquisition module 20 , a deformation field determination module 30 , and a target area contour determination module 40 .

[0042] The first acquisition module 10 is used to acquire a tumor target area contour set before drug treatment and a first contour set related to the organ where the tumor is located.

[0043] The first contour set may include at least one of the contour information of the subregion of the organ before drug treatment and the contour information of the combination of adjacent subregions of the organ before drug treatment. In specific implementation, the contour information before drug treatment can be obtained as needed, and the present invention does not limit this. For example, if the tumor is distributed in a certain subregion, only the contour information of the subregion where the tumor is located can be obtained, or the contour information of each subregion of the organ can be obtained. If the subregion contour is unclear and difficult to distinguish, the contour information of the combination of adjacent subregions can be obtained to obtain a recognizable contour.

[0044] In addition, the first contour set may also include contour information of other tissues or organs related to the anatomical structure of the organ where the tumor is located, such as contour information of the seminal vesicle that is closely related to the anatomical structure of the prostate.

[0045] The second acquisition module 20 is used to acquire a second contour set related to the organ where the tumor is located after drug treatment.

[0046] The second contour set may include at least one of the contour information of the subregion of the organ after drug treatment and the contour information of the combination of adjacent subregions of the organ after drug treatment. In specific implementation, the contour information after drug treatment can be obtained as needed, and the present invention does not limit this. For example, if the tumor is distributed in a certain subregion, only the contour information of the subregion where the tumor is located can be obtained, or the contour information of each subregion of the organ can be obtained. If the subregion contour is unclear and difficult to distinguish, the contour information of the combination of adjacent subregions can be obtained to obtain a recognizable contour.

[0047] The deformation field determination module 30 is used to determine the deformation field that can convert the first contour set into the second contour set based on the first contour set and the second contour set. Specifically, the deformation field determination module 30 selects at least the contour information of the sub-region where the tumor is located before drug treatment from the first contour set; correspondingly, selects at least the contour information of the sub-region where the tumor is located after drug treatment from the second contour set; and uses the deformation registration model to perform deformation registration on the selected contour information to obtain the deformation field.

[0048] Assuming that a tumor grows in a subregion of an organ, the deformation field determination module 30 selects contour information of the subregion where the tumor is located before drug treatment and contour information of the subregion where the tumor is located after drug treatment from the first contour set and the second contour set respectively, and uses a deformation registration model to perform deformation registration on the contour information of the subregion where the tumor is located before and after drug treatment, and obtains a deformation field that can convert the contour information of the subregion where the tumor is located before drug treatment into the contour information of the subregion where the tumor is located after drug treatment.

[0049] Assuming that a tumor grows in adjacent subregions of an organ, the deformation field determination module 30 selects contour information of the adjacent subregion combination of the tumor before drug treatment and contour information of the subregion combination of the tumor after drug treatment from the first contour set and the second contour set respectively, and uses the deformation registration model to perform deformation registration on the contour information of the subregion combination of the tumor before and after drug treatment, and obtains a deformation field that can convert the contour information of the subregion combination of the tumor before drug treatment into the contour information of the subregion combination of the tumor after drug treatment.

[0050] It should be pointed out that the contour information used for deformation registration by the deformation field determination module 30 must include the contour information of the sub-region where the tumor is located or the contour information of the combination of sub-regions where the tumor is located. Whether the contour information of other sub-regions or the contour information of other sub-region combinations is used or not is determined by the doctor according to the circumstances. For example, when necessary, the deformation field determination module 30 can use the deformation registration model to perform deformation registration on all contour information in the first contour set and the second contour set to obtain a deformation field.

[0051] The target contour determination module 40 is used to determine the tumor target contour set after drug treatment based on the deformation field and the tumor target contour set before drug treatment. Specifically, the target contour determination module is used to transform the tumor target contour set before drug treatment using the deformation field to obtain the transformed tumor target contour set as the tumor target contour set after drug treatment.

[0052] The system may further include: an output module 50, which is used to output the tumor target area contour set after the drug treatment, so as to implement corresponding clinical operations based on the output tumor target area contour set after the drug treatment, such as conventional radiotherapy, close-range radiotherapy, ablation therapy, or puncture.

[0053] Figure 2b is based on Figure 2a The flow chart of the system for treating tumors in patients after drug treatment is as follows: Figure 2b As shown, the steps may include:

[0054] Step S201: obtaining a medical image set before drug treatment, a tumor target area contour set before drug treatment, and a first contour set related to the organ where the tumor is located.

