FAPI PET image-based tumor treatment dose optimization method and system

Abstract

The application discloses a FAPI PET image-based tumor treatment dose optimization method and system, and belongs to the technical field of tumor treatment dose optimization. The application aims at the problems of poor applicability of the existing targeted fibroblast activation protein biological image evaluation method, lagging evaluation of treatment sensitivity, and limitations of the FDG PET tumor dose-response matrix model, and through collecting baseline FAPI PET images of tumor patients or experimental animals and mid-treatment feedback images, image deformation registration and target delineation are carried out; the SUV value and volume of the image target region voxels are obtained, and the tumor treatment sensitivity is calculated; the tumor voxel control probability is calculated according to the sensitivity, and then the tumor treatment dose is obtained, so that the early and accurate monitoring of tumor sensitivity is realized, and the adjustment of the tumor treatment scheme and the optimization of the dose are effectively guided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of tumor treatment dose optimization, and particularly relates to a FAPI PET image-based tumor treatment dose optimization method and system. BACKGROUND

[0002] Tumor sensitivity is an important factor for determining treatment decisions and influencing treatment outcomes, but the in vivo monitoring of the same in the prior art still has deficiencies.

[0003] 1. The existing evaluation standard based on the change ratio of the uptake value of the imaging agent of the tumor during the treatment process: such as the PERCIST1.0, EORTC and other solid tumor evaluation standards, which adopt the method of comparing the change degree of the standard uptake value (SUV) of the lesion on the Fluorodeoxyglucose Positron Emission Tomography (FDG PET) image before and after the treatment to represent the sensitivity of the lesion to the treatment. However, this method is only applicable to FDG PET, and the applicability of the biological image of the target Fibroblast-activated protein (FAP) is unknown. Moreover, this method needs to be evaluated at least one month after the treatment is completed, which has a lagging nature and cannot provide early feedback of the lesion information, so the guiding value for the timely adjustment of the treatment scheme is limited.

[0004] 2. The existing tumor dose-response matrix model based on the FDG PET image: the tumor dose response matrix (DRM) model can dynamically monitor the survival fraction in the human body from the image level and reveal the tumor sensitivity. The model has the advantages of early, dynamic evaluation, quantification, voxel-level evaluation and interpretability. However, the DRM model constructed by using the FDG PET still has the following deficiencies: (1) the model is based on the assumption that the change of the SUV value of the FDG PET image can simulate the change of the tumor cell density. However, the level of cell metabolism cannot directly correspond to the size of cell density. In addition, the uptake of FDG is not only from the tumor cells, but also from inflammation, immune cell activity and non-tumor cell metabolism. Therefore, it is not certain whether the assumption is completely true in the tumor tissue. (2) Application limitation: the existing model is not applicable to the FDG low-uptake tumor. SUMMARY

[0005] In view of the above problems in the prior art, the FAPI PET image-based tumor treatment dose optimization method and system provided by the present application solves the problems that the prior art has poor applicability of the targeted fibroblast activation protein (FAP) biological image, the evaluation of treatment sensitivity has a lag, the assumption of the tumor dose-response matrix model based on the fluorodeoxyglucose positron emission tomography image (FDG PET) is not accurate, the applicability of the model to some tumors is limited, it is difficult to early, accurately and widely monitor the tumor sensitivity, and thus the tumor treatment scheme adjustment and dose optimization cannot be timely and effectively guided.

[0006] In order to achieve the above-mentioned purpose, the technical scheme adopted by the present application is as follows: a FAPI PET image-based tumor treatment dose optimization method, comprising the following steps:

[0007] S1, collecting baseline FAPI PET images and mid-treatment FAPI PET feedback images of tumor patients or experimental animals, and performing deformation registration and target region delineation on the two images;

[0008] S2, obtaining the SUV value and volume of each voxel from the target region of the baseline FAPI PET image and the mapped target region of the mid-treatment FAPI PET feedback image, respectively, and calculating the tumor treatment sensitivity;

[0009] S3, calculating the tumor voxel control probability according to the tumor treatment sensitivity;

[0010] S4, calculating the tumor treatment dose according to the tumor voxel control probability.

