Tumor growth prediction method and system based on digital twinning

Abstract

The invention relates to the field of tumor growth prediction, and discloses a tumor growth prediction method and system based on digital twinning, and the method comprises the steps: constructing a first digital twinning model of brain image data, clinical data and physiological data of a patient; constructing a second digital twinborn model of the blood brain barrier data, the drug interaction data and the dynamic microenvironment data of the patient; determining a first treatment prediction result according to the treatment scheme of the patient through the first digital twinborn model; determining a second treatment prediction result according to the treatment scheme of the patient through a second digital twinborn model; and through a treatment scheme correction model, according to the treatment scheme of the patient, the first treatment prediction result and the second treatment prediction result, determining a correction treatment scheme of the patient. The accuracy of tumor growth prediction is improved through the pharmacokinetic model, interaction mechanism simulation between drugs and dynamic microenvironment parameters.

Description

Technical Field

[0001] This application relates to the technical field of tumor growth prediction, and more specifically, to a tumor growth prediction method and system based on digital twin. Background Art

[0002] Tumor growth prediction is one of the core challenges in the field of precision medicine. Traditional methods mainly rely on two-dimensional medical image analysis and static mathematical models, and it is difficult to dynamically simulate the spatio-temporal evolution process of tumor cells and the interaction of complex microenvironments. Existing technologies usually based on simplified physical equations or statistical learning models, ignoring key biological mechanisms such as tumor heterogeneity, angiogenesis, and immune escape, resulting in limited prediction accuracy. In addition, traditional methods lack real-time and individualized adaptation capabilities, unable to dynamically adjust the prediction model according to the patient's treatment response, and it is difficult to meet the needs of clinical personalized treatment.

[0003] In recent years, the rise of digital twin technology has provided a new paradigm for tumor growth prediction. Digital twin can simulate the evolution trajectory of tumors under different treatment interventions by constructing a three-dimensional dynamic simulation model of tumors and their microenvironments and integrating data such as multi-modal medical images, gene sequencing, and pathological reports. However, how to effectively couple the prediction results with the clinical decision-making system in existing digital twin models to achieve intelligent optimization of treatment plans is also a current research difficulty. For example, Patent CN119092136B (Application No.: CN202411586012.7) provides a modeling method based on multi-modal digital twin technology, which can perform multi-modal digital twin simulation of tumors through multi-modal data of images, genes, and clinics to generate prediction results of treatment plan effects. The method in Patent CN119092136B only relies on a simplified pharmacokinetic model to predict the treatment results, resulting in poor intelligent optimization effect of the multi-modal digital twin technology on treatment plans. Summary of the Invention

[0004] The purpose of this application is to provide a tumor growth prediction method and system based on digital twin, which solves the technical problem of only relying on a simplified pharmacokinetic model to predict treatment results, and achieves the technical effect of improving the accuracy of tumor growth prediction through pharmacokinetic models, simulation of the interaction mechanism between drugs, and dynamic microenvironment parameters.

[0005] A tumor growth prediction method based on digital twin provided by an embodiment of the present application, the method includes: constructing a digital twin database of multimodal pathological data of a patient, where the multimodal pathological data includes the patient's brain image data, clinical data, patient physiological data, blood-brain barrier data, drug interaction data, and dynamic microenvironment data; constructing a first digital twin model of the patient's brain image data, clinical data, and patient physiological data, where the first digital twin model reflects the growth direction and evolution speed characteristics of the tumor at different stages of the patient; and constructing a second digital twin model of the patient's blood-brain barrier data, drug interaction data, and dynamic microenvironment data, where the second digital twin model reflects the treatment microenvironment characteristics of the tumor at different stages of the patient; determining a first treatment prediction result through the first digital twin model according to the patient's treatment plan; and determining a second treatment prediction result through the second digital twin model according to the patient's treatment plan; determining the revised treatment plan of the patient through a treatment plan correction model according to the patient's treatment plan, the first treatment prediction result, and the second treatment prediction result.

[0006] In a possible implementation manner, the method further includes: determining the blood-brain barrier characteristics of the patient through a blood-brain barrier detection model according to the patient's brain image data and blood-brain barrier data; determining the revised treatment plan of the patient through a treatment plan correction model according to the patient's treatment plan, the first treatment prediction result, the second treatment prediction result, and the blood-brain barrier characteristics.

[0007] In another possible implementation manner, the method further includes: constructing drug concentration gradient brain image data corresponding to the patient's brain image data according to the patient's brain image data, blood-brain barrier data, and dynamic microenvironment data; constructing a first digital twin model of the patient's brain image data, clinical data, and patient physiological data, including: constructing a first digital twin model of the patient's drug concentration gradient brain image data, clinical data, and patient physiological data.

[0008] In another possible implementation manner, the method further includes: determining the normal tissue region and tumor infiltration region of the patient's brain image data, and obtaining the normal tissue characteristics corresponding to the normal tissue region and the infiltration region characteristics corresponding to the tumor infiltration region; constructing the normal tissue drug concentration gradient brain image data of the patient according to the patient's normal tissue region, normal tissue characteristics, blood-brain barrier data, and dynamic microenvironment data; constructing the infiltration region drug concentration gradient brain image data of the patient according to the patient's tumor infiltration region, infiltration region characteristics, blood-brain barrier data, and dynamic microenvironment data; constructing a first digital twin model of the patient's brain image data, clinical data, and patient physiological data, including: constructing a first digital twin model of the patient's normal tissue drug concentration gradient brain image data, infiltration region drug concentration gradient brain image data, clinical data, and patient physiological data.

[0009] In another possible implementation, the method further includes: through the first digital twin model, determining the blood-brain barrier correction data of the patient according to the blood-brain barrier data of the patient, where the blood-brain barrier correction data represents the change information of the blood-brain barrier data; constructing a second digital twin model of the patient's blood-brain barrier data, drug interaction data, and dynamic microenvironment data, including: constructing a second digital twin model of the patient's blood-brain barrier correction data, drug interaction data, and dynamic microenvironment data.

[0010] In another possible implementation, the method further includes: obtaining the normal region blood-brain barrier characteristics corresponding to the normal tissue region and the infiltrated region blood-brain barrier characteristics corresponding to the tumor-infiltrated region; determining the normal region blood-brain barrier data corresponding to the normal tissue region according to the patient's blood-brain barrier data, the normal tissue region of the brain imaging data, and the normal region blood-brain barrier characteristics; determining the infiltrated region blood-brain barrier data corresponding to the tumor-infiltrated region according to the patient's blood-brain barrier data, the tumor-infiltrated region of the brain imaging data, and the infiltrated region blood-brain barrier characteristics; through the first digital twin model, determining the normal region blood-brain barrier correction data of the patient according to the normal region blood-brain barrier data of the patient; through the first digital twin model, determining the infiltrated region blood-brain barrier correction data of the patient according to the infiltrated region blood-brain barrier data of the patient; constructing a second digital twin model of the patient's blood-brain barrier data, drug interaction data, and dynamic microenvironment data, including: constructing a second digital twin model of the patient's normal region blood-brain barrier correction data, infiltrated region blood-brain barrier correction data, drug interaction data, and dynamic microenvironment data.