[0055] Step S202: obtaining a medical image set after drug treatment and a second contour set related to the organ where the tumor is located after drug treatment.

[0056] Step S203: determining, according to the first contour set and the second contour set, a deformation field capable of transforming the first contour set into the second contour set.

[0057] Step S204: determining the tumor target area contour set after drug treatment on the positioning image or the guiding image of the medical image set after drug treatment according to the deformation field and the tumor target area contour set before drug treatment.

[0058] Step S205: outputting the tumor target area contour set after the drug treatment for subsequent clinical operations.

[0059] The above steps S201-S205 can be respectively performed by Figure 2a The first acquisition module 10, the second acquisition module 20, the deformation field determination module 30, the target area contour determination module 40 and the output module 50 are implemented as shown.

[0060] Figure 2c and 2d is the MRI tumor target contour before treatment, using Figure 2aThe first and second comparison images of the tumor target area contour after drug treatment determined by the system, the postoperative pathological section contour (gold standard), such as Figure 2c and Figure 2d As shown, Ta is the contour of the tumor target area before drug treatment, and Tb is the contour of the tumor target area after drug treatment determined by the present invention. It can be seen that Tb has a high similarity with the contour of the postoperative pathological section (gold standard), indicating that the accuracy of the tumor target area after drug treatment determined and outlined by the present invention is high. Compared with the tumor contour manually outlined by the prior art or the expansion of the intended irradiation range of the tumor target area to include the entire anatomical part of the tumor target area, the tumor target area contour set after drug treatment determined by the present invention is more accurate, so that during clinical operations (including conventional radiotherapy, brachytherapy, ablation therapy, or puncture), it can not only effectively and accurately locate the tumor range clinically, but also reduce the probability of various side effects during implementation.

[0061] An embodiment of the present invention further provides a control system for treating a tumor after drug treatment, and the system may include the aforementioned system for determining a tumor target area contour set after drug treatment.

[0062] It should be pointed out that the tumor treatment control system can be a conventional radiotherapy control system, or it can be other tumor treatment control systems involved in tumor targeted puncture, radiofrequency ablation, particle implantation and other close-range radiotherapy. For example, the tumor treatment control system after drug treatment is a conventional radiotherapy control system after drug treatment, which uses the aforementioned system for determining the tumor target area contour set after drug treatment to determine the contour of the diffuse tumor after drug treatment, thereby ensuring the prescribed dose of the tumor and reducing the irradiation dose of the endangered organ during conventional radiotherapy. For another example, the tumor treatment control system after drug treatment is a system related to tumor targeted puncture after drug treatment, which uses the aforementioned system for determining the tumor target area contour set after drug treatment to determine the contour of the diffuse tumor after drug treatment, thereby accurately navigating to the tumor site during tumor puncture biopsy, so that the puncture range is small, or the puncture points are few, and the patient's trauma is reduced.

[0063] The following is an example of conventional radiotherapy. Figures 3 to 10d The present invention is described in detail.

[0064] Figure 3 is a conventional radiotherapy flow chart for patients after drug treatment provided by an embodiment of the present invention, and Figure 1 Compared with the existing technology, this embodiment can generate a tumor target area contour set after drug treatment, such as Figure 3 As shown, the steps include:

[0065] Step S201: Acquire images and related clinical data.

[0066] Obtain medical image sets before and after drug treatment and related clinical data. The medical image set after drug treatment includes at least one set of positioning images for outlining and planning, and related clinical data include age, medical history, pathology report, imaging report, and test report.

[0067] Step S202: Design a treatment plan.

[0068] According to the images acquired in step S201 and the relevant clinical data, a treatment plan is designed, which generally includes the intended irradiation range of the tumor target area, the prescribed dose, the fractionated dose, the dose limit, etc.

[0069] In this embodiment, the proposed irradiation range of the tumor target area in the treatment plan will be designed as the outline of the tumor target area after drug treatment, rather than the entire anatomical part containing the tumor target area. For example, for prostate cancer, the treatment area is no longer the entire anatomical part containing the tumor target area, but the outline of the tumor after endocrine therapy. The volume of the irradiation range is often smaller and farther away from the relevant organs at risk, which will bring two clinical benefits: first, when designing the prescription dose and fractionated dose, since the irradiation range is farther away from the relevant organs at risk, there is an opportunity to increase the prescription dose and fractionated dose of the irradiation range clinically, so as to effectively implement high-dose, radical radiotherapy; second, when designing the treatment plan, since the irradiation range is smaller, there is an opportunity to effectively reduce the irradiation dose of the relevant organs at risk, thereby reducing the probability of organ-threatening toxic and side effects during the treatment implementation.