[0011] Further, S1 comprises the following steps:

[0012] S11, collecting baseline FAPI PET images of tumor patients or experimental animals;

[0013] S12, collecting mid-treatment FAPI PET feedback images after treatment with a treatment dose D;

[0014] S13, performing deformation registration and target region delineation on the baseline FAPI PET images and the mid-treatment FAPI PET feedback images.

[0015] Further, S13 comprises the following steps:

[0016] S131, performing registration on the baseline FAPI PET images and the mid-treatment FAPI PET feedback images using a deformation registration technology;

[0017] S132, semi-automatically delineating the target region of the baseline FAPI PET images with SUV2.0 as the absolute threshold;

[0018] S133, mapping the deformation vector of the target region on the baseline FAPI PET image to the registered mid-treatment FAPI PET feedback image to obtain a mapped target region on the mid-treatment FAPI PET feedback image.

[0019] Further, the formula for calculating the tumor treatment sensitivity in S2 is: ,

[0020] wherein DRM(v) is the tumor treatment sensitivity, SUV D (v) is the SUV value of the vth voxel in the target region of the baseline FAPI PET image mapped to the vth voxel in the mapped target region of the mid-treatment FAPI PET feedback image, V D (v) is the volume of the vth voxel in the target region of the baseline FAPI PET image mapped to the vth voxel in the mapped target region of the mid-treatment FAPI PET feedback image, SUV0(v) is the SUV value of the vth voxel in the target region of the baseline FAPI PET image, V0(v) is the volume of the vth voxel in the target region of the baseline FAPI PET image, k is a correction constant, D is the treatment dose, VOI is the target region, and v is the number of voxels.

[0021] Further, the formula for calculating the tumor voxel control probability in S3 is: ,

[0022] wherein TVCP[SUV0(v), DRM(v), D] is the tumor voxel control probability under the condition of [SUV0(v), DRM(v), D], exp is the exponential function, γ 50 is the increase in tumor control dose percentage caused by an increase of 1 Gy dose when the tumor voxel control probability is equal to 50%, TCD 50 is the tumor control dose 50%, D is the treatment dose, ln is the logarithmic function, SUV0(v) is the SUV value of the vth voxel in the target region of the baseline FAPI PET image, and DRM(v) is the tumor treatment sensitivity.

[0023] Further, the formula for calculating the tumor treatment dose in S4 is: wherein DPF TCP=P [SUV0(v), DRM(v)] is the tumor treatment dose, d * is the treatment dose determined for the tumor voxel with [SUV0(v), DRM(v)], TVCP[SUV0(v), DRM(v), d * ] is the tumor voxel control probability under the condition of [SUV0(v), DRM(v), d * ], P is the desired tumor voxel control probability, N is the number of voxels of a tumor or lesion, To find that specific dose d * So that when using dose d * The tumor voxel control probability TVCP is equal to .

[0024] A FAPI PET image-based tumor treatment dose optimization system, comprising: a deformation registration and target delineation subsystem, a tumor treatment sensitivity calculation subsystem, a tumor voxel control probability calculation subsystem, and a tumor treatment dose calculation subsystem;

[0025] The deformation registration and target delineation subsystem is used to collect baseline FAPI PET images and treatment mid-term FAPI PET feedback images of tumor patients or experimental animals, and perform deformation registration and target delineation of the two images;

[0026] The tumor treatment sensitivity calculation subsystem is used to obtain the SUV value and volume of each voxel from the target region of the baseline FAPI PET image and the treatment mid-term FAPI PET feedback image, respectively, and calculate the tumor treatment sensitivity;

[0027] The tumor voxel control probability calculation subsystem is used to calculate the tumor voxel control probability according to the tumor treatment sensitivity;

[0028] The tumor treatment dose calculation subsystem is used to calculate the tumor treatment dose according to the tumor voxel control probability.