[0011] In another possible implementation, the method further includes: obtaining multiple similar first digital twin models of the first digital twin model and multiple similar second digital twin models of the second digital twin model, obtaining the treatment decision directions corresponding to the multiple similar first digital twin models respectively and the subsequent first digital twin models corresponding to the treatment decision directions; and obtaining the treatment decision directions corresponding to the multiple similar second digital twin models respectively and the subsequent second digital twin models corresponding to the treatment decision directions; where the treatment decision directions include aggressive treatment methods and conservative treatment methods; among the multiple similar second digital twin models, determining the target similar second digital twin models corresponding to the multiple similar first digital twin models respectively according to the parameter similarity of the brain imaging data of the similar first digital twin models and the blood-brain barrier data of the similar second digital twin models; when the treatment decision directions of the similar first digital twin model and the target similar second digital twin model are the same, taking the similar first digital twin model and the target similar second digital twin model as a predicted treatment decision model group, and taking the treatment decision direction corresponding to the predicted treatment decision model group as the recommended treatment decision direction.

[0012] In another possible implementation, the method further includes: obtaining a first subsequent tumor growth amplitude corresponding to a similar first digital twin model in each predictive treatment decision model group, and a second subsequent tumor growth amplitude corresponding to a target similar second digital twin model; determining an absolute value of the difference between the first subsequent tumor growth amplitude and the second subsequent tumor growth amplitude, and a ratio of the sum of the first subsequent tumor growth amplitude and the second subsequent tumor growth amplitude, as a subsequent tumor growth amplitude difference degree; recommending a recommended treatment decision direction in ascending order of the subsequent tumor growth amplitude difference degrees of multiple predictive treatment decision model groups.

[0013] In another possible implementation, the method further includes: obtaining a first complication assessment value corresponding to a similar first digital twin model in each predictive treatment decision model group, and a second complication assessment value corresponding to a target similar second digital twin model; wherein, the complication assessment value is used to characterize the degree of influence of complications on the patient's quality of life; determining an absolute value of the difference between the first complication assessment value and the second complication assessment value, and a ratio of the sum of the first complication assessment value and the second complication assessment value, as a complication assessment value difference degree; determining a sum of the subsequent tumor growth amplitude difference degree and the complication assessment value difference degree of each predictive treatment decision model group, as a comprehensive treatment direction assessment value; recommending a recommended treatment decision direction in ascending order of the comprehensive treatment direction assessment values of multiple predictive treatment decision model groups.

[0014] An embodiment of the present application further provides a tumor growth prediction system based on digital twins, including units for executing the method described in any one of the above.

[0015] The beneficial effects of the embodiment of the present application compared with the prior art are: The embodiments of the present application provide a tumor growth prediction method based on digital twin. The method includes: constructing a digital twin database of multimodal pathological data of a patient, where the multimodal pathological data includes the patient's brain imaging data, clinical data, patient physiological data, blood-brain barrier data, drug interaction data, and dynamic microenvironment data; constructing a first digital twin model of the patient's brain imaging data, clinical data, and patient physiological data, where the first digital twin model reflects the growth direction and evolution speed characteristics of the tumor in different stages of the patient; and constructing a second digital twin model of the patient's blood-brain barrier data, drug interaction data, and dynamic microenvironment data, where the second digital twin model reflects the treatment microenvironment characteristics of the tumor in different stages of the patient; determining a first treatment prediction result through the first digital twin model according to the patient's treatment plan; and determining a second treatment prediction result through the second digital twin model according to the patient's treatment plan; determining a revised treatment plan for the patient through a treatment plan correction model according to the patient's treatment plan, the first treatment prediction result, and the second treatment prediction result. In the embodiments of the present application, the growth direction and evolution speed characteristics of the tumor in different stages of the patient can be reflected through the first digital twin model, the treatment microenvironment characteristics of the tumor in different stages of the patient can be reflected through the second digital twin model, and the first treatment prediction result and the second treatment prediction result can be corrected through the treatment plan correction model, so as to realize the determination of the tumor treatment plan according to the growth direction and evolution speed characteristics of the tumor and the treatment microenvironment characteristics, and be able to compare and verify the treatment effects of the first treatment prediction result and the second treatment prediction result on the treatment plan, thereby improving the simulation effect of the treatment effect of brain tumors. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 It is a schematic flowchart of the first tumor growth prediction method based on digital twin provided by the embodiments of the present application; Figure 2 It is a schematic workflow diagram of the first tumor growth prediction method based on digital twin provided by the embodiments of the present application; Figure 3 It is a schematic flowchart of the second tumor growth prediction method based on digital twin provided by the embodiments of the present application; Figure 4 It is a schematic flowchart of the third tumor growth prediction method based on digital twin provided by the embodiments of the present application; Figure 5 A tissue schematic diagram of a normal tissue region and a tumor infiltration region provided by an embodiment of the present application; Figure 6 A flowchart of a fourth digital-twin-based tumor growth prediction method provided by an embodiment of the present application; Figure 7 A flowchart of a fifth digital-twin-based tumor growth prediction method provided by an embodiment of the present application; Figure 8 A flowchart of a sixth digital-twin-based tumor growth prediction method provided by an embodiment of the present application; Figure 9 A flowchart of a seventh digital-twin-based tumor growth prediction method provided by an embodiment of the present application; Figure 10 A logical structure diagram of a digital-twin-based tumor growth prediction system provided by an embodiment of the present application. Detailed implementation manners

[0018] It should be understood that when used in the specification and appended claims of the present application, the term "comprising" indicates the presence of the described features, wholes, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0019] It should also be understood that the term "and / or" used in the specification and appended claims of the present application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0020] As used in the specification and appended claims of the present application, the term "if" may be interpreted as "when", "once", "in response to determining", or "in response to detecting" according to the context. Similarly, the phrase "if determined" or "if [the described condition or event] is detected" may be interpreted as meaning "once determined", "in response to determining", "once [the described condition or event] is detected", or "in response to detecting [the described condition or event]" according to the context.

[0021] In addition, in the description of the specification and appended claims of the present application, the terms "first", "second", "third", etc. are only used for differential description and cannot be understood as indicating or implying relative importance.

[0022] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that a particular feature, structure, or characteristic described in connection with the embodiment is included in one or more embodiments of this application. Thus, statements such as "in one embodiment", "in some embodiments", "in other some embodiments", "in still other embodiments", etc. that appear in different places in this specification are not necessarily all referring to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized. The terms "comprising", "including", "having" and their variants all mean "including but not limited to", unless otherwise specifically emphasized.

[0023] Existing modeling methods based on multimodal digital twin technology can perform multimodal digital twin simulation on tumors through multimodal data of images, genes, and clinics to generate prediction results of the effects of treatment plans. However, the current modeling methods based on multimodal digital twin technology only rely on simplified pharmacokinetic models to predict treatment results, resulting in poor intelligent optimization effects of this multimodal digital twin technology on treatment plans.

[0024] For the above reasons, the embodiments of this application provide a tumor growth prediction method based on digital twin. The method includes: constructing a digital twin database of multimodal pathological data of a patient, where the multimodal pathological data includes the patient's brain image data, clinical data, patient physiological data, blood-brain barrier data, drug interaction data, and dynamic microenvironment data; constructing a first digital twin model of the patient's brain image data, clinical data, and patient physiological data, where the first digital twin model reflects the growth direction and evolution speed characteristics of the tumor in different stages of the patient; and constructing a second digital twin model of the patient's blood-brain barrier data, drug interaction data, and dynamic microenvironment data, where the second digital twin model reflects the treatment microenvironment characteristics of the tumor in different stages of the patient; determining a first treatment prediction result through the first digital twin model according to the patient's treatment plan; and determining a second treatment prediction result through the second digital twin model according to the patient's treatment plan; determining the patient's modified treatment plan through a treatment plan correction model according to the patient's treatment plan, the first treatment prediction result, and the second treatment prediction result. In the embodiments of this application, the growth direction and evolution speed characteristics of the tumor in different stages of the patient can be reflected through the first digital twin model, the treatment microenvironment characteristics of the tumor in different stages of the patient can be reflected through the second digital twin model, and the first treatment prediction result and the second treatment prediction result can be corrected through the treatment plan correction model, realizing the determination of the tumor treatment plan according to the growth direction and evolution speed characteristics of the tumor and the treatment microenvironment characteristics, and being able to compare and verify the treatment effects of the first treatment prediction result and the second treatment prediction result on the treatment plan, improving the simulation effect of the treatment effect of brain tumors.