[0070] Step S203: Generate a tumor target area contour set after drug treatment.

[0071] According to the proposed irradiation range of the tumor target area designed in step S203, the contour of the tumor target area is determined on the positioning image of the medical image set after drug treatment.

[0072] Step S204: Outline the contours of the organs at risk or normal tissues.

[0073] According to the proposed irradiation range of the tumor target area designed in step S203, an organ at risk (OAR) contour set is outlined on the positioning image of the medical image set after drug treatment. Generally speaking, OAR refers to the contours of various related organs at risk or surrounding normal tissues. For example, during radiotherapy of prostate cancer, OAR generally includes the rectum, bladder, small intestine, femoral head, etc.

[0074] Step S205: Design a treatment plan.

[0075] The same as the existing scheme, that is, according to the treatment plan and outline in the above steps, the radiotherapy planning system is used to design the treatment plan and generate a plan execution file that can be recognized by the radiotherapy machine.

[0076] Step S206: Treatment implementation.

[0077] The same as the existing scheme, that is, using the plan execution file generated by the above steps, radiotherapy is implemented on the radiotherapy machine.

[0078] Figure 4 FIG. 1 is a flow chart of generating a tumor target area contour set after drug treatment provided by an embodiment of the present invention, such as Figure 4 As shown, it involves an input module, a processing module and an output module, wherein the input module can realize the functions of the first acquisition module 10 and the second acquisition module 20 in Figure 2, the processing module can realize the functions of the deformation field determination module 30 and the target area contour determination module 40 in Figure 2, and the output module can realize the functions of the output module 50 in Figure 2.

[0079] 1. Input module

[0080] The input module inputs the following:

[0081] (1) The medical image set Ia before drug treatment and the sub-region contour set Pa (including n contours, P) for deformation registration drawn based on the image set a，1 , P a，2 ,……,P a，n ) and the tumor target contour set Ta (containing m contours, T a，1 、T a，2 ,……,T a，m ).

[0082] Medical image set Ia before drug treatment refers to the collection of medical images collected before the patient undergoes drug treatment. At least one of CT images, magnetic resonance images, nuclear medicine images or ultrasound images is selected according to the location and type of the tumor. For example, magnetic resonance images are selected for prostate cancer. If necessary, the remaining images can be used as references for outlining. These images can be two-dimensional images, three-dimensional images or four-dimensional images. Taking three-dimensional images as an example, if each pixel on the image is regarded as an element in a matrix, then this three-dimensional image can be mathematically regarded as a three-dimensional matrix, and the element value of a certain layer, row and column in the matrix is ​​equal to the pixel value of a certain layer, row and column in the three-dimensional image.

[0083] The sub-region contour set Pa used for deformable registration refers to the contours to be used in the deformable registration of the processing module, and includes n sub-region contours in total, which are recorded as Pa1, Pa,2, ..., Pa,n. The coverage of this contour set must include the tumor target contour set Ta, that is, from the three-dimensional volume, the geometry formed by the tumor target contour set Ta is a subset of the geometry formed by the sub-region contour set Pa. Taking prostate cancer as an example, the tumor target must be located in the prostate structure. Therefore, the sub-region contour set that needs to be outlined for deformable registration here can include five sub-regions of the prostate anatomical structure: the peripheral zone (PZ), the transition zone (TZ), the central zone (CZ), the anterior fibrous stroma zone (AFS), and the central urethra (US). If some sub-regions are not easy to distinguish, they can also be adjacent combinations, such as TZ+CZ, and can also include the seminal vesicle (SV) closely related to the prostate anatomical structure.

[0084] Before drug treatment, doctors can use clinical knowledge to identify the boundaries of the tumor target area on the above medical image set. The tumor target area may be an independent target area or multiple scattered target areas, so it is necessary to outline the tumor target area contour set Ta, which contains a total of m tumor target area contours, denoted as T a，1 、T a，2 ,……,T a，m Among them, outlining, in mathematical terms, means assigning the same label to certain elements in the three-dimensional matrix of the image, which corresponds to the contour to be outlined. In this way, the pixels corresponding to these matrix elements on the three-dimensional image also have the same label, and these pixels will form a geometric structure with a three-dimensional volume on the three-dimensional image. This geometric structure is called a contour, and this operation process is called outlining.

[0085] (2) The medical image set Ib after drug treatment and the sub-region contour set Pb (including n contours, Pb) for deformation registration drawn based on the image set b，1 , P b，2 ,……,P b，n ).

