A three-dimensional dose reconstruction method and system
By using imaging data to match pelvic model and voxel model in close range radiation therapy of cervical cancer, update the radio source position and apply heterogeneity correction factors, the unreliable problem of dose measurement caused by radio source deviation and human heterogeneity is solved. , improves the accuracy and reliability of treatment.
Patent Information
- Application Number
- CN202410588677.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-13
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-05-13
AI Technical Summary
In the near-range radiation therapy of cervical cancer, the prior art fails to effectively consider the radiation source deviation and human heterogeneity, resulting in unreliable dose measurement results in body, affecting the target area and the evaluation of radiation doses that endanger the organ.
By matching the pelvic dimension model based on patient image data, converting it into voxel model, obtaining the true location of the radioactive source, updating the planned location, and using heterogeneity correction factors to correct the three-dimensional dose distribution, considering human heterogeneity and radioactive source shift, improve Reliability of dose measurement results.
It improves the reliability of in-volume dose measurement results, reduces the difference in dose distribution between target areas and hazardous organs, and enhances the accuracy of treatment.
Smart Images

Figure CN118320329B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the application field of brachytherapy, and in particular to a three-dimensional dose reconstruction method and system. Background Art
[0002] For radiotherapy of advanced cervical cancer, external beam radiotherapy combined with afterloading brachytherapy is a crucial component of radical radiotherapy.
[0003] At present, in terms of radiation source tracking, the treatment process in the clinical treatment of cervical cancer at close range, including target area delineation, source reconstruction, plan design and optimization, is very time-consuming. The patient's bladder filling or gastrointestinal motility may change; after the source is inserted, it is difficult for the patient to maintain the same position during the long positioning scan; in addition, the patient will inevitably move during the transfer to the treatment room, affecting the position of the source; these may cause the patient's bladder and rectum, etc., to receive doses at risk. The difference with the prescribed dose distribution is more than 10%. In terms of dose calculation, for nearly 30 years, the dose calculation in clinical cervical cancer brachytherapy has always followed the AAPMTG-43 recommended method, which gives the dose distribution in a uniform water phantom based on the relationship between the activity of the radiation source and the reference air kerma rate. However, since the radiation source offset caused by the patient's anatomical changes during treatment and the dose distribution due to human heterogeneity (bones, cavities, etc.) are not taken into account, there are large errors between the actual in-body dose measurement results and the planned dose distribution, and the radiation doses of the tumor target area and organs at risk cannot be accurately assessed, making the in-body dose measurement results unreliable. Summary of the invention
[0004] Based on the above problems, the present invention proposes a three-dimensional dose reconstruction method and system to solve the problem that the prior art does not consider radiation source offset and human body heterogeneity, resulting in unreliable in-body dose measurement results.
[0005] To achieve the above object, an embodiment of the present invention provides a three-dimensional dose reconstruction method, comprising:
[0006] Matching a pelvic size model that meets preset requirements according to the patient image data, and converting the pelvic size model that meets the preset requirements into a number of voxel models of corresponding sizes;
[0007] According to the voxel models of the corresponding sizes, the planned position of the radiation source is updated by acquiring the real position of the radiation source to obtain an updated real position of the radiation source;
[0008] According to the voxel models of the corresponding sizes, the type of the radiation source at the current moment and the updated real position of the radiation source, a corresponding inhomogeneity correction factor is selected to correct the three-dimensional dose distribution calculation result of each radiation source residence point to obtain a three-dimensional dose distribution correction result of each radiation source residence point;
[0009] The true result of the three-dimensional dose distribution is obtained by accumulating the three-dimensional dose distribution correction results of each radiation source residence point.
[0010] The embodiment of the present invention proposes a three-dimensional dose reconstruction method, which matches a corresponding pelvic size model according to the patient's image data, wherein the pelvic size model considers voxels and thus human body heterogeneity; then the planned position is updated by obtaining the real position of the radiation source, and the offset problem of the radiation source is considered; finally, based on the factors considering the human body heterogeneity and the radiation source offset, the three-dimensional dose is corrected by a non-homogeneous correction factor to obtain the real distribution result of the three-dimensional dose; therefore, the present invention further improves the reliability of the in-body dose measurement result by considering the human body heterogeneity and the radiation source offset problem and then using the non-homogeneous correction factor.