[0025] In some scenarios, a tumor growth prediction method based on digital twin in an embodiment of the present application can be applied to the treatment of brain tumors, which can improve the digital twin simulation effect in the treatment of brain tumors and enhance the treatment prediction effect for brain tumors.

[0026] The following specifically describes a tumor growth prediction method based on digital twin provided by the embodiments of the present application with specific examples.

[0027] Figure 1 FIG. is a schematic flowchart of the first tumor growth prediction method based on digital twin provided by the embodiments of the present application. As Figure 1 shown, this method includes S110 to S130, and the following specifically describes S110 to S130.

[0028] S110. Construct a digital twin database of the patient's multimodal pathological data, where the multimodal pathological data includes the patient's brain image data, clinical data, patient physiological data, blood-brain barrier data, drug interaction data, and dynamic microenvironment data.

[0029] Since the current modeling method based on multimodal digital twin technology only relies on a simplified pharmacokinetic model to predict the treatment outcome, resulting in poor intelligent optimization effect of the multimodal digital twin technology on the treatment plan. In the embodiments of the present application, in order to avoid predicting the treatment outcome only through a simplified pharmacokinetic model, a digital twin database of the patient's multimodal pathological data is first constructed. The multimodal pathological data includes the patient's brain image data, clinical data, patient physiological data, blood-brain barrier data, drug interaction data, and dynamic microenvironment data. Furthermore, the accuracy of predicting the treatment outcome can be improved through the blood-brain barrier data, drug interaction data, and dynamic microenvironment data in the digital twin database, avoiding the problem of poor simulation effect of tumor treatment through a simplified pharmacokinetic model.

[0030] Exemplarily, the brain image data can be obtained by magnetic resonance imaging (MRI). Specifically, high-resolution brain structure images are generated through magnetic fields and radio waves. It can also quickly obtain brain tomographic images through computed tomography (CT), which is suitable for preliminary diagnosis in emergency situations. At the same time, the brain image data can be collected by positron emission tomography (PET) combined with radioactive tracers to detect tumor metabolic activity (such as FDG-PET).

[0031] Exemplarily, clinical data can be obtained through electronic health records (EHRs). Specifically, patient medical history, diagnosis results, and treatment records can be extracted from the hospital HIS system to collect clinical data. At the same time, patient symptoms and medication history can be recorded through structured forms using standardized tables (such as CRFs, case report forms) to obtain clinical data.

[0032] Exemplarily, patient physiological data can be measured by conventional detection devices such as sphygmomanometers, electrocardiographs, and oximeters to obtain vital signs. Patient physiological data can also be obtained through laboratory tests such as blood tests (complete blood count, liver and kidney function) and urine tests.

[0033] Exemplarily, blood-brain barrier data can be obtained through enhanced MRI scans. Specifically, after injecting gadolinium contrast agent, the permeability of the blood-brain barrier (such as the peritumoral edema situation) is evaluated to obtain blood-brain barrier data. Blood-brain barrier data can also be obtained through lumbar puncture (LP) by detecting indicators such as proteins and glucose in cerebrospinal fluid (CSF) to indirectly reflect the function of the blood-brain barrier. In addition, blood-brain barrier data can also be obtained by detecting biomarkers such as S100B protein and vascular endothelial growth factor (VEGF) in blood or cerebrospinal fluid. Blood-brain barrier data can also be obtained through molecular imaging such as ultrasmall superparamagnetic iron oxide (USPIO) nanoparticle MRI to evaluate the integrity of the blood-brain barrier.

[0034] Exemplarily, drug interaction data can be obtained through drug databases such as Micromedex, Lexicomp, and UpToDate to provide information on interactions between drugs. Drug interaction data can also be obtained through literature searches to track the latest research to supplement rare interactions not covered by the database.

[0035] Exemplarily, dynamic microenvironment data can be obtained through tumor biopsies by analyzing tumor cells, immune cells, and angiogenesis (such as immunohistochemistry, gene sequencing) in tissue samples. Dynamic microenvironment data can also be obtained through liquid biopsies to detect circulating tumor DNA (ctDNA) and exosomes to monitor the dynamic changes in the microenvironment and achieve the detection of dynamic microenvironment data.

[0036] S120. Construct a first digital twin model of the patient's brain imaging data, clinical data, and patient physiological data. The first digital twin model reflects the growth direction and evolution speed characteristics of the tumor in different stages of the patient. And construct a second digital twin model of the patient's blood-brain barrier data, drug interaction data, and dynamic microenvironment data. The second digital twin model reflects the treatment microenvironment characteristics of the tumor in different stages of the patient.

[0037] After obtaining the digital twin database, a first digital twin model of the patient's brain imaging data, clinical data, and patient physiological data can be constructed. Since the brain imaging data, clinical data, and patient physiological data can predict the tumor growth process at the macroscopic level, the first digital twin model can extract the tumor volume change through image segmentation technology, and combine with the clinical chemotherapy plan to predict the growth rate, so that the first digital twin model can reflect the growth direction and evolution speed characteristics of the tumor in different stages of the patient.

[0038] After obtaining the digital twin database, a second digital twin model of the patient's blood-brain barrier data, drug interaction data, and dynamic microenvironment data can also be constructed, so that the second digital twin model reflects the treatment microenvironment characteristics of the tumor in different stages of the patient. The blood-brain barrier data, drug interaction data, and dynamic microenvironment data in the second digital twin model are more suitable for microscopic mechanism modeling, and can be updated in real time dynamically, realizing multi-scale correlation analysis from molecules (such as drug targets) to tissues (such as tumor vascular density), and at the same time can optimize the treatment for the patient-specific microenvironment (such as immunosuppressive TME), facilitating the design of combination treatment plans, such as the combination treatment method of immune checkpoint inhibitors and anti-angiogenic drugs.

[0039] S130: Through the first digital twin model, according to the patient's treatment plan, determine the first treatment prediction result. And through the second digital twin model, according to the patient's treatment plan, determine the second treatment prediction result. Through the treatment plan correction model, according to the patient's treatment plan, the first treatment prediction result, and the second treatment prediction result, determine the patient's corrected treatment plan.

[0040] After obtaining the first digital twin model and the second digital twin model, through the first digital twin model, according to the patient's treatment plan, determine the first treatment prediction result, and the first treatment prediction result can determine the first treatment prediction result for the treatment plan from the growth direction and evolution speed characteristics of the tumor in different stages.

[0041] After obtaining the first digital twin model and the second digital twin model, through the second digital twin model, according to the patient's treatment plan, determine the second treatment prediction result, and the second treatment prediction result can determine the second treatment prediction result for the treatment plan from the treatment microenvironment characteristics of the tumor in different stages of the patient.

[0042] Since the first treatment prediction result characterizes the treatment effect obtained from the growth direction and evolution speed of the tumor at different stages, and since the second treatment prediction result characterizes the treatment effect obtained from the treatment microenvironment characteristics of the tumor at different stages, after obtaining the first treatment prediction result and the second treatment prediction result, a treatment plan correction model can be used to determine the corrected treatment plan for the patient according to the patient's treatment plan, the first treatment prediction result, and the second treatment prediction result, achieving the purpose of correcting the treatment measures in the patient's treatment plan according to the patient's treatment plan, the first treatment prediction result, and the second treatment prediction result.