[0086] The medical image set Ib after drug treatment refers to a set of medical images collected after the patient receives drug treatment, including but not limited to CT images, magnetic resonance images, nuclear medicine images, ultrasound images, etc.

[0087] The sub-region contour set Pb used for deformation registration refers to the contours to be used in the subsequent deformation registration, which contains n sub-region contours in total, denoted as P b，1 , P b，2 ,……,P b，n , and these n contours are consistent with the n contours P in Pa a，1 , P a，2 ,……,Pa，n There is a one-to-one correspondence. Taking prostate cancer as an example, if Pa includes the peripheral zone (PZ), transition zone (TZ), central zone (cZ), anterior fibrous stroma zone (AFS), central urethra (US) and seminal vesicle (SV), or adjacent combinations such as TZ+CZ, then Pb must also include the peripheral zone (PZ), transition zone (TZ), central zone (CZ), anterior fibrous stroma zone (AFS), central urethra (US) and seminal vesicle (SV), or adjacent combinations such as TZ+CZ.

[0088] Since the tumor target area after drug treatment is diffuse in the medical image set after drug treatment, its contour boundary is difficult to be directly identified from the medical image through existing clinical medical knowledge. Therefore, the embodiment of the present invention establishes a mapping relationship between Pa and Pb, and maps the subset Ta of Pa to the medical image set after drug treatment, thereby generating a tumor target area contour set Tb after drug treatment.

[0089] (3) The parameter set C of the deformable registration model M (including k parameters, C1, C2, ..., C k ).

[0090] The deformable registration model M can adopt an existing model, which is a known technology. The parameter set C of the deformable registration model M is a parameter set corresponding to the selected model M, which is also a known technology. Taking the demons deformable registration model as an example, its parameters include but are not limited to the number of iteration levels, the maximum number of iterations, the diffusion speed coefficient, the similarity index, etc.

[0091] 2. Processing Module

[0092] The processing module performs the following steps:

[0093] Step 1: Use the deformable registration model M with a preset parameter set C to perform deformable registration on the contour set Pa and the contour set Pb to obtain the deformation field D.

[0094] The preset parameter set C is related to the selected deformable registration model. Taking the demons deformable registration model as an example, its parameters include but are not limited to the number of iteration levels, the maximum number of iterations, the diffusion velocity coefficient, the similarity index, etc. For the specific assignment of model parameters, the default values ​​can be used, or the common techniques in this field can be used to assign parameters, such as finding appropriate parameter values ​​for different cases through manual trials and iterative searches, or using artificial intelligence methods to train to obtain parameter values ​​for a certain disease.

[0095] Deformation registration is to use certain features in the image (such as marker coordinates, pixel values, etc.) to establish a mapping relationship between the original image and the target image through an optimization algorithm. Through this mapping relationship, a mapping image of the original image can be obtained, and the mapping image is very similar to the target image, satisfying the similarity index preset by the optimization algorithm. This mapping relationship is usually called a deformation field, and the mapping process is called registration. Since there are usually differences in shape between images, it is often called deformation registration. Generally speaking, the deformation field conforms to the affine transformation (AffineTransformation), which is a linear transformation between two-dimensional / three-dimensional coordinates. It can be achieved by the composite superposition of a series of sub-transformations, including: translation, scaling, flipping, rotation, and shearing. Mathematically speaking, deformation registration is to obtain an affine transformation matrix through an optimization search algorithm. After applying this matrix to the original image, a registered image with a high degree of similarity to the target image can be obtained. Obviously, any sub-image area in the original image can be transformed into a deformed registered sub-image area through the deformation field and mapped onto the target image.

[0096] Figure 5a and Figure 5b This is an example of deformation registration. The original image and the target image are triangles of different shapes. Now we need to establish a deformation field so that each point in the triangle of the original image can be mapped to each corresponding point in the triangle of the target image in the form of affine transformation. In this example, the deformation field is a small rhombus. By applying this small rhombus to the original image, the original image will produce two sub-transformations: reduction and shearing. The superposition of these two sub-transformations maps the triangle of the original image to the triangle of the target image. This process is called deformation registration. It can be seen that through this deformation field, not only the vertices of the external large triangle can be mapped one by one, but also the vertices of the internal small triangle can be mapped one by one.

[0097] In combination with the embodiments of the present invention, if the outer large triangle is regarded as a sub-region contour set for deformable registration, and the inner small triangle is regarded as a tumor target contour set, then the tumor target contour set in the medical image set before drug treatment can be mapped to the medical image set after drug treatment through deformable registration of the sub-region contour set.