[0011] Furthermore, the patient image data is obtained and a pelvic model that meets preset requirements is matched according to the patient image data, wherein the pelvic size model that meets the preset requirements includes: a plurality of voxel models of different sizes, specifically:
[0012] Acquire historical patient pelvic imaging data, and obtain upper and lower limits of the patient's pelvic size distribution by performing size analysis on the historical patient pelvic imaging data;
[0013] According to the upper and lower limits of the patient's pelvic size distribution, several pelvic size models are constructed by presetting incremental intervals;
[0014] Matching the pelvic size model according to the patient's image data to obtain a pelvic size model that meets the preset requirements;
[0015] According to a first preset algorithm, a voxel model conversion is performed on the pelvic size model that meets the preset requirements to obtain a plurality of voxel models of different sizes, wherein the plurality of voxel models of different sizes include a first model and a second model.
[0016] Furthermore, according to the first preset algorithm, the voxel model conversion is performed on the pelvic size model that meets the preset requirements to obtain a plurality of voxel models of different sizes, wherein the plurality of voxel models of different sizes include a first model and a second model, specifically:
[0017] Acquire corresponding patient image data according to a plurality of pelvic dimension models, and convert the corresponding pelvic dimension model that meets the preset requirements through a first preset algorithm to obtain a corresponding voxel model;
[0018] A first model is obtained by assigning the HU value of the patient's image data to the preset physical density of the voxel model; wherein the first model is a non-homogeneous model;
[0019] The second model is obtained by assigning water density to the voxel model; wherein the second model is a water homogeneous model.
[0020] Furthermore, according to the voxel models of the corresponding sizes, the type of radiation source at the current moment and the updated real position of the radiation source, a corresponding inhomogeneity correction factor is selected to correct the three-dimensional dose distribution calculation result of each radiation source residence point, so as to obtain the three-dimensional dose distribution correction result of each radiation source residence point, specifically:
[0021] By using a preset algorithm and patient image data, respectively calculating the inhomogeneity correction factor for the first model and the second model and constructing a inhomogeneity correction factor database;
[0022] Matching corresponding inhomogeneity correction factors according to the voxel models of the corresponding sizes, the type of radiation source at the current moment and the updated real position of the radiation source;
[0023] According to the corresponding inhomogeneity correction factor, the three-dimensional dose distribution calculation result of each radiation source residence point is corrected by the GPU algorithm to obtain the three-dimensional dose distribution correction result of each radiation source residence point.
[0024] Furthermore, the preset algorithm and the patient image data are used to calculate the inhomogeneity correction factor for the first model and the second model respectively and construct a inhomogeneity correction factor database, specifically:
[0025] According to the first preset algorithm and the second preset algorithm, three-dimensional voxel dose distribution calculations are performed on first models of several sizes and second models of several sizes respectively, to obtain first calculation results corresponding to each size and second calculation results corresponding to each size;
[0026] According to the first calculation results corresponding to the sizes and the second calculation results corresponding to the sizes, the corresponding inhomogeneity correction factors are obtained by calculation, and a inhomogeneity correction factor database is constructed.
[0027] The embodiment of the present invention also provides a three-dimensional dose reconstruction system, including: a pelvic model matching module, a radiation source position updating module, a dose distribution result correction module and a real result acquisition module;
[0028] The pelvic model matching module is used to match a pelvic size model that meets preset requirements according to the patient image data, and convert the pelvic size model that meets the preset requirements into a number of voxel models of corresponding sizes;
[0029] The radiation source position updating module is used to update the planned position of the radiation source by acquiring the real position of the radiation source according to the voxel models of the corresponding sizes, so as to obtain the updated real position of the radiation source;
[0030] The dose distribution result correction module is used to select a corresponding inhomogeneity correction factor to correct the three-dimensional dose distribution calculation result of each radiation source residence point according to the voxel models of the corresponding sizes, the type of radiation source at the current moment and the updated true position of the radiation source, so as to obtain the three-dimensional dose distribution correction result of each radiation source residence point;
[0031] The real result acquisition module is used to obtain the real result of the three-dimensional dose distribution by accumulating the three-dimensional dose distribution correction results of each radiation source residence point.