[0029] The beneficial effects of the present application are:

[0030] 1、The present application gets rid of the dependence of FDG PET on tumor metabolic activity, realizes specific imaging of FAP high-expression tumors (including FDG low-uptake tumors) through FAPI PET imaging targeting fibroblast activation protein (FAP), solves the problem of unknown suitability of existing standards (such as PERCIST1.0) for FAPI PET, and expands the technical application range.

[0031] 2、The FAPI PET of the present application focuses on tumor stromal fibroblasts, reduces the interference of non-tumor factors such as inflammation and immune cell metabolism on SUV values, is closer to the real biological characteristics of tumors compared to FDG PET, and makes the assumption that "SUV value changes reflect tumor sensitivity" more reliable.

[0032] 3、The present application analyzes the treatment mid-term FAPI PET feedback image, captures the response signal of the tumor to the treatment in advance, overcomes the evaluation lag of the prior art, and provides a basis for clinical real-time optimization of dose or switching of treatment plan.

[0033] 4, The present application is based on the voxel-level registration and target delineation of baseline and mid-term images, and calculates the tumor treatment sensitivity (DRM) and voxel control probability (TVCP) by combining SUV value and volume data, realizes the upgrade from "lesion overall evaluation" to "single voxel precise modeling", and avoids the evaluation deviation caused by cell metabolic heterogeneity of FDG PET.

[0034] 5, The present application reverses the optimal treatment dose (DPF) of tumor voxel by TVCP, and puts parameters such as tumor sensitivity (DRM) and baseline SUV value into the dose optimization system, breaks the limitation of traditional fixed dose scheme, realizes "individualized dose planning based on tumor biological characteristics", and can provide quantitative "dose-therapeutic effect" prediction results for doctors, which has more scientific basis than empirical dose scheme and reduces the risk of excessive treatment or insufficient treatment. Thus, the closed loop of "early monitoring-precise modeling-dose optimization" is achieved, and the problem of insufficient clinical guidance is solved. BRIEF DESCRIPTION OF DRAWINGS

[0035] Figure 1 The flowchart of the method of the present application is shown in Figure 1.

[0036] Figure 2 The overall flowchart of Example 1 and Example 2 is shown in Figure 2.

[0037] Figure 3 The DRM result graph in the target area of the patient at another shooting angle is shown in Figure 4.

[0038] Figure 4 The DRM result graph in the target area of the patient at another shooting angle is shown in Figure 4.

[0039] Figure 5 The dose adjustment scheme graph of the patient in the coronal position according to the tumor control probability and the dose irradiation of 32Gy, 50Gy and 86Gy is shown in Figure 5.

[0040] Figure 6 The dose adjustment scheme graph of the patient in the coronal position according to the tumor control probability and the dose irradiation of 32Gy, 50Gy and 86Gy is shown in Figure 5.

[0041] Figure 7 The dose adjustment scheme graph of the patient in the coronal position according to the tumor control probability and the dose irradiation of 32Gy, 50Gy and 86Gy is shown in Figure 5. DETAILED DESCRIPTION

[0042] The specific embodiments of the present application are described below to facilitate the understanding of the present application for those skilled in the art, but it should be clear that the present application is not limited to the scope of the specific embodiments, and for those skilled in the art, it is obvious that various changes are within the spirit and scope of the present application defined and determined by the appended claims, and all the inventions utilizing the concept of the present application are within the scope of protection.

[0043] CAF is an important component of the tumor microenvironment, which can interact with the tumor through various mechanisms: affecting tumor angiogenesis, shaping the immunosuppressive microenvironment, and promoting tumor formation and drug resistance. Fibroblast-activated protein inhibitor (Fibroblast-activated protein inhibitor, FAPI) PET targets the FAP of CAF, which has the following advantages in esophageal cancer evaluation: (1) FAPI and FDG PET reveal different tumor characteristics of esophageal cancer, and have complementary value. (2) The tumor microenvironment and tumor cells do not respond consistently during neoadjuvant therapy of esophageal cancer. FAPI and FDG PET can reveal the response of both after esophageal cancer treatment. (3) FAPI PET has lower background uptake in normal tissues, and has a higher tumor-to-background ratio.