[0098] Step 2: Apply the deformation field D to the contour set Ta to obtain the deformed contour set Tb (containing m contours, T b，1 、T b，2 ,……,T b，m ), as the tumor target contour set after drug treatment.

[0099] The m contours contained in Tb correspond one-to-one to the m contours contained in Ta, and are mapped one-to-one to the image set Ib through the deformation field D.

[0100] The deformation field D is applied to the contour set Ta. The "action" here refers to performing an affine transformation on Ta, or a matrix operation, usually the product of the affine transformation matrix and the image matrix. Of course, the "action" mentioned here is also part of the existing deformable registration model, that is, a forward transformation function. Directly applying this function can realize the above-mentioned "action" process.

[0101] Step 3: Integrate the contour set Tb into the medical image set Ib after drug treatment.

[0102] The medical image set after drug treatment includes at least one set of positioning images used for contouring and plan design, so step 3 specifically refers to integrating the contour set Tb into the positioning images used for contouring and plan design in Ib.

[0103] There are many methods of integration. A commonly used method in this field is to store the image set Ib as a file in dicom format (dicom is a common format for data in the medical field), and then use the writing rules of the dicom file or existing software or functions to write the contour set Tb into another dicom file, which can match the dicom file of Ib (requirements include but are not limited to header file information matching), so that the dicom files of Ib and Tb can be simultaneously imported into the planning system or related radiotherapy software for subsequent contouring and planning design. Of course, other existing technologies can also be used to implement the "integration" operation.

[0104] 3. Output Module

[0105] The output module is used to output a medical image set Ib after drug treatment that includes a tumor target area contour set Tb after drug treatment.

[0106] The output here can be a set of matching dicom files, or it can be the simultaneous display of image set Ib and contour set Tb directly on the planning system.

[0107] Application examples:

[0108] Because the prostate anatomy includes five sub-zones, namely the peripheral zone (PZ), transitional zone (TZ), central zone (CZ), anterior fibrous stromal zone (AFS), and central urethra (US), the division is more complicated. Solving the deformation of prostate cancer and its various sub-zones can cover other tumors with relatively simple anatomy. Therefore, this application example takes prostate cancer as an example.

[0109] Will Figure 3Conventional radiotherapy protocols for patients after drug treatment when applied to prostate cancer may include:

[0110] Step S301: Acquire images and related clinical data.

[0111] Multiparameter magnetic resonance images (mpMRI) of a patient with prostate cancer before and after endocrine therapy, as well as positioning CT images and clinical data used for outlining and planning during radiotherapy positioning, were obtained.

[0112] The multi-parameter magnetic resonance images include the following different sequences: small field high-definition T2-weighted (T2WI) sequence images, diffusion weighted imaging (Diffusion Weighted Imaging, DWI), and apparent diffusion coefficient (Apparent Diffusion Coefficient, ADC) images.

[0113] Among them, clinical data include age, pathology report, imaging report and tumor marker prostate-specific antigen (PSA).

[0114] Step S302: Design a treatment plan.

[0115] Design a treatment plan based on the acquired medical image sets before and after drug treatment and related clinical data. The treatment plan generally includes the intended irradiation range of the tumor target area, the prescribed dose, the dose in fractions, the dose limit, etc. For example, according to the basic information of the patient obtained in step S301, determine the patient's stage and risk classification and design a radiotherapy plan. Assume that the patient has extremely high-risk prostate cancer, but is old and in poor physical condition. Therefore, the tumor target area is intended to be irradiated with a high dose of 70Gy, and the prostate and seminal vesicle area are irradiated with 65Gy, with 25 fractions each, to ensure that the tumor can be controlled without damaging surrounding organs.

[0116] Step S303: Generate a prostate tumor target area contour set after endocrine therapy.

[0117] Step S304: Outline the organs at risk (OR).

[0118] OARs are outlined on the positioning images of the medical image set after drug treatment to obtain an OAR contour set. Generally speaking, OARs refer to various related organs at risk or surrounding normal tissues. For example, during radiotherapy for prostate cancer, OARs generally include the rectum, bladder, small intestine, femoral head, etc., so outlining the OAR contours includes outlining the contours of the rectum, bladder, small intestine, femoral head, etc., which are organs at risk that need to be protected.

[0119] Step S305: Design a treatment plan.

[0120] According to the treatment plan and tumor contour, a treatment planning system dedicated to radiotherapy is used to design a treatment plan for this patient and generate a plan execution file that can be recognized by the treatment machine.