[0032] The embodiment of the present invention proposes a three-dimensional dose reconstruction system, which matches the corresponding pelvic size model according to the patient's image data through a pelvic model matching module, taking into account the voxels and thus the heterogeneity of the human body; the radiation source position updating module obtains the real position of the radiation source to update the planned position, taking into account the offset problem of the radiation source; and then the dose distribution result correction module corrects the three-dimensional dose through a non-homogeneous correction factor based on the factors taking into account the heterogeneity of the human body and the offset of the radiation source to obtain the real distribution result of the three-dimensional dose; therefore, the present invention further improves the reliability of the in-body dose measurement result by taking into account the heterogeneity of the human body and the offset problem of the radiation source and then using the non-homogeneous correction factor.
[0033] Furthermore, the pelvic model matching module is used to match a pelvic size model that meets preset requirements according to the patient image data, and convert the pelvic size model that meets the preset requirements into a number of voxel models of corresponding sizes, and also includes:
[0034] A pelvic size upper and lower limit acquisition unit, a pelvic size model acquisition unit, a pelvic size model matching unit and a voxel model conversion unit;
[0035] The pelvic size upper and lower limit acquisition unit is used to acquire historical patient pelvic image data, and obtain the upper and lower limits of the patient's pelvic size distribution by performing size analysis on the historical patient pelvic image data;
[0036] The pelvic cavity size model acquisition unit is used to construct and obtain a plurality of pelvic cavity size models by preset increment intervals according to the upper and lower limits of the patient's pelvic cavity size distribution;
[0037] The pelvic cavity size model matching unit is used to match the pelvic cavity size model according to the patient image data to obtain a pelvic cavity size model that meets the preset requirements;
[0038] The voxel model conversion unit is used to perform voxel model conversion on the pelvic size model that meets the preset requirements according to a first preset algorithm to obtain a plurality of voxel models of different sizes, wherein the plurality of voxel models of different sizes include a first model and a second model.
[0039] Further, the voxel model conversion unit is used to perform voxel model conversion on the pelvic size model that meets the preset requirements according to a first preset algorithm to obtain a plurality of voxel models of different sizes, wherein the plurality of voxel models of different sizes include a first model and a second model, and further include:
[0040] a voxel model building subunit, a first model building subunit, and a second model building subunit;
[0041] The voxel model construction subunit is used to obtain corresponding patient image data according to a plurality of pelvic dimension models, and convert the corresponding pelvic dimension model that meets the preset requirements through a first preset algorithm to obtain a corresponding voxel model;
[0042] The first model building subunit is used to obtain a first model by assigning the HU value of the patient image data to the preset physical density of the voxel model; wherein the first model is a non-homogeneous model;
[0043] The second model building subunit is used to obtain a second model by assigning water density to the voxel model; wherein the second model is a water homogeneous model.
[0044] Further, the dose distribution result correction module is used to select a corresponding inhomogeneity correction factor to correct the three-dimensional dose distribution calculation result of each radiation source residence point according to the voxel models of the corresponding sizes, the radiation source type at the current moment and the updated true position of the radiation source, so as to obtain the three-dimensional dose distribution correction result of each radiation source residence point, and also includes:
[0045] A heterogeneous correction factor database construction unit, a heterogeneous correction factor matching unit and a result correction unit;
[0046] The inhomogeneity correction factor database construction unit is used to calculate a number of inhomogeneity correction factors and construct a inhomogeneity correction factor database through a preset algorithm and patient image data;
[0047] The inhomogeneity correction factor matching unit is used to match the corresponding inhomogeneity correction factor according to the voxel models of the corresponding sizes, the type of the radiation source at the current moment and the updated real position of the radiation source;
[0048] The result correction unit is used to correct the three-dimensional dose distribution calculation result of each radiation source residence point through the GPU algorithm according to the corresponding inhomogeneity correction factor to obtain the three-dimensional dose distribution correction result of each radiation source residence point.