[0043] Exemplarily, the treatment plan correction model can be a neural network model trained with the labeled patient's treatment plan, the first treatment prediction result, the second treatment prediction result, and the labeled patient's corrected treatment plan. The labeled patient's corrected treatment plan is used to correct the content in the labeled patient's treatment plan that conflicts with the first treatment prediction result and the second treatment prediction result, achieving the purpose of correcting the patient's treatment plan according to the first treatment prediction result and the second treatment prediction result, and improving the credibility and treatment effect of the patient's treatment plan.

[0044] The beneficial effect of the above implementation manner is that the first digital twin model can reflect the growth direction and evolution speed characteristics of the patient's tumor at different stages, and the blood-brain barrier data, drug interaction data, and dynamic microenvironment data in the second digital twin model are more suitable for microscopic mechanism modeling. By simulating the development process of the tumor through the characteristics in different dimensions, the digital simulation effect of the tumor treatment process is improved.

[0045] The beneficial effect of the above implementation manner is also that after obtaining the first treatment prediction result and the second treatment prediction result, a treatment plan correction model can be used to determine the corrected treatment plan for the patient according to the patient's treatment plan, the first treatment prediction result, and the second treatment prediction result, and it is possible to comprehensively correct the treatment plan according to the growth direction and evolution speed of the tumor at different stages and the treatment microenvironment characteristics, further improving the treatment effect of the patient's treatment plan on the tumor.

[0046] Figure 3 The flowchart of the second digital-twin-based tumor growth prediction method provided by the embodiments of the present application is shown as Figure 3 As shown, the above method further includes S210 to S220, and the following is a specific description of S210 to S220.

[0047] S210. Through the blood-brain barrier detection model, determine the blood-brain barrier characteristics of the patient according to the patient's brain image data and blood-brain barrier data.

[0048] The brain imaging data of the patient in the first digital twin model and the blood-brain barrier data in the second digital twin model can both characterize the blood-brain barrier characteristics of the patient at the imaging level and the blood-brain barrier characteristics of the patient at the microenvironment level. In the embodiments of the present application, the blood-brain barrier detection model can be used to determine the blood-brain barrier characteristics of the patient according to the brain imaging data and the blood-brain barrier data of the patient, and then the treatment plan for the patient can be further adjusted comprehensively according to the blood-brain barrier characteristics of the patient.

[0049] Exemplarily, brain imaging data (such as MRI, CT) provides spatial localization and anatomical structure information (such as tumor location, edema range), while blood-brain barrier data (such as dynamic contrast-enhanced parameters, S100B protein level) reflects the functional state (such as permeability, integrity).

[0050] S220. Through the treatment plan correction model, according to the treatment plan of the patient, the first treatment prediction result, the second treatment prediction result, and the blood-brain barrier characteristics, determine the corrected treatment plan of the patient.

[0051] In tumor treatment, the blood-brain barrier characteristics have a great impact on the process of tumor treatment. Therefore, in the embodiments of the present application, the treatment plan correction model can be used to determine the corrected treatment plan of the patient according to the treatment plan of the patient, the first treatment prediction result, the second treatment prediction result, and the blood-brain barrier characteristics, realizing the further correction of the treatment plan of the patient in combination with the blood-brain barrier characteristics, which is applicable to the individualized treatment of complex central nervous system diseases (such as malignant brain tumors, autoimmune encephalitis).

[0052] The beneficial effect of the above implementation manner is that the blood-brain barrier characteristics characterize the blood-brain barrier characteristics of the patient at the imaging level and the blood-brain barrier characteristics of the patient at the microenvironment level, and can further correct the treatment plan of the patient in combination with the blood-brain barrier characteristics, which is applicable to the individualized treatment of complex central nervous system diseases and improves the digital simulation effect of the treatment effect of brain tumors.

[0053] In some implementation manners, the above method further includes: constructing drug concentration gradient brain imaging data corresponding to the brain imaging data of the patient according to the brain imaging data, the blood-brain barrier data, and the dynamic microenvironment data of the patient.

[0054] In the process of treating brain tumors, since brain imaging data, blood-brain barrier data, and dynamic microenvironment data can affect the drug concentration in different tissue regions, in the embodiments of the present application, the drug concentration gradient brain imaging data corresponding to the patient's brain imaging data can be constructed according to the patient's brain imaging data, blood-brain barrier data, and dynamic microenvironment data. The drug concentration gradient brain imaging data characterizes the drug concentration information in different tissue regions, and thus can further improve the simulation effect of the tumor treatment process for the patient based on the patient's drug concentration gradient brain imaging data.

[0055] Exemplarily, when constructing the drug concentration gradient brain imaging data corresponding to the patient's brain imaging data according to the patient's brain imaging data, blood-brain barrier data, and dynamic microenvironment data, first, imaging data such as MRI / CT (structure), PET (metabolism), DCE-MRI (blood flow / permeability) can be spatially aligned to a unified coordinate system (such as the MNI standard space), and blood-brain barrier (BBB)-related molecular markers (such as VEGF, MMP-9), gene expression profiles (such as ABCB1 transporter) can be superimposed on the imaging space through immunohistochemistry or in situ hybridization results, and parameters such as blood flow velocity (PWI-MRI), tissue interstitial pressure (TIP), pH value (1H-MRS) can be extracted to construct a voxel-level microenvironment feature vector to obtain the drug concentration gradient brain imaging data.

[0056] In some implementation manners, in S120 above, constructing the first digital twin model of the patient's brain imaging data, clinical data, and patient physiological data includes: constructing the first digital twin model of the patient's drug concentration gradient brain imaging data, clinical data, and patient physiological data.

[0057] After obtaining the drug concentration gradient brain imaging data, the first digital twin model of the patient's drug concentration gradient brain imaging data, clinical data, and patient physiological data can be constructed, so that the first digital twin model can accurately predict the tumor treatment effect by combining the drug concentration gradient brain imaging data, and can improve the digital simulation effect of the tumor treatment of the first digital twin model.

[0058] The beneficial effect of the above implementation manner is that based on the first digital twin model that can characterize the growth direction and evolution speed characteristics of tumors at different stages, the tumor treatment effect can be accurately predicted by combining the drug concentration gradient brain imaging data, and the digital simulation effect of the tumor treatment of the first digital twin model can be improved.

[0059] Figure 4 This is a schematic flowchart of the third tumor growth prediction method based on digital twins provided by the embodiments of the present application, as Figure 4As shown above, the above method further includes S310 to S320, and the following is a specific description of S310 to S320.

[0060] S310. Determine the normal tissue region and the tumor infiltration region of the patient's brain imaging data, and obtain the normal tissue characteristics corresponding to the normal tissue region and the infiltration region characteristics corresponding to the tumor infiltration region.

[0061] In the first digital twin model, due to the differences in the integrity of the blood-brain barrier, the characteristics of the tumor microenvironment, and the drug delivery mechanism between the normal tissue region and the tumor infiltration region, in order to further improve the prediction effect of the first digital twin model on the tumor growth state, in the embodiments of the present application, digital simulations can be performed on the normal tissue region and the tumor infiltration region respectively. First, the normal tissue region and the tumor infiltration region of the patient's brain imaging data can be determined, and the normal tissue drug concentration characteristics corresponding to the normal tissue region and the infiltration region drug concentration characteristics corresponding to the tumor infiltration region can be obtained. The normal tissue drug concentration characteristics characterize the drug diffusion characteristics of the normal tissue region, and the infiltration region drug concentration characteristics characterize the drug diffusion characteristics of the infiltration region.