[0044] Therefore, the present application can realize non-invasive, repeatable, dynamic, and evaluation of the sensitivity of the tumor microenvironment to treatment by using the biological image based on FAP targeting and combining the biological principle of tumor dose-response, and can more comprehensively improve the evaluation accuracy of the tumor dose-response situation, retain the heterogeneity in the space (improve the accuracy at the spatial level), and is a technical method of image mining and utilization analysis.

[0045] Embodiment 1, as shown in the following, a FAPI PET image-based tumor treatment dose optimization method, comprising the following steps: Figure 1

[0046] S1, collecting baseline FAPI PET images and mid-treatment FAPI PET feedback images of tumor patients or experimental animals, and performing image deformation registration and target delineation;

[0047] S2, obtaining the SUV value and volume of each voxel from the target region of the baseline FAPI PET image and the mapping target region of the mid-treatment FAPI PET feedback image, respectively, and calculating the tumor treatment sensitivity;

[0048] S3, calculating the tumor voxel control probability according to the tumor treatment sensitivity;

[0049] S4, calculating the tumor treatment dose according to the tumor voxel control probability.

[0050] ​In this embodiment, S1 comprises the following sub-steps:

[0051] S11, collecting baseline FAPI PET images of tumor patients or experimental animals;

[0052] S12, collecting mid-treatment FAPI PET feedback images after treatment with a therapeutic dose D;

[0053] S13, performing deformation registration and target delineation on the baseline FAPI PET images and the mid-treatment FAPI PET feedback images.

[0054] The subject of this embodiment is individuals / lesions that are positive on the baseline FAPI PET images.

[0055] In this embodiment, the medical image acquisition process of the clinical standard is adopted, and the acquisition and reconstruction parameter settings of the baseline and mid-treatment images should be as consistent as possible, and the basic conditions of the patients / experimental animals should be consistent. Measures such as body position fixation can be taken.

[0056] In this embodiment, S13 comprises the following sub-steps:

[0057] S131, performing registration of the baseline FAPI PET images and the mid-treatment FAPI PET feedback images using deformation registration technology;

[0058] S132, semi-automatically delineating the target region of the baseline FAPI PET images with SUV2.0 as the absolute threshold;

[0059] S133, mapping the deformation vector of the target region on the baseline FAPI PET images to the registered mid-treatment FAPI PET feedback images to obtain the mapped target region on the mid-treatment FAPI PET feedback images.

[0060] In step S131, the type of image post-processing technology is not limited, and existing deformation registration technologies can be tried. The technical purpose of this step is to achieve one-to-one correspondence of voxels of two scan images. For example, deformation registration methods include B-spline, Demons algorithm, LDDMM-Net, etc.

[0061] The target delineation standard can be adjusted by the user as needed. This embodiment provides a reference delineation standard: semi-automatic delineation of PET images with SUV2.0 as the absolute threshold, which can obtain the corresponding target region on the anatomical images by mapping to the CT / MRI anatomical images of the same scan. The target region of the FAPI image after treatment is obtained by deformation mapping of the deformation vector of the deformation registration. The actual mapped target region can be corrected to limit the vector of the deformation registration.

[0062] The specific process of step S2 is: the embodiment designs to read the SUV value (SUV0) of each voxel in the target region of the baseline FAPI PET image, and the volume (V0) of the voxel. And through the deformation registration, the SUV value (SUV D ) and the volume (V D ) of the corresponding voxel of the mid-treatment image FAPI PET are located. Wherein D refers to the treatment dose received by the lesion between the two scan images. If the tumor shrinks during treatment, the volume V D of the corresponding voxel on the mid-treatment FAPI PET image should be reduced. If the tumor tissue microenvironment corresponding to the voxel v of the tumor has a down-regulation of FAP expression during treatment, then SUV D (v) should be less than SUV0(v).