[0121] Step S306: Treatment implementation.

[0122] The radiotherapy is performed on the radiotherapy machine using the plan execution file generated in step S305.

[0123] Combine the following Figure 4 The process of generating the prostate tumor target area contour set after endocrine therapy in step S303 is described in detail. Figure 4 As shown,

[0124] 1. Input module

[0125] 1. Input the contour sets Pa drawn before endocrine therapy.

[0126] Referring to the signal characteristics in all sequence images, different sub-zones in the prostate were outlined on T2WI images, including the peripheral zone (PZ) and the anterior fibrous stroma zone (AFS). Assuming that the transition zone (TZ) and central zone (CZ) of this patient were not easy to distinguish, TZ+CZ were outlined and named Pa, PZ; Pa, AFS; Pa, TZ+CZ respectively.

[0127] 2. Input the outlined tumor contour set Ta.

[0128] Different subregional contours and tumor contours before endocrine therapy Figure 6a , 6b , as shown in 6c, where Figure 6a This is a high-definition T2WI sequence image of Ono before endocrine therapy. Figure 6b This is the DWI sequence image before endocrine therapy. Figure 6c These are ADC sequence images before endocrine therapy.

[0129] 3. Input the outlined contour set Pb after endocrine therapy.

[0130] Different subzones within the prostate were delineated, including the peripheral zone (PZ), the anterior fibrous stromal zone (AFS), and the transition zone (TZ) + central zone (CZ), which were named Pb, PZ; Pb, AFS; Pb, TZ + CZ, respectively.

[0131] Different subregional contours and tumor contours after endocrine therapy Figure 7a , 7b , as shown in 7c, where Figure 7a This is a high-definition T2WI sequence image of Xiaoye after endocrine therapy. Figure 7b This is a DWI sequence image after endocrine therapy. Figure 7c These are ADC sequence images after endocrine therapy.

[0132] 4. Input the parameter set of the deformable registration model M.

[0133] The deformable registration model M of this application example adopts the demons deformable registration model, and its parameters include the number of iteration levels (set to 3), the maximum number of iterations (set to 100), the diffusion speed coefficient (set to 1), and the similarity index (set to mutual information index).

[0134] 2. Processing Module

[0135] 1. Perform deformation registration on the contour set Pa and the contour set Pb to obtain the deformation field D;

[0136] 2. Apply the deformation field D to Ta to obtain the tumor target area contour set Tb after endocrine therapy;

[0137] 3. Integrate Tb into MRI images after endocrine therapy.

[0138] 3. Output Module

[0139] The MRI images of the tumor target contour set Tb after endocrine therapy are included. The image sets before and after the operation are in the system as follows Figure 8a , 8b , 8c, 8d, where Figure 8a It is the map of each sub-region and tumor volume before endocrine therapy; Figure 8b This is the volume map of each subregion after endocrine therapy; Figure 8c It is the sub-region cluster and deformed tumor contour map after endocrine therapy; Figure 8d This is the subregion cluster and deformed tumor volume map after endocrine therapy.

[0140] The Tb contour set is transmitted back to the radiation therapy planning system. The MRI image with the Tb contour set includes Figure 9a , 9b9c show the small field high-definition T2WI sequence images, DWI sequence images, and ADC sequence images with Tb contour set after endocrine therapy, respectively.

[0141] This image can be matched to the CT image used for planning using existing treatment planning systems, such as Fig.10a , 10b 10c and 10d respectively show a CT image dedicated to radiotherapy planning, an original MRI image after color adjustment before fusion with CT (i.e., an MRI image used for outlining and deformation), a fusion image of CT and MRI, and an MRI image after registration with CT.

[0142] The above gives a detailed process of applying the system for determining the tumor target contour set after drug treatment to radiotherapy during conventional radiotherapy. Since the present invention can more accurately determine the tumor target contour set after drug treatment, it can ensure the prescribed dose for the tumor and reduce the irradiation dose for the endangered organ. In addition, the system for determining the tumor target contour set after drug treatment can also be applied to close-range radiotherapy such as particle implantation, ablation therapy, and puncture. The implementation process of the above operations is simpler than the radiotherapy implementation process and can be applied with reference to the radiotherapy implementation process.