[0049] Furthermore, the inhomogeneity correction factor database construction unit is used to calculate a number of inhomogeneity correction factors and construct a inhomogeneity correction factor database through a preset algorithm and patient image data, and also includes:
[0050] A three-dimensional voxel dose distribution calculation subunit and a non-homogeneous correction factor calculation subunit;
[0051] The three-dimensional voxel dose distribution calculation subunit is used to perform three-dimensional voxel dose distribution calculations on first models of several sizes and second models of several sizes according to the first preset algorithm and the second preset algorithm, respectively, to obtain first calculation results corresponding to each size and second calculation results corresponding to each size;
[0052] The heterogeneity correction factor calculation subunit is used to obtain the corresponding heterogeneity correction factor through calculation according to the first calculation result corresponding to each size and the second calculation result corresponding to each size, and to construct a heterogeneity correction factor database. BRIEF DESCRIPTION OF THE DRAWINGS
[0053] Figure 1 A schematic flow chart of the steps of a three-dimensional dose reconstruction method provided by an embodiment of the present invention;
[0054] Figure 2 A schematic diagram of the flow structure of a three-dimensional dose reconstruction method provided by an embodiment of the present invention;
[0055] Figure 3 A schematic diagram of the module structure of a three-dimensional dose reconstruction system provided by an embodiment of the present invention;
[0056] Figure 4 A schematic diagram of the structure of a pelvic model matching module of a three-dimensional dose reconstruction system provided by an embodiment of the present invention;
[0057] Figure 5 A schematic diagram of the structure of a voxel model conversion unit of a three-dimensional dose reconstruction system provided by an embodiment of the present invention;
[0058] Figure 6 A schematic diagram of the structure of a dose distribution result correction module of a three-dimensional dose reconstruction system provided by an embodiment of the present invention;
[0059] Figure 7 A schematic diagram of a structure of a non-homogeneous correction factor database construction unit of a three-dimensional dose reconstruction system provided by an embodiment of the present invention;
[0060] Figure 8 A schematic diagram of the application process of a three-dimensional dose reconstruction method in afterloading therapy provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0061] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0062] At present, there is a lack of in-vivo online dose verification method flow that can be used for brachytherapy of cervical cancer in clinical practice. The present invention uses a previously proposed multi-channel in-vivo dose verification device, which can run through the entire process of cervical cancer radiotherapy, and integrates three functions: real-time monitoring of in-vivo dose, real-time tracking and positioning of radiation sources, and rapid three-dimensional dose reconstruction. In addition, a uniform water phantom is used clinically to verify the consistency of brachytherapy dose delivery. However, the human body is composed of organs and tissues of different densities, and has different X-ray absorption abilities, which will lead to inconsistencies between the actual dose distribution and the verification results. In addition, during the treatment process, it is assumed that the actual position of the radiation source is consistent with the plan, and the possible displacement of the radiation source during the actual treatment process is not considered. The present invention takes into account the heterogeneity of human tissue by measuring the position of the radiation source in real time and the pre-established HCF database, and calculates the real three-dimensional spatial dose distribution that takes into account the heterogeneity of the human body based on the actual source tracking information.
[0063] In this embodiment, DICOM (Digital Imaging and Communications Medical) is a commonly used file format in the field of medical imaging, which is used to store medical imaging data, such as CT scans, MRI, etc. HU value (Hounsfield unit) is a density unit commonly used in CT scans, which is used to measure the X-ray absorption capacity of tissues. In DICOM files, HU values are usually stored in the pixel values of CT images. CT scans pass X-rays through human tissues, measure the absorption of X-rays, and convert them into grayscale values. These grayscale values are usually associated with HU values. In DICOM files, the HU value of each pixel in the image can be obtained through specific tags or fields.
[0064] In order to better explain the three-dimensional dose reconstruction method of the present invention, this embodiment will be combined with Figure 2 and Figure 8 One example of the method of the present invention is explained. Figure 2 A schematic diagram of the flow structure of a three-dimensional dose reconstruction method provided by an embodiment of the present invention; Figure 8A schematic diagram of the application process of a three-dimensional dose reconstruction method in afterloading therapy provided by an embodiment of the present invention.
[0065] Example 1
[0066] See also Figure 1 , Figure 1 A schematic flow chart of a three-dimensional dose reconstruction method provided by an embodiment of the present invention. Figure 1 As shown, the present invention proposes a three-dimensional dose reconstruction method, including steps 101 to 104, each step is specifically as follows:
[0067] Step 101, matching a pelvic size model that meets preset requirements according to patient image data, and converting the pelvic size model that meets preset requirements into a number of voxel models of corresponding sizes;
[0068] As an example of this embodiment, historical patient pelvic image data is obtained, and the upper and lower limits of the patient's pelvic size distribution are obtained by performing size analysis on the historical patient pelvic image data;
[0069] According to the upper and lower limits of the patient's pelvic size distribution, several pelvic size models are constructed by presetting incremental intervals;
[0070] Matching the pelvic size model according to the patient's image data to obtain a pelvic size model that meets the preset requirements;
[0071] According to a first preset algorithm, a voxel model conversion is performed on the pelvic size model that meets the preset requirements to obtain a plurality of voxel models of different sizes, wherein the plurality of voxel models of different sizes include a first model and a second model.