[0063] In this embodiment, the formula for calculating the tumor treatment sensitivity in S2 is: ,

[0064] Wherein DRM(v) is the tumor treatment sensitivity, SUV D (v) is the SUV value of the vth voxel in the target region of the baseline FAPI PET image mapped to the vth voxel in the mapped target region of the mid-treatment FAPI PET feedback image, V D (v) is the volume of the vth voxel in the target region of the baseline FAPI PET image mapped to the vth voxel in the mapped target region of the mid-treatment FAPI PET feedback image, SUV0(v) is the SUV value of the vth voxel in the target region of the baseline FAPI PET image, V0(v) is the volume of the vth voxel in the target region of the baseline FAPI PET image, k is a correction constant, D is the treatment dose, VOI is the target region, and v is the number of voxels.

[0065] The voxels in the target region of the baseline FAPI PET image and the mapped target region of the mid-treatment FAPI PET feedback image are in one-to-one correspondence.

[0066] k is used to normalize DRM, so that the average DRM value of all tumor voxels is equal to the average cell survival fraction of 2Gy irradiation measured in vitro esophageal cancer tissue.

[0067] In this embodiment, the formula for calculating the tumor voxel control probability in S3 is: Wherein TVCP[SUV0(v), DRM(v), D] is the tumor voxel control probability under the condition of [SUV0(v), DRM(v), D], exp is the exponential function, γ 50 is the increase in tumor control dose percentage caused by an increase of 1 Gy dose when the tumor voxel control probability is equal to 50%, and TCD 50D50% = 50% of tumor control dose, D is the treatment dose, ln is the natural logarithm, SUV0(v) is the SUV value of the vth voxel in the target region in the baseline FAPI PET image, DRM(v) is the tumor treatment sensitivity.

[0068] (TCD 50 ,γ 50 ) refers to the tumor dose that achieves TVCP = 50% and the increase in TCD% that results from each 1 Gy dose increase at TVCP = 50%. By maximizing the likelihood of tumor voxel control / failure at different dose levels, the relationship between (TCD 50 ,γ 50 ) and (SUV0, DRM) can be determined. Their correspondence can be determined by a given look-up table. The individual tumor control probability (TCP) is the product of all voxel TVCPs.

[0069] In this embodiment, the formula for calculating the tumor treatment dose in S4 is: ,

[0070] where DPF TCP=P [SUV0(v), DRM(v)] is the tumor treatment dose, d * is the treatment dose determined for the tumor voxel with [SUV0(v), DRM(v)], TVCP[SUV0(v), DRM(v), d * ] is the tumor voxel control probability at [SUV0(v), DRM(v), d * ], P is the desired tumor voxel control probability, N is the number of voxels in a tumor or lesion, is the specific dose d * that is found such that when using the dose d * , the tumor voxel control probability TVCP is equal to .

[0071] d * is the corresponding prescription dose determined from DPF for a given tumor voxel with (SUV0, DRM) values. DPF provides a prescription dose for each tumor voxel based on the voxel (SUV0, DRM) values and the desired TCP level. The desired TCP represents the tumor control rate that the clinic expects to achieve for this lesion using this method.

[0072] (TCD 50 ,γ 50 ) and (SUV0, DRM) are related as shown in Table 1.

[0073] Table 1

[0074] TCD 50 ]]> SUV0= 4.5 SUV0= 8.5 SUV0= 12.5 [ S U V 0 = 16.5 ] DRM = 0.2 3.72 7.07 13.62 20.34 DRM = 0.3 4.11 7.79 14.9 22.05 DRM = 0.4 4.6 8.8 17.69 27.18 DRM = 0.5 5.16 10.41 21.91 34.01 DRM = 0.6 5.81 12.77 29.71 40.78 DRM = 0.7 6.45 17.44 39.7 47.65 DRM = 0.8 7.46 22.4 46.08 53.99 DRM = 0.9 8.93 28.03 50.65 58.92 DRM = 1.0 10.58 31.38 55.51 62.74 DRM = 1.1 14.08 36.75 63.1 69.37 DRM = 1.2 15.87 41.53 69.35 74.89