[0143] An embodiment of the present invention also provides a storage medium, on which a program is stored. When the program is executed by a processor, the following steps are implemented: obtaining a tumor target area contour set before drug treatment and a first contour set related to the organ where the tumor is located; obtaining a second contour set related to the organ where the tumor is located after drug treatment; determining a deformation field that can convert the first contour set into the second contour set based on the first contour set and the second contour set; and determining the tumor target area contour set after drug treatment based on the deformation field and the tumor target area contour set before drug treatment.

[0144] Among them, the first contour set may include at least one of the contour information of the subregion of the organ before drug treatment and the contour information of the combination of adjacent subregions of the organ before drug treatment. In specific implementation, the contour information before drug treatment can be obtained as needed, and the present invention does not limit this. For example, if the tumor is distributed in a certain subregion, only the contour information of the subregion where the tumor is located can be obtained, or the contour information of each subregion of the organ can be obtained. If the subregion contour is unclear and difficult to distinguish, the contour information of the combination of adjacent subregions can be obtained to obtain a recognizable contour. In addition, the first contour set can also include contour information of other tissues or organs related to the anatomical structure of the organ where the tumor is located, such as contour information of the seminal vesicle that is closely related to the anatomical structure of the prostate.

[0145] The second contour set may include at least one of the contour information of the subregion of the organ after drug treatment and the contour information of the combination of adjacent subregions of the organ after drug treatment. In specific implementation, the contour information after drug treatment can be obtained as needed, and the present invention does not limit this. For example, if the tumor is distributed in a certain subregion, only the contour information of the subregion where the tumor is located can be obtained, or the contour information of each subregion of the organ can be obtained. If the subregion contour is unclear and difficult to distinguish, the contour information of the combination of adjacent subregions can be obtained to obtain a recognizable contour.

[0146] Among them, the step of determining the deformation field capable of converting the first contour set into the second contour set according to the first contour set and the second contour set implemented when the program is executed by the processor includes: selecting at least the contour information of the subregion where the tumor is located before drug treatment from the first contour set; correspondingly, selecting at least the contour information of the subregion where the tumor is located after drug treatment from the second contour set; using the deformation registration model, deforming and registering the selected contour information to obtain the deformation field. For example, if a tumor grows alone in a subregion of an organ, the contour information of the subregion where the tumor is located before drug treatment and the contour information of the subregion where the tumor is located after drug treatment can be selected from the first contour set and the second contour set, and the contour information of the subregion where the tumor is located before and after drug treatment can be deformed and registered using the deformation registration model to obtain a deformation field capable of converting the contour information of the subregion where the tumor is located before drug treatment into the contour information of the subregion where the tumor is located after drug treatment. For another example, when a tumor grows in an adjacent subregion of an organ, the contour information of the combination of adjacent subregions where the tumor is located before drug treatment and the contour information of the combination of subregions where the tumor is located after drug treatment can be selected from the first contour set and the second contour set, and the contour information of the combination of subregions where the tumor is located before and after drug treatment can be deformed and registered using the deformation registration model to obtain a deformation field that can convert the contour information of the combination of subregions where the tumor is located before drug treatment into the contour information of the combination of subregions where the tumor is located after drug treatment. It should be pointed out that the contour information used for deformation registration must include the contour information of the subregion where the tumor is located or the contour information of the combination of subregions where the tumor is located, and whether the contour information of other subregions or the contour information of other subregion combinations is selected or not is determined by the doctor according to the situation. For example, when necessary, the deformation field determination module 30 can use the deformation registration model to perform deformation registration on all contour information in the first contour set and the second contour set to obtain a deformation field.

[0147] Among them, the step of determining the tumor target contour set after drug treatment based on the deformation field and the tumor target contour set before drug treatment implemented when the program is executed by the processor includes: using the deformation field to transform the tumor target contour set before drug treatment, and obtaining the transformed tumor target contour set as the tumor target contour set after drug treatment.

[0148] When the program is executed by the processor, the following steps are also implemented: outputting the tumor target area contour set after the drug treatment, so as to implement corresponding clinical operations based on the output tumor target area contour set after the drug treatment, such as conventional radiotherapy, or close-range radiotherapy such as particle implantation, or ablation therapy, or puncture.

[0149] Compared with the existing technology of manually outlining the tumor contour or expanding the intended irradiation range of the tumor target area to include the entire anatomical part of the tumor target area, the tumor target area contour set after drug treatment determined in this embodiment is more accurate, can ensure the prescribed dose for the tumor, and reduce the irradiation dose to the endangered organs.

[0150] Among them, the above-mentioned storage media may include computer storage media (or non-transitory media), as known to those of ordinary skill in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules or other data). Storage media include but are not limited to RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer.