[0072] As another example of this embodiment, clinical cervical cancer patient imaging data is summarized, and data analysis is performed based on different patient body types (such as pelvic dimensions) to determine the upper and lower limits of the patient's pelvic dimension distribution; based on the patient's pelvic dimension data distribution, a series of pelvic model sizes are determined. One specific implementation method is: using the patient's pelvic upper and lower limits as boundaries, and using increments of 1 cm or 2 cm as intervals, a series of pelvic model sizes are determined; based on the determined model sizes, the patient's corresponding imaging data (DICOM file) is obtained.
[0073] As another example of this embodiment, the corresponding patient image data is obtained according to several pelvic size models, and the corresponding pelvic size model that meets the preset requirements is converted by a first preset algorithm to obtain a corresponding voxel model; the first model is obtained by assigning the HU value of the patient image data to the voxel model with a preset physical density; wherein the first model is a non-homogeneous model; the second model is obtained by assigning the voxel model with water density; wherein the second model is a water homogeneous model. A specific implementation method is: using a Monte Carlo program (such as FLUKA or GEANT4) to convert DICOM data into a series of voxel models of different body sizes; establishing two models: a Monte Carlo non-homogeneous model and a Monte Carlo water homogeneous model. Specifically, for the non-homogeneous model, the HU value in the DICOM file is used to assign different physical densities to the voxels; for the water homogeneous model, all voxels are set to water, that is, the density is 1.0g / cm3.
[0074] Step 102, updating the planned position of the radiation source by acquiring the actual position of the radiation source according to the voxel models of the corresponding sizes, to obtain an updated actual position of the radiation source;
[0075] As an example of this embodiment, the radiation source tracking device gives the real three-dimensional position of the radiation source in the patient's body; replacing the planned position of the radiation source in the afterloading treatment plan. A specific implementation method is as follows: Step 1: Multi-channel scintillator-fiber optic detector in-body dose measurement equipment construction:
[0076] Considering the size of the source applicator, select a cylindrical inorganic or organic scintillator (such as BCF-60 and CsI crystal, etc.) and use high-grit sandpaper to polish the end face, control the height at the millimeter level, and use it as a radiation detector for dose measurement. Select a cylindrical white optical fiber of the same size (such as Mitsubishi CK-40) and polish the two end faces as a photon transmission medium, and use a silicon photomultiplier tube (SiPM) for photoelectric conversion and signal measurement. After completing the above steps, a single-channel scintillator-optical fiber-silicon photomultiplier tube measurement module is obtained, and n identical modules are processed to build a multi-channel scintillator-optical fiber detector in-body dose measurement device. Use high-permeability epoxy resin glue to couple the scintillator and the optical fiber, and use silicone grease to couple the optical fiber and the SiPM. The back-end electronics use an integral circuit to collect the SiPM output signal, and write and design the host computer software to collect and display multi-channel data;
[0077] Step 2: Determine the timing and location of the radiation source during brachytherapy by placing detectors inside and outside the body and 3D printing, as follows:
[0078] According to the patient's CT scan DICOM file, design the .stl file of the personalized source applicator 3D printing, embed the n-1 channel detector into the personalized 3D printed source applicator, determine the patient's body surface reference point position A and use lead points to mark it for easy resetting in the next treatment;
[0079] Before the patient is treated, the 3D printed applicator equipped with n-1 channel detectors is placed in the patient's cervix, and the remaining single-channel detectors are arranged at point A;
[0080] After completing the operation, perform a CT scan, obtain the position information of the lead point relative to the n-1 channel detector in the applicator according to the DICOM file, and establish a patient coordinate system with point A as the origin (0,0,0);
[0081] Step 3: Dose rate-distance calibration curve calibration acquisition:
[0082] Using the commonly used radiation sources in radiotherapy (such as Ir-192), an experimental measurement method is used to independently calibrate each channel of the in-body dose measurement device. The calibration method is as follows:
[0083] The position of the radiation source is fixed in water medium or simulated human body medium; the output dose rate data m1, m2, ..., m of each channel is measured at a distance of 1 mm within the range of 0 to 1 cm. 10 And record; measure the output dose rate data m of each channel at 1 cm intervals within the range of 1 to 10 cm 11 , m 12 ,…,m 20 and record;
[0084] Output data for any channel (m1, m2, ..., m 20 ), the dose rate-distance calibration curve f(r) is obtained by exponential fitting, where r is the distance between the scintillator and the center of the radiation source; the n-channel calibration curves f1(r), f2(r), …, f n (r);
[0085] Step 4: During the afterloading treatment, the detectors at different positions in the applicator and point A on the pelvis measure the dose rate signals D1, D2, ..., D in real time based on the data collected in real time by the host computer. n-1 , D A ;
[0086] According to the calibration curve obtained in step 2, the distances R1, R2, ..., R n-1 , R A ;
[0087] Through the multi-envelope positioning method, such as Figure 8As shown, multiple spherical envelope surfaces centered on the position of the scintillator detector can be established, and the radius thereof is the distance from the detector to the radiation source. The first n-1 data can determine the spatial position of the radiation source relative to the applicator coordinate system;
[0088] The measurement data at point A is used to verify whether the intrauterine applicator has shifted relative to the patient, and the three-dimensional spatial position of the radiation source relative to the patient is determined by the profile positioning method.