[0075] The embodiment 2 is a FAPI PET image-based tumor treatment dose optimization system, comprising: a deformation registration and target delineation subsystem, a tumor treatment sensitivity calculation subsystem, a tumor voxel control probability calculation subsystem and a tumor treatment dose calculation subsystem;

[0076] The deformation registration and target delineation subsystem is used to collect baseline FAPI PET images and treatment mid-term FAPI PET feedback images of tumor patients or experimental animals, and to perform deformation registration and target delineation of the two images;

[0077] The tumor treatment sensitivity calculation subsystem is used to obtain the SUV value and volume of each voxel from the target region of the baseline FAPI PET image and the treatment mid-term FAPI PET feedback image respectively, and to calculate the tumor treatment sensitivity;

[0078] The tumor voxel control probability calculation subsystem is used to calculate the tumor voxel control probability according to the tumor treatment sensitivity;

[0079] The tumor treatment dose calculation subsystem is used to calculate the tumor treatment dose according to the tumor voxel control probability.

[0080] The specific implementation of the embodiment 2 is consistent with that of the embodiment 1.

[0081] As shown in the following formula (1), the output results are: (1) DRM, representing tumor sensitivity; (2) TVCP, representing the complete control probability of the tumor under a given treatment plan, i.e., the predicted treatment outcome; (3) Dose prescription function (DPF): the prescribed dose that should be given under the condition that the lesion / voxel reaches the expected control probability. Figure 2 Actual clinical application can guide subsequent treatment according to the prediction results to improve the treatment effect: judge the treatment outcome under the existing plan. If it cannot be completely controlled or the treatment effect is not ideal, the plan can be changed or the dose can be increased; if the treatment effect is very good, the treatment dose can be appropriately reduced on the premise of achieving the treatment effect, so as to reduce the toxic and side effects on normal tissues. That is, the calculated tumor treatment sensitivity DRM represents whether the tumor is sensitive, the calculated tumor voxel control probability TVCP represents whether the tumor is cured, and the calculated tumor treatment dose DPF represents how to treat subsequently.

[0082] In this embodiment, a patient with esophageal cancer who received neoadjuvant therapy (including 2 cycles of paclitaxel + carboplatin + tiragolumab + 40 Gy concurrent radiotherapy) is taken as an example. The patient received a baseline 68Ga-FAPI PET / CT image scan before neoadjuvant therapy and another 68Ga-FAPI PET / CT image scan after the completion of neoadjuvant therapy. The patient preparation, scanning, and reconstruction parameters of the two scans were basically kept consistent, making the two images comparable.

[0083] Image registration: Taking the CT of the baseline scan as the reference image, the CT after treatment was deformed and registered based on the finite element method to obtain the deformation vector based on the images of the two scans. This vector acts on the FAPI PET after treatment to achieve voxel-level registration with the baseline FAPI PET image.

[0084] Target delineation: On the baseline FAPI PET image, the absolute threshold of 2.0 was used to semi-automatically delineate the primary lesion of esophageal cancer. The target area contained a total of 272 voxels of 2.68x2.68x3.13mm 3 . It was mapped to the FAPI PET after treatment through the deformation vector. According to the actual mapping result, manual correction was performed. According to the correction result, the deformation vector was limited to make it more consistent with the registration result of manual correction. The running result shows that the target area after actual registration is consistent with the target area defined by hand: Dice similarity coefficient = 0.90. The average Jacobian coefficient is 0.87, indicating that the lesion as a whole shows a shrinking trend.

[0085] Data extraction: The SUV0 and SUV D of all voxels were extracted one by one, and their average values were 3.81 and 1.97, respectively; the maximum values were 7.74 and 3.71, respectively.