[0062] Figure 5 This is a schematic diagram of tissues in the normal tissue region and the tumor infiltration region provided by the embodiments of the present application. As Figure 5 shown in FIG. a, the tumor cells 1a in the normal tissue region are located within the tissue epithelium and do not penetrate the tissue basement membrane, and the tumor cells 1b in the tumor infiltration region have infiltrated outward into the surrounding tissues to form an invasive tumor.

[0063] Exemplarily, the normal tissue drug concentration characteristics corresponding to the normal tissue region and the infiltration region drug concentration characteristics corresponding to the tumor infiltration region can be obtained by detecting tissue samples, or the normal tissue drug concentration characteristics and the infiltration region drug concentration characteristics can be obtained from a tissue sample database.

[0064] S320. Construct the patient's normal tissue drug concentration gradient brain imaging data according to the patient's normal tissue region, normal tissue drug concentration characteristics, blood-brain barrier data, and dynamic microenvironment data. Construct the patient's infiltration region drug concentration gradient brain imaging data according to the patient's tumor infiltration region, infiltration region drug concentration characteristics, blood-brain barrier data, and dynamic microenvironment data.

[0065] After obtaining the normal tissue drug concentration characteristics, the patient's normal tissue drug concentration gradient brain imaging data can be constructed according to the patient's normal tissue region, normal tissue drug concentration characteristics, blood-brain barrier data, and dynamic microenvironment data. The normal tissue drug concentration gradient brain imaging data characterizes the data of the drug concentration at each position in the normal tissue region.

[0066] After obtaining the drug concentration characteristics of the infiltrated area, the drug concentration gradient brain image data of the patient can be constructed based on the tumor infiltrated area of the patient, the drug concentration characteristics of the infiltrated area, the blood-brain barrier data, and the dynamic microenvironment data. The drug concentration gradient brain image data of the infiltrated area represents the data of the drug concentration at each position in the infiltrated area.

[0067] In some implementation manners, in the above S120, constructing a first digital twin model of the patient's brain image data, clinical data, and patient physiological data includes: constructing a first digital twin model of the drug concentration gradient brain image data of the patient's normal tissue, the drug concentration gradient brain image data of the infiltrated area, clinical data, and patient physiological data.

[0068] After obtaining the drug concentration gradient brain image data of the normal tissue and the drug concentration gradient brain image data of the infiltrated area, a first digital twin model of the drug concentration gradient brain image data of the patient's normal tissue, the drug concentration gradient brain image data of the infiltrated area, clinical data, and patient physiological data can be constructed. Furthermore, the tumor treatment effect simulation of the drug concentration in the normal area and the infiltrated area can be carried out according to the drug concentration gradient brain image data of the normal tissue and the drug concentration gradient brain image data of the infiltrated area respectively, which can improve the digital simulation effect of tumor treatment of the first digital twin model.

[0069] The beneficial effect of the above implementation manner is that the tumor treatment effect simulation of the drug concentration in the normal area and the infiltrated area can be carried out according to the drug concentration gradient brain image data of the normal tissue and the drug concentration gradient brain image data of the infiltrated area respectively. Based on the first digital twin model that can characterize the growth direction and evolution speed characteristics of tumors at different stages, the digital simulation effect of tumor treatment of the first digital twin model is improved.

[0070] In some implementation manners, the above method further includes: determining the blood-brain barrier correction data of the patient through the first digital twin model according to the blood-brain barrier data of the patient. The blood-brain barrier correction data represents the change information of the blood-brain barrier data.

[0071] In order to improve the accuracy of the blood-brain barrier data, in the embodiments of the present application, the blood-brain barrier correction data of the patient can also be determined through the first digital twin model. Since the first digital twin model includes the growth direction and evolution speed characteristics of tumors at different stages, the growth direction and evolution speed characteristics of tumors at different stages can represent the change situation of the blood-brain barrier data to a certain extent. Furthermore, the blood-brain barrier correction data of the patient can be determined according to the blood-brain barrier data of the patient. The blood-brain barrier correction data represents the change information of the blood-brain barrier data. Furthermore, the simulation effect of the second digital twin model can be improved through the blood-brain barrier correction data.

[0072] In some implementations, in the above S120, constructing a second digital twin model of the patient's blood-brain barrier data, drug interaction data, and dynamic microenvironment data includes: constructing a second digital twin model of the patient's blood-brain barrier correction data, drug interaction data, and dynamic microenvironment data.

[0073] After obtaining the blood-brain barrier correction data, a second digital twin model of the patient's blood-brain barrier correction data, drug interaction data, and dynamic microenvironment data can be constructed, and then the effect of microscopic mechanism modeling can be further improved through the blood-brain barrier correction data in the second digital twin model, and the digital simulation effect of the tumor treatment process can be improved.

[0074] The beneficial effect of the above implementation is that the effect of microscopic mechanism modeling is further improved through the blood-brain barrier correction data in the second digital twin model, and the digital simulation effect of the tumor treatment process is improved.

[0075] Figure 6 The flowchart of the fourth tumor growth prediction method based on digital twin provided by the embodiments of the present application is shown in Figure 6 As shown, the above method further includes S410 to S420, and the following is a specific description of S410 to S420.

[0076] S410. Obtain the normal region blood-brain barrier characteristics corresponding to the normal tissue region and the infiltrated region blood-brain barrier characteristics corresponding to the tumor infiltrated region. According to the patient's blood-brain barrier data, the normal tissue region of the brain imaging data, and the normal region blood-brain barrier characteristics, determine the normal region blood-brain barrier data corresponding to the normal tissue region. According to the patient's blood-brain barrier data, the tumor infiltrated region of the brain imaging data, and the infiltrated region blood-brain barrier characteristics, determine the infiltrated region blood-brain barrier data corresponding to the tumor infiltrated region.

[0077] After determining the normal tissue region and the tumor infiltrated region, since the normal tissue region and the tumor infiltrated region have different blood-brain barrier characteristics respectively, the blood-brain barrier simulation effect of the second digital twin model on the normal tissue region and the tumor infiltrated region can be further improved.

[0078] When further improving the blood-brain barrier simulation effect of the first digital twin model on the normal tissue region and the tumor infiltrated region, the normal region blood-brain barrier characteristics corresponding to the normal tissue region and the infiltrated region blood-brain barrier characteristics corresponding to the tumor infiltrated region can be obtained, and then the blood-brain barrier simulation effect of the second digital twin model can be based on the normal region blood-brain barrier characteristics and the infiltrated region blood-brain barrier characteristics.

[0079] Exemplarily, the normal regional blood-brain barrier characteristics corresponding to the normal tissue region and the infiltrated regional blood-brain barrier characteristics corresponding to the tumor-infiltrated region can be determined by dynamic contrast-enhanced MRI. For example, the normal regional blood-brain barrier characteristics can reflect the permeability of the blood-brain barrier by measuring the transfer coefficient (Ktrans) of the contrast agent leaking from the blood vessel lumen to the tissue interstitium in the normal tissue region.

[0080] After obtaining the normal regional blood-brain barrier characteristics and the infiltrated regional blood-brain barrier characteristics, the normal regional blood-brain barrier data corresponding to the normal tissue region can be determined based on the patient's blood-brain barrier data, the normal tissue region of the brain imaging data, and the normal regional blood-brain barrier characteristics, realizing the digital simulation of the normal regional blood-brain barrier data.

[0081] After obtaining the normal regional blood-brain barrier characteristics and the infiltrated regional blood-brain barrier characteristics, the infiltrated regional blood-brain barrier data corresponding to the tumor-infiltrated region can be determined based on the patient's blood-brain barrier data, the tumor-infiltrated region of the brain imaging data, and the infiltrated regional blood-brain barrier characteristics, realizing the digital simulation of the infiltrated regional blood-brain barrier data.