[0151] The embodiments of the present invention are described above with reference to the accompanying drawings, but the scope of the present invention is not limited thereby. Any modification, equivalent substitution and improvement made by those skilled in the art without departing from the scope and essence of the present invention shall be within the scope of the present invention.

Claims

1. A system for determining a tumor target area contour set after drug treatment, characterized in that: The system comprises: A first acquisition module is used to acquire a tumor target area contour set before drug treatment and a sub-area contour set of the organ where the tumor is located for deformation registration before drug treatment, wherein the geometric body formed by the tumor target area contour set before drug treatment is a subset of the geometric body formed by the sub-area contour set of the organ where the tumor is located for deformation registration before drug treatment; The second acquisition module is used to acquire a sub-region contour set of the organ where the tumor is located for deformable registration after drug treatment; A deformation field determination module, used for determining, based on the sub-region contour set of the organ where the tumor is located before the drug treatment for deformation registration and the sub-region contour set of the organ where the tumor is located after the drug treatment, a deformation field capable of converting the sub-region contour set of the organ where the tumor is located before the drug treatment for deformation registration into the sub-region contour set of the organ where the tumor is located after the drug treatment; The target area contour determination module is used to determine the tumor target area contour set after the drug treatment according to the deformation field and the tumor target area contour set before the drug treatment.

2. The system according to claim 1, characterized in that The sub-region contour set of the organ where the tumor is located before the drug treatment and used for deformable registration acquired by the first acquisition module includes at least one of contour information of the sub-region of the organ before the drug treatment and contour information of a combination of adjacent sub-regions of the organ before the drug treatment; The sub-region contour set of the organ where the tumor is located after drug treatment for deformable registration acquired by the second acquisition module includes at least one of contour information of the sub-region of the organ after drug treatment and contour information of a combination of adjacent sub-regions of the organ after drug treatment.

3. The system according to claim 1, characterized in that The target area contour determination module is specifically used to transform the tumor target area contour set before drug treatment by using the deformation field, and obtain the transformed tumor target area contour set as the tumor target area contour set after drug treatment.

4. The system according to any one of claims 1 to 3, characterized in that: The system further comprises: The output module is used to output the tumor target area contour set after the drug treatment, so as to implement corresponding clinical operations based on the output tumor target area contour set after the drug treatment.

5. A tumor treatment control system after drug treatment, characterized in that: The system comprises the system for determining a tumor target area contour set after drug treatment as described in any one of claims 1-4.

6. A storage medium, characterized in that: The storage medium stores a program, and when the program is executed by the processor, the following steps are implemented: Acquire a tumor target area contour set before drug treatment and a sub-area contour set of the organ where the tumor is located for deformation registration before drug treatment, wherein the geometric body formed by the tumor target area contour set before drug treatment is a subset of the geometric body formed by the sub-area contour set of the organ where the tumor is located for deformation registration before drug treatment; Obtaining a sub-region contour set of the organ where the tumor is located for deformable registration after drug treatment; Determine, according to the subregion contour set of the organ where the tumor is located for deformation registration before the drug treatment and the subregion contour set of the organ where the tumor is located for deformation registration after the drug treatment, a deformation field capable of converting the subregion contour set of the organ where the tumor is located for deformation registration before the drug treatment into the subregion contour set of the organ where the tumor is located for deformation registration after the drug treatment; The tumor target area contour set after drug treatment is determined according to the deformation field and the tumor target area contour set before drug treatment.

7. The storage medium according to claim 6, characterized in that The sub-region contour set of the organ where the tumor is located before drug treatment for deformable registration includes at least one of contour information of the sub-region of the organ before drug treatment and contour information of a combination of adjacent sub-regions of the organ before drug treatment; The sub-region contour set of the organ where the tumor is located after the drug treatment for deformable registration includes at least one of contour information of the sub-region of the organ after the drug treatment and contour information of a combination of adjacent sub-regions of the organ after the drug treatment.

8. The storage medium according to claim 7, characterized in that The step of determining the tumor target area contour set after drug treatment based on the deformation field and the tumor target area contour set before drug treatment, which is implemented when the program is executed by the processor, includes: using the deformation field to transform the tumor target area contour set before drug treatment, and obtaining the transformed tumor target area contour set as the tumor target area contour set after drug treatment.

9. The storage medium according to any one of claims 6 to 8, characterized in that: When the program is executed by the processor, the following steps are also implemented: The tumor target area contour set after the drug treatment is output, so as to implement corresponding clinical operations based on the output tumor target area contour set after the drug treatment.

Images