[0086] Output result calculation: DRM and TVCP were calculated voxel by voxel to obtain the DRM result of the target area of the patient, as shown in Figure 3 and 4 , Figure 3 and Figure 4 The colored area is the target area, Figure 3 and Figure 4 are the fusion images of DRM and baseline CT, Figure 3 and Figure 4 are the fusion images of DRM and post-treatment CT. The red area is the area where DRM > 1.0.

[0087] The specific statistical results show that the average DRM of the patient is 0.55, the median DRM is 0.48, the maximum DRM is 1.62, the coefficient of variation is 0.49, there are 50 voxels with DRM greater than 0.8, and 20 voxels with DRM greater than 1.0. These voxels are low-sensitivity voxels to treatment and also correspond to high-risk tumor tissue sites that may lead to treatment failure. It can be known from Figure 3 and Figure 4 that the spatial distribution is that the main distribution site is the tumor edge, and the sensitivity of the tumor center site is relatively good.

[0088] The lesion still has 214 voxels remaining in the middle of the treatment, that is, the volume of the tumor remaining on the image is 4810.9 cc. The average dose for complete control of all voxels is 42.56 Gy of biological equivalent dose, but the most resistant voxels to achieve complete control require 102 Gy of biological equivalent dose. That is, a higher treatment dose should be given to the high-risk area of the patient where the treatment fails to achieve complete control. The prescription dose for all voxels is designed according to the actual spatial distribution to design a three-level dose scheme, which can be divided into 86 Gy, 50 Gy and 32 Gy, as shown by the color areas in Figure 5 、 Figure 6 and Figure 7 .

[0089] As shown in Figures 5-7 , it can be known that the volume requiring 86 Gy of biological equivalent dose is only 0.494 cm 3 , and the remaining most of the volume only needs to be given a dose of 50 Gy and below to achieve tumor control, which provides a practical and operable guidance scheme for subsequent treatment.

[0090] The FAPI PET image can be replaced by other biological images targeting CAF, and the purpose of the present application of monitoring the tumor microenvironment treatment dose-response from the perspective of the treatment response of tumor fibroblasts can also be achieved.

[0091] From the accuracy point of view, the best existing technical solution is the tumor dose-response relationship model based on FDG PET and the TVCP / TCP, DPF model derived therefrom. However, this model is limited by the shortcomings of FDG PET, such as: ignoring the role of tumor microenvironment, the information represented is not comprehensive; easily interfered by inflammation during treatment, thereby causing errors in the correspondence between the image and the tumor; some tumors do not uptake FDG, making it impossible to use FDG PET for evaluation; and the physiological uptake of FDG PET in the liver and gastrointestinal tract is obvious, which affects the judgment of the corresponding tumor site.

[0092] The FAPI PET replacing the FDG PET has the following advantages: targeting CAF in the tumor microenvironment, revealing the response of the tumor microenvironment in the treatment process; less interference from inflammation; better diagnostic ability than FDG in some tumor types, especially gastrointestinal tumors, so the image correspondence with the tumor is better than that of FDG PET; the physiological uptake value of FAPI PET in the intracranial, liver and gastrointestinal tract is low, and the interference with tumor judgment is small. Therefore, FAPI PET has better evaluation ability than FDG PET for tumors in the above scenarios, and the tumor dose-response relationship established therefrom is also optimized, including the tumor control probability model and the dose prescription function further derived therefrom.

[0093] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. The present application can be variously changed and modified by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for optimizing tumor treatment dosage based on FAPI PET imaging, characterized in that, Includes the following steps: S1. Acquire baseline FAPI PET images and mid-treatment FAPI PET feedback images of tumor patients or experimental animals, and perform deformable registration and target delineation on the two images; S2. Obtain the SUV value and volume of each voxel from the target area of ​​the baseline FAPI PET image and the mapped target area of ​​the mid-treatment FAPI PET feedback image, respectively, and calculate the tumor treatment sensitivity. The formula for calculating tumor treatment sensitivity in S2 is as follows: , Wherein, DRM(v) represents tumor treatment sensitivity, and SUV D (v) represents the SUV value of the vth ... D (v) represents the volume of the vth vth voxel in the target area mapped from the baseline FAPI PET image to the mid-treatment FAPI PET feedback image; SUV0(v) represents the SUV value of the vth voxel in the target area of ​​the baseline FAPI PET image; V0(v) represents the volume of the vth voxel in the target area of ​​the baseline FAPI PET image; k is the correction factor; D is the treatment dose; VOI is the target area; and v is the voxel number. S3. Calculate the probability of tumor control by voxels based on tumor treatment sensitivity; S4. Calculate the tumor treatment dose based on the tumor voxel control probability.