[0082] S420. Through the first digital twin model, determine the corrected normal regional blood-brain barrier data of the patient based on the patient's normal regional blood-brain barrier data. Through the first digital twin model, determine the corrected infiltrated regional blood-brain barrier data of the patient based on the patient's infiltrated regional blood-brain barrier data.

[0083] After obtaining the normal regional blood-brain barrier data, the corrected normal regional blood-brain barrier data of the patient can be determined through the first digital twin model based on the patient's normal regional blood-brain barrier data, realizing the correction of the normal regional blood-brain barrier data.

[0084] After obtaining the infiltrated regional blood-brain barrier data, the corrected infiltrated regional blood-brain barrier data of the patient can be determined through the first digital twin model based on the patient's infiltrated regional blood-brain barrier data, realizing the correction of the infiltrated regional blood-brain barrier data.

[0085] In some implementation manners, in the above S120, constructing the second digital twin model of the patient's blood-brain barrier data, drug interaction data, and dynamic microenvironment data includes: constructing the second digital twin model of the patient's corrected normal regional blood-brain barrier data, corrected infiltrated regional blood-brain barrier data, drug interaction data, and dynamic microenvironment data.

[0086] After obtaining the blood-brain barrier correction data for the normal region and the blood-brain barrier correction data for the infiltrated region, a second digital twin model of the patient's blood-brain barrier correction data for the normal region, blood-brain barrier correction data for the infiltrated region, drug interaction data, and dynamic microenvironment data can be constructed, enabling the second digital twin model to improve the simulation effect of the blood-brain barrier in the normal tissue region and the tumor-infiltrated region.

[0087] The beneficial effect of the above implementation is that the normal tissue region and the tumor-infiltrated region have different blood-brain barrier characteristics. By constructing the blood-brain barrier data for the normal tissue region and the blood-brain barrier data for the infiltrated region and correcting the blood-brain barrier, a second digital twin model of the patient's blood-brain barrier correction data for the normal region, blood-brain barrier correction data for the infiltrated region, drug interaction data, and dynamic microenvironment data is further constructed, enabling the second digital twin model to improve the simulation effect of the blood-brain barrier in the normal tissue region and the tumor-infiltrated region, and improving the digital simulation effect of tumor treatment.

[0088] Figure 7 The following is a schematic flowchart of the fifth tumor growth prediction method based on digital twins provided by the embodiments of the present application. As Figure 7 shown, the above method further includes S510 to S530, which will be specifically described below.

[0089] S510. Obtain multiple similar first digital twin models of the first digital twin model and multiple similar second digital twin models of the second digital twin model, and obtain the treatment decision directions corresponding to the multiple similar first digital twin models respectively and the subsequent first digital twin models corresponding to the treatment decision directions. Also obtain the treatment decision directions corresponding to the multiple similar second digital twin models respectively and the subsequent second digital twin models corresponding to the treatment decision directions. Among them, the treatment decision directions include aggressive treatment methods and conservative treatment methods.

[0090] When performing digital twin simulation on tumors, multiple similar first digital twin models of the first digital twin model and multiple similar second digital twin models of the second digital twin model can be obtained, and then the simulation effect of the tumor treatment process can be further improved based on the multiple similar first digital twin models and multiple similar second digital twin models.

[0091] When performing digital twin simulation, the treatment decision directions corresponding to the multiple similar first digital twin models respectively and the subsequent first digital twin models corresponding to the treatment decision directions can also be obtained. The subsequent first digital twin model is the digital twin model of the similar first digital twin model after tumor development in tumor treatment. Then, the subsequent development of the first digital twin model can be simulated with the assistance of the subsequent first digital twin model.

[0092] Similarly, treatment decision directions corresponding to multiple similar second digital twin models and subsequent second digital twin models corresponding to the treatment decision directions can be obtained, and the subsequent development of the tumors of the second digital twin models can be simulated based on the subsequent second digital twin models.

[0093] It should be noted that the treatment decision directions include radical treatment methods and conservative treatment methods. Furthermore, the treatment decision directions can be assisted in judgment based on the radical treatment methods or conservative treatment methods to improve the more accurate judgment of the treatment directions.

[0094] S520. Among multiple similar second digital twin models, target similar second digital twin models corresponding to multiple similar first digital twin models are determined according to the parameter similarity of the brain imaging data of the similar first digital twin models and the blood-brain barrier data of the similar second digital twin models.

[0095] After obtaining multiple similar first digital twin models of the first digital twin model and multiple similar second digital twin models of the second digital twin model, in order to determine a similar model group that can best simulate the first digital twin model and the second digital twin model, among multiple similar second digital twin models, target similar second digital twin models corresponding to multiple similar first digital twin models are determined according to the parameter similarity of the brain imaging data of the similar first digital twin models and the blood-brain barrier data of the similar second digital twin models. The blood-brain barrier data of each target similar second digital twin model is similar to the brain imaging data of the corresponding similar first digital twin model.

[0096] Exemplarily, when determining target similar second digital twin models corresponding to multiple similar first digital twin models according to the parameter similarity of the brain imaging data of the similar first digital twin models and the blood-brain barrier data of the similar second digital twin models, a parameter similarity comparison model can be used to extract the parameter similarity of the brain imaging data of the similar first digital twin models and the blood-brain barrier data of the similar second digital twin models, and target similar second digital twin models corresponding to multiple similar first digital twin models are determined in the order from high to low of the parameter similarity.

[0097] Exemplarily, the parameter similarity comparison model can be trained according to the labeled brain imaging data of the similar first digital twin models, the blood-brain barrier data of the similar second digital twin models, and the parameter similarity.

[0098] S530. When the treatment decision directions of the similar first digital twin model and the target similar second digital twin model are the same, the similar first digital twin model and the target similar second digital twin model are used as a predicted treatment decision model group, and the treatment decision direction corresponding to the predicted treatment decision model group is used as the recommended treatment decision direction.

[0099] After obtaining the similar first digital twin model and the target similar second digital twin model, when the treatment decision directions of the similar first digital twin model and the target similar second digital twin model are the same, it indicates that the similar first digital twin model and the target similar second digital twin model are subsequent digital twin models obtained according to the same treatment method. Therefore, the similar first digital twin model and the target similar second digital twin model can be used as a predicted treatment decision model group, and the treatment decision direction corresponding to the predicted treatment decision model group is used as the recommended treatment decision direction to realize the recommendation of the treatment decision direction.

[0100] The beneficial effect of the above implementation method is that the similar first digital twin model and the similar second digital twin model are respectively determined, and the tumor treatment process is simulated according to the similar first digital twin model and the similar second digital twin model, improving the simulation effect of the tumor treatment process.

[0101] The beneficial effect of the above implementation method is also that by determining the similar first digital twin model and the target similar second digital twin model, and determining the same treatment decision direction of the similar first digital twin model and the target similar second digital twin model as the recommended treatment decision direction, the accuracy of the recommendation of the treatment decision direction is improved.

[0102] Figure 8 This is the flowchart of the sixth tumor growth prediction method based on digital twins provided by the embodiments of the present application. As Figure 8 shown, the above method further includes S610 to S620, and the following is a specific description of S610 to S620.

[0103] S610. Obtain the first subsequent tumor growth amplitude corresponding to the similar first digital twin model and the second subsequent tumor growth amplitude corresponding to the target similar second digital twin model in each predicted treatment decision model group.