[0089] Step 103, selecting a corresponding inhomogeneity correction factor to correct the three-dimensional dose distribution calculation result of each radiation source residence point according to the voxel models of the corresponding sizes, the radiation source type at the current moment and the updated true position of the radiation source, so as to obtain the three-dimensional dose distribution correction result of each radiation source residence point;
[0090] As an example of this embodiment, through a preset algorithm and patient image data, the non-homogeneous correction factors are calculated for the first model and the second model respectively and a non-homogeneous correction factor database is constructed; according to the voxel models of the corresponding sizes, the type of radiation source at the current moment and the updated true position of the radiation source, the corresponding non-homogeneous correction factors are matched; according to the corresponding non-homogeneous correction factors, the three-dimensional dose distribution calculation results of each radiation source residence point are corrected through the GPU algorithm to obtain the three-dimensional dose distribution correction results of each radiation source residence point.
[0091] As another example of this embodiment, a Monte Carlo program (such as FLUKA or GEANT4) and the TG-43 dose formula are used to calculate the three-dimensional dose distribution in two different density models (TG-43 formula is used to calculate the dose distribution in the water homogeneous model), including the radiation source at different positions in the model and different radiation sources (i.e., different radiation energies). Through this calculation, the radiation source (such as 192 Ir) The three-dimensional voxel dose distribution in the two different density models at a certain position in the pelvic cavity. The three-dimensional voxel dose distribution in the inhomogeneous model is denoted as DOSE DICOM , the three-dimensional voxel dose distribution in the water homogeneous model is denoted as DOSE WATER . The ratio of the two is defined as the inhomogeneous correction factor HCF under the conditions of the pelvic model of that size, the radiation source energy and the radiation source position. Since it is a three-dimensional dose distribution, HCF is a three-dimensional matrix, and each HCF corresponds to only one situation: that is, the correction factor under the conditions of a pelvic model of a certain size, a certain radiation source and a radiation source position. When the pelvic model (voxel model of the corresponding size), the radiation source type and the radiation source position at the current moment are determined, the corresponding HCF matrix is selected. Furthermore, the three-dimensional dose distribution calculation results of each radiation source residence point are corrected to obtain DOSE. MEASURE =DOSETG-43 HCF, where the TG-43 formula is used to calculate the dose distribution and obtain the three-dimensional dose distribution DOSE TG-43 In addition, since the time required for matrix operations increases with the dimension in single-threaded CPU operations, in order to improve the operation efficiency, the GPU is used to perform matrix operations in this embodiment.
[0092] As another example of this embodiment, according to the first preset algorithm and the second preset algorithm, three-dimensional voxel dose distribution calculations are performed on first models of several sizes and second models of several sizes, respectively, to obtain first calculation results corresponding to each size and second calculation results corresponding to each size; according to the first calculation results corresponding to each size and the second calculation results corresponding to each size, corresponding inhomogeneous correction factors are calculated, and a database of inhomogeneous correction factors is constructed. Specifically, one possible implementation method in this embodiment is to use a Monte Carlo program (such as FLUKA or GEANT4) and a TG-43 dose formula to calculate the three-dimensional dose distribution in two different density models respectively. Repeated calculations can obtain a series of inhomogeneous correction factors HCF of pelvic models of different sizes under different radiation source energies and radiation source positions, and finally establish an HCF database.