2. The method for optimizing tumor treatment dosage based on FAPI PET imaging according to claim 1, characterized in that, S1 includes the following steps: S11. Acquire baseline FAPI PET images of tumor patients or laboratory animals; S12. Acquire mid-treatment FAPI PET feedback images after treatment with therapeutic dose D; S13. Perform deformable registration and target delineation on baseline FAPI PET images and mid-treatment FAPI PET feedback images.

3. The method for optimizing tumor treatment dosage based on FAPI PET imaging according to claim 2, characterized in that, S13 includes the following steps: S131. The baseline FAPI PET image and the mid-treatment FAPI PET feedback image are registered using the deformation registration technique; S132, The target area of ​​the baseline FAPI PET image will be semi-automatically delineated using SUV2.0 as the absolute threshold; S133. Map the deformation vector of the target area on the baseline FAPI PET image to the registered mid-treatment FAPI PET feedback image to obtain the mapped target area on the mid-treatment FAPI PET feedback image.

4. The method for optimizing tumor treatment dosage based on FAPI PET imaging according to claim 1, characterized in that, The formula for calculating the tumor voxel control probability in S3 is as follows: , Where TVCP[SUV0(v),DRM(v),D] is the tumor voxel control probability under the condition [SUV0(v),DRM(v),D], exp is the exponential function, and γ 50 The percentage increase in tumor control dose per 1 Gy dose when the probability of tumor control is 50% is represented by the TCD. 50 The tumor control dose is 50%, D is the treatment dose, ln is the logarithmic function, SUV0(v) is the SUV value of the vth vth voxel in the target area of ​​the baseline FAPI PET image, and DRM(v) is the tumor treatment sensitivity.

5. The method for optimizing tumor treatment dosage based on FAPI PET imaging according to claim 1, characterized in that, The formula for calculating the tumor treatment dose in S4 is as follows: , Among them, DPF TCP=P [SUV0(v),DRM(v)] represents the tumor treatment dose, d * The therapeutic dose determined for a tumor voxel with [SUV0(v), DRM(v)] is TVCP[SUV0(v), DRM(v), d * ] is in [SUV0(v), DRM(v), d * The probability of tumor voxel control under certain conditions, where P is the expected probability of tumor voxel control and N is the number of voxels in a tumor or lesion. To find that specific dose d * This makes it possible to use a dose d * At that time, the tumor voxel control probability TVCP equals .

6. A tumor treatment dose optimization system based on FAPI PET imaging, implemented based on the tumor treatment dose optimization method based on FAPI PET imaging as described in any one of claims 1 to 5, characterized in that, include: Subsystems for deformable registration and target delineation, tumor treatment sensitivity calculation, tumor voxel control probability calculation, and tumor treatment dose calculation; The deformable registration and target delineation subsystem is used to acquire baseline FAPI PET images and mid-treatment FAPI PET feedback images of tumor patients or experimental animals, and to perform deformable registration and target delineation on the two images. The tumor treatment sensitivity calculation subsystem is used to obtain the SUV value and volume of each voxel in the target area from the baseline FAPI PET image and the mid-treatment FAPI PET feedback image, respectively, and calculate the tumor treatment sensitivity. The tumor voxel control probability calculation subsystem is used to calculate the tumor voxel control probability based on tumor treatment sensitivity. The tumor treatment dose calculation subsystem is used to calculate the tumor treatment dose based on the tumor voxel control probability.

Images