[0104] When simulating the tumor treatment process, it is also possible to further obtain the first subsequent tumor growth amplitude corresponding to the similar first digital twin model and the second subsequent tumor growth amplitude corresponding to the target similar second digital twin model in each predicted treatment decision model group, and evaluate the tumor treatment effect according to the first subsequent tumor growth amplitude and the second subsequent tumor growth amplitude.

[0105] Exemplarily, the first subsequent tumor growth amplitude and the second subsequent tumor growth amplitude can be the corresponding area increase amplitudes in the tumor region images of the similar first digital twin model and the target similar second digital twin model, respectively.

[0106] S620. Determine the absolute value of the difference between the first subsequent tumor growth amplitude and the second subsequent tumor growth amplitude, and the ratio of the sum of the first subsequent tumor growth amplitude and the second subsequent tumor growth amplitude as the subsequent tumor growth amplitude difference degree. Recommend the proposed treatment decision direction in ascending order of the subsequent tumor growth amplitude difference degrees of multiple prediction treatment decision model groups.

[0107] After obtaining the first subsequent tumor growth amplitude and the second subsequent tumor growth amplitude, the absolute value of the difference between the first subsequent tumor growth amplitude and the second subsequent tumor growth amplitude, and the ratio of the sum of the first subsequent tumor growth amplitude and the second subsequent tumor growth amplitude can be determined as the subsequent tumor growth amplitude difference degree. The subsequent tumor growth amplitude difference degree characterizes the ratio of the difference and the sum of the increase amplitudes corresponding to the first subsequent tumor growth amplitude and the second subsequent tumor growth amplitude, that is, it can characterize the difference in the tumor increase amplitudes corresponding to the first subsequent tumor growth amplitude and the second subsequent tumor growth amplitude.

[0108] After obtaining the subsequent tumor growth amplitude difference degree, when the subsequent tumor growth amplitude difference degree is lower, it indicates that the difference degree of the tumor increase amplitudes simulated by the similar first digital twin model and the target similar second digital twin model is smaller, that is, the simulation effect of the tumor growth process is better; when the subsequent tumor growth amplitude difference degree is higher, it indicates that the difference degree of the tumor increase amplitudes simulated by the similar first digital twin model and the target similar second digital twin model is larger, that is, the simulation effect of the tumor growth process is worse; thus, the proposed treatment decision direction can be recommended in ascending order of the subsequent tumor growth amplitude difference degrees of multiple prediction treatment decision model groups, improving the reliability of the proposed treatment decision direction.

[0109] The beneficial effect of the above implementation method is that the proposed treatment decision direction is recommended in ascending order of the subsequent tumor growth amplitude difference degrees of multiple prediction treatment decision model groups, improving the reliability of the proposed treatment decision direction.

[0110] Figure 9 It is a schematic flowchart of the seventh digital twin-based tumor growth prediction method provided by the embodiments of the present application. As Figure 9 shown, the above method further includes S710 to S720, which will be specifically described below.

[0111] S710. Obtain the first complication assessment value corresponding to the similar first digital twin model in each predicted treatment decision model group and the second complication assessment value corresponding to the target similar second digital twin model. The complication assessment value is used to characterize the degree of impact of complications on the patient's quality of life.

[0112] When evaluating the direction of tumor treatment, it is also possible to obtain the first complication assessment value corresponding to the similar first digital twin model in each predicted treatment decision model group and the second complication assessment value corresponding to the target similar second digital twin model. Furthermore, the complications in the patient's tumor treatment can be considered based on the first complication assessment value and the second complication assessment value, reducing the impact of complications on the quality of life suffered by the patient. The complication assessment value is used to characterize the degree of impact of complications on the patient's quality of life.

[0113] Exemplarily, the first complication assessment value corresponding to the similar first digital twin model in each predicted treatment decision model group and the second complication assessment value corresponding to the target similar second digital twin model can be determined based on the experience corresponding to the treatment direction.

[0114] Exemplarily, the complication assessment value can be represented by the sum of the impact score values of all complications on life.

[0115] S720. Determine the absolute value of the difference between the first complication assessment value and the second complication assessment value and the ratio of the sum of the first complication assessment value and the second complication assessment value as the complication assessment value difference degree. Determine the sum of the subsequent tumor growth amplitude difference degree and the complication assessment value difference degree of each predicted treatment decision model group as the comprehensive treatment direction assessment value. Recommend the recommended treatment decision direction in ascending order of the comprehensive treatment direction assessment values of multiple predicted treatment decision model groups.

[0116] After obtaining the first complication assessment value and the second complication assessment value, it is possible to determine the absolute value of the difference between the first complication assessment value and the second complication assessment value and the ratio of the sum of the first complication assessment value and the second complication assessment value as the complication assessment value difference degree, and the complication assessment value difference degree characterizes the difference between the first complication assessment value and the second complication assessment value.

[0117] After obtaining the difference degree of the complication evaluation value, the sum of the subsequent tumor growth amplitude difference degree and the complication evaluation value difference degree of each predicted treatment decision model group can be determined as the comprehensive treatment direction evaluation value. The comprehensive treatment direction evaluation value represents the comprehensive evaluation value obtained from the perspective of the tumor increase amplitude difference and the complication difference. When the comprehensive treatment direction evaluation value is larger, it indicates that the simulation effect of the predicted treatment decision model group is worse; when the comprehensive treatment direction evaluation value is smaller, it indicates that the simulation effect of the predicted treatment decision model group is better. Furthermore, the recommended treatment decision direction can be recommended in ascending order of the comprehensive treatment direction evaluation values of multiple predicted treatment decision model groups, so as to recommend the treatment decision direction of the predicted treatment decision model group with a better simulation effect to the doctor and improve the reliability of the recommended treatment decision direction.

[0118] The beneficial effect of the above implementation manner is that the complication evaluation value difference degree represents the difference between the first complication evaluation value and the second complication evaluation value. The sum of the subsequent tumor growth amplitude difference degree and the complication evaluation value difference degree of each predicted treatment decision model group is determined as the comprehensive treatment direction evaluation value. The recommended treatment decision direction is recommended in ascending order of the comprehensive treatment direction evaluation values of multiple predicted treatment decision model groups, so as to recommend the treatment decision direction of the predicted treatment decision model group with a better simulation effect to the doctor and improve the reliability of the recommended treatment decision direction.

[0119] The embodiment of the present application further provides a tumor growth prediction system based on digital twins, including a unit for executing the method described in any one of the above.

[0120] Figure 10 It is a schematic logical structure diagram of a tumor growth prediction system based on digital twins provided by an embodiment of the present application, as Figure 10 shown, the system 1 of this embodiment includes a processing unit 11, a storage unit 12, and a transceiver unit 13. The processing unit 11 is used to process data, the storage unit 12 is used to store data, and the transceiver unit 13 is used to send and receive data. The processing unit 11, the storage unit 12, and the transceiver unit 13 cooperate with each other to implement the above method. The beneficial effects of the embodiments of the present application have been described in the above method and will not be repeated here.

[0121] It should be noted that the information interaction, execution process, etc. between the above devices / units, due to being based on the same concept as the method embodiment of the present application, for their specific functions and the technical effects brought, please refer to the method embodiment part specifically, and will not be repeated here.

[0122] Those skilled in the art can clearly understand that, for the convenience and conciseness of description, only the above division of each functional unit and module is used as an example. In actual applications, the above functions can be allocated to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit. In addition, the specific names of each functional unit and module are only for the convenience of mutual distinction and do not limit the protection scope of this application. The specific working process of the units and modules in the above system can refer to the corresponding process in the foregoing method embodiment and will not be elaborated herein.