[0093] Step 104 , obtaining a true result of the three-dimensional dose distribution by accumulating the three-dimensional dose distribution correction results of the radiation source residence points.
[0094] As an example of this embodiment, considering that the radiation source may have n residence points in the cervical cavity, the three-dimensional dose distribution calculation result is corrected for each residence point to obtain the DOSE at the corresponding residence point. MEASURE,i , the dose DOSE received by cervical cancer patients during the entire radiotherapy period is accumulated, as shown in the following formula:
[0095] DOSE=DOSE MEASURE,1 +DOSE MEASURE,2 +···+DOSE MEASURE,i +···+
[0096] DOSE MEASURE,n
[0097] Among them, DOSE MEASURE,i Represents the true three-dimensional dose distribution of the radiation source in the patient's pelvis at a certain dwelling point.
[0098] The embodiment of the present invention proposes a three-dimensional dose reconstruction method, which matches a corresponding pelvic size model according to the patient's image data, wherein the pelvic size model considers voxels and thus human body heterogeneity; then the planned position is updated by obtaining the real position of the radiation source, and the offset problem of the radiation source is considered; finally, based on the factors considering the human body heterogeneity and the radiation source offset, the three-dimensional dose is corrected by a non-homogeneous correction factor to obtain the real distribution result of the three-dimensional dose; therefore, the present invention further improves the reliability of the in-body dose measurement result by considering the human body heterogeneity and the radiation source offset problem and then using the non-homogeneous correction factor.
[0099] Example 2
[0100] See also Figure 3 , Figure 3 The following is a schematic diagram of the module structure of a three-dimensional dose reconstruction system provided by an embodiment of the present invention. Figure 3 As shown, the present invention proposes a three-dimensional dose reconstruction system, including: a pelvic model matching module 301, a radiation source position updating module 302, a dose distribution result correction module 303 and a real result acquisition module 304;
[0101] The pelvic model matching module 301 is used to match a pelvic size model that meets preset requirements according to the patient image data, and convert the pelvic size model that meets the preset requirements into a number of voxel models of corresponding sizes;
[0102] As an example of this embodiment, see Figure 4 , Figure 4 A schematic diagram of the structure of a pelvic model matching module of a three-dimensional dose reconstruction system provided by an embodiment of the present invention; Figure 4 As shown, the pelvic model matching module 301 is used to match a pelvic size model that meets preset requirements according to the patient image data, and convert the pelvic size model that meets the preset requirements into a number of voxel models of corresponding sizes, and also includes:
[0103] A pelvic size upper and lower limit acquisition unit 401, a pelvic size model acquisition unit 402, a pelvic size model matching unit 403 and a voxel model conversion unit 404;
[0104] The pelvic size upper and lower limit acquisition unit 401 is used to acquire historical patient pelvic image data, and obtain the upper and lower limits of the patient's pelvic size distribution by performing size analysis on the historical patient pelvic image data;
[0105] The pelvic cavity size model acquisition unit 402 is used to construct and obtain a plurality of pelvic cavity size models by preset increment intervals according to the upper and lower limits of the patient's pelvic cavity size distribution;
[0106] The pelvic cavity size model matching unit 403 is used to match the pelvic cavity size model according to the patient image data to obtain a pelvic cavity size model that meets the preset requirements;
[0107] The voxel model conversion unit 404 is used to perform voxel model conversion on the pelvic size model that meets the preset requirements according to a first preset algorithm to obtain a plurality of voxel models of different sizes, wherein the plurality of voxel models of different sizes include a first model and a second model.
[0108] As another example of this embodiment, see Figure 5 , Figure 5 A schematic diagram of the structure of a voxel model conversion unit of a three-dimensional dose reconstruction system provided by an embodiment of the present invention; Figure 5 As shown, the voxel model conversion unit 404 is used to perform voxel model conversion on the pelvic size model that meets the preset requirements according to a first preset algorithm to obtain a plurality of voxel models of different sizes, wherein the plurality of voxel models of different sizes include a first model and a second model, and also include:
[0109] Voxel model building subunit 501, first model building subunit 502 and second model building subunit 503;
[0110] The voxel model construction subunit 501 is used to obtain corresponding patient image data according to a plurality of pelvic dimension models, and convert the corresponding pelvic dimension model that meets the preset requirements through a first preset algorithm to obtain a corresponding voxel model;
[0111] The first model building subunit 502 is used to obtain a first model by assigning the HU value of the patient image data to the preset physical density of the voxel model; wherein the first model is a non-homogeneous model;
[0112] The second model building subunit 503 is used to obtain a second model by assigning water density to the voxel model; wherein the second model is a water homogeneous model.