[0123] If the above integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, to implement all or part of the processes in the above method embodiments of this application, a computer program can be used to instruct the relevant hardware to complete. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium can at least include: any entity or device that can carry the computer program code to the photographing device / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal, and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disc, etc. In some jurisdictions, according to legislation and patent practice, the computer-readable medium cannot be an electrical carrier signal and a telecommunication signal.

[0124] In the above embodiments, the descriptions of each embodiment have their own emphases. For the parts not detailed or recorded in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0125] Those of ordinary skill in the art may be aware that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0126] In the embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules or units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of the devices or units can be in an electrical, mechanical or other form.

[0127] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place, or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0128] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features. And these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included in the protection scope of this application.

Claims

1. A tumor growth prediction method based on digital twin, characterized in that, The method includes: Constructing a digital twin database of the multimodal pathological data of a patient, where the multimodal pathological data includes the patient's brain imaging data, clinical data, patient physiological data, blood-brain barrier data, drug interaction data, and dynamic microenvironment data; Constructing a first digital twin model of the patient's brain imaging data, clinical data, and patient physiological data, where the first digital twin model reflects the growth direction and evolution speed characteristics of the tumor at different stages of the patient; and constructing a second digital twin model of the patient's blood-brain barrier data, drug interaction data, and dynamic microenvironment data, where the second digital twin model reflects the treatment microenvironment characteristics of the tumor at different stages of the patient; Through the first digital twin model, determining a first treatment prediction result according to the patient's treatment plan; and through the second digital twin model, determining a second treatment prediction result according to the patient's treatment plan; through a treatment plan correction model, determining a corrected treatment plan for the patient according to the patient's treatment plan, the first treatment prediction result, and the second treatment prediction result.

2. The method according to claim 1, characterized in that The method further includes: Through a blood-brain barrier detection model, determining the blood-brain barrier characteristics of the patient according to the patient's brain imaging data and blood-brain barrier data; Through a treatment plan correction model, determining a corrected treatment plan for the patient according to the patient's treatment plan, the first treatment prediction result, the second treatment prediction result, and the blood-brain barrier characteristics.

3. The method according to claim 2, wherein The method further includes: According to the patient's brain imaging data, blood-brain barrier data, and dynamic microenvironment data, constructing drug concentration gradient brain imaging data corresponding to the patient's brain imaging data; Constructing a first digital twin model of the patient's brain imaging data, clinical data, and patient physiological data, including: Constructing a first digital twin model of the patient's drug concentration gradient brain imaging data, clinical data, and patient physiological data.

4. The method according to claim 3, wherein The method further includes: Determining the normal tissue region and tumor infiltration region of the patient's brain imaging data, and obtaining the normal tissue characteristics corresponding to the normal tissue region and the infiltration region characteristics corresponding to the tumor infiltration region; According to the patient's normal tissue region, normal tissue characteristics, blood-brain barrier data, and dynamic microenvironment data, constructing normal tissue drug concentration gradient brain imaging data for the patient; according to the patient's tumor infiltration region, infiltration region characteristics, blood-brain barrier data, and dynamic microenvironment data, constructing infiltration region drug concentration gradient brain imaging data for the patient; Constructing a first digital twin model of the patient's brain imaging data, clinical data, and patient physiological data, including: Constructing a first digital twin model of the patient's normal tissue drug concentration gradient brain imaging data, infiltration region drug concentration gradient brain imaging data, clinical data, and patient physiological data.

5. The method according to claim 4, wherein The method further includes: Through the first digital twin model, determining blood-brain barrier correction data of the patient according to the patient's blood-brain barrier data, where the blood-brain barrier correction data represents the change information of the blood-brain barrier data; Constructing a second digital twin model of the patient's blood-brain barrier data, drug interaction data, and dynamic microenvironment data, including: Construct a second digital twin model for the patient's blood-brain barrier correction data, drug interaction data, and dynamic microenvironment data.

6. The method according to claim 5, characterized in that, The method further includes: Obtain the normal region blood-brain barrier characteristics corresponding to the normal tissue region and the infiltrated region blood-brain barrier characteristics corresponding to the tumor-infiltrated region; determine the normal region blood-brain barrier data corresponding to the normal tissue region according to the patient's blood-brain barrier data, the normal tissue region of the brain imaging data, and the normal region blood-brain barrier characteristics; determine the infiltrated region blood-brain barrier data corresponding to the tumor-infiltrated region according to the patient's blood-brain barrier data, the tumor-infiltrated region of the brain imaging data, and the infiltrated region blood-brain barrier characteristics; Through the first digital twin model, determine the normal region blood-brain barrier correction data of the patient according to the normal region blood-brain barrier data of the patient; through the first digital twin model, determine the infiltrated region blood-brain barrier correction data of the patient according to the infiltrated region blood-brain barrier data of the patient; Construct a second digital twin model for the patient's blood-brain barrier data, drug interaction data, and dynamic microenvironment data, including: Construct a second digital twin model for the patient's normal region blood-brain barrier correction data, infiltrated region blood-brain barrier correction data, drug interaction data, and dynamic microenvironment data.

7. The method according to claim 6, characterized in that The method further includes: Obtain multiple similar first digital twin models of the first digital twin model and multiple similar second digital twin models of the second digital twin model, obtain the treatment decision directions corresponding to the multiple similar first digital twin models respectively and the subsequent first digital twin models corresponding to the treatment decision directions; and obtain the treatment decision directions corresponding to the multiple similar second digital twin models respectively and the subsequent second digital twin models corresponding to the treatment decision directions; wherein, the treatment decision directions include aggressive treatment methods and conservative treatment methods; Among the multiple similar second digital twin models, determine the target similar second digital twin models corresponding to the multiple similar first digital twin models respectively according to the parameter similarity of the brain imaging data of the similar first digital twin models and the blood-brain barrier data of the similar second digital twin models; When the treatment decision directions of the similar first digital twin model and the target similar second digital twin model are the same, use the similar first digital twin model and the target similar second digital twin model as a prediction treatment decision model group, and use the treatment decision direction corresponding to the prediction treatment decision model group as the recommended treatment decision direction.

8. The method according to claim 7, characterized in that, The method further includes: Obtain the first subsequent tumor growth amplitude corresponding to the similar first digital twin model in each prediction treatment decision model group and the second subsequent tumor growth amplitude corresponding to the target similar second digital twin model; Determine the absolute value of the difference between the first subsequent tumor growth amplitude and the second subsequent tumor growth amplitude and the ratio of the sum of the first subsequent tumor growth amplitude and the second subsequent tumor growth amplitude as the subsequent tumor growth amplitude difference degree; recommend the recommended treatment decision direction in ascending order of the subsequent tumor growth amplitude difference degree of the multiple prediction treatment decision model groups.

9. The method according to claim 8, characterized in that, The method further includes: Obtain the first complication evaluation value corresponding to the similar first digital twin model in each predicted treatment decision model group and the second complication evaluation value corresponding to the target similar second digital twin model; wherein, the complication evaluation value is used to characterize the degree of impact of the complication on the patient's quality of life. Determine the absolute value of the difference between the first complication evaluation value and the second complication evaluation value and the ratio of the sum of the first complication evaluation value and the second complication evaluation value as the complication evaluation value difference degree; determine the sum of the subsequent tumor growth amplitude difference degree and the complication evaluation value difference degree of each predicted treatment decision model group as the comprehensive treatment direction evaluation value; recommend the recommended treatment decision direction in ascending order of the comprehensive treatment direction evaluation value of multiple predicted treatment decision model groups.

10. A tumor growth prediction system based on digital twin, characterized in that, It includes units for performing the method according to any one of claims 1 to 9.

Images