[0113] The radiation source position updating module 302 is used to update the planned position of the radiation source by acquiring the real position of the radiation source according to the voxel models of the corresponding sizes, so as to obtain the updated real position of the radiation source;
[0114] The dose distribution result correction module 303 is used to select a corresponding inhomogeneity correction factor to correct the three-dimensional dose distribution calculation result of each radiation source residence point according to the voxel models of the corresponding sizes, the type of radiation source at the current moment and the updated true position of the radiation source, so as to obtain the three-dimensional dose distribution correction result of each radiation source residence point;
[0115] As an example of this embodiment, see Figure 6 , Figure 6 A schematic diagram of the structure of a dose distribution result correction module of a three-dimensional dose reconstruction system provided by an embodiment of the present invention; Figure 6 As shown, the dose distribution result correction module 303 is used to select a corresponding inhomogeneity correction factor to correct the three-dimensional dose distribution calculation result of each radiation source residence point according to the voxel models of the corresponding sizes, the radiation source type at the current moment and the updated radiation source real position, so as to obtain the three-dimensional dose distribution correction result of each radiation source residence point, and also includes:
[0116] A non-homogeneous correction factor database construction unit 601, a non-homogeneous correction factor matching unit 602 and a result correction unit 603;
[0117] The inhomogeneity correction factor database construction unit 601 is used to calculate a number of inhomogeneity correction factors and construct a inhomogeneity correction factor database through a preset algorithm and patient image data;
[0118] The inhomogeneity correction factor matching unit 602 is used to match the corresponding inhomogeneity correction factor according to the voxel models of the corresponding sizes, the type of the radiation source at the current moment and the updated real position of the radiation source;
[0119] The result correction unit 603 is used to correct the three-dimensional dose distribution calculation result of each radiation source residence point through the GPU algorithm according to the corresponding inhomogeneity correction factor to obtain the three-dimensional dose distribution correction result of each radiation source residence point.
[0120] As another example of this embodiment, see Figure 7 , Figure 7 A schematic diagram of a structure of a non-homogeneous correction factor database construction unit of a three-dimensional dose reconstruction system provided by an embodiment of the present invention; Figure 7 As shown, the inhomogeneity correction factor database construction unit 601 is used to calculate a number of inhomogeneity correction factors and construct a inhomogeneity correction factor database through a preset algorithm and patient image data, and also includes:
[0121] A three-dimensional voxel dose distribution calculation subunit 701 and a non-homogeneous correction factor calculation subunit 702;
[0122] The three-dimensional voxel dose distribution calculation subunit 701 is used to perform three-dimensional voxel dose distribution calculations on first models of several sizes and second models of several sizes according to the first preset algorithm and the second preset algorithm, respectively, to obtain first calculation results corresponding to each size and second calculation results corresponding to each size;
[0123] The non-homogeneity correction factor calculation subunit 702 is used to obtain the corresponding non-homogeneity correction factor through calculation according to the first calculation result corresponding to each size and the second calculation result corresponding to each size, and to construct a non-homogeneity correction factor database.
[0124] The real result acquisition module 304 is used to obtain the real result of the three-dimensional dose distribution by accumulating the three-dimensional dose distribution correction results of each radiation source residence point.
[0125] The embodiment of the present invention proposes a three-dimensional dose reconstruction system, which matches the corresponding pelvic size model according to the patient's image data through a pelvic model matching module, taking into account the voxels and thus the heterogeneity of the human body; the radiation source position updating module obtains the real position of the radiation source to update the planned position, taking into account the offset problem of the radiation source; and then the dose distribution result correction module corrects the three-dimensional dose through a non-homogeneous correction factor based on the factors taking into account the heterogeneity of the human body and the offset of the radiation source to obtain the real distribution result of the three-dimensional dose; therefore, the present invention further improves the reliability of the in-body dose measurement result by taking into account the heterogeneity of the human body and the offset problem of the radiation source and then using the non-homogeneous correction factor.
[0126] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.
[0127] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.
[0128] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.
Claims
1. A three-dimensional dose reconstruction method, characterized in that: include: A pelvic size model that meets preset requirements is matched according to the patient image data, and the pelvic size model that meets the preset requirements is converted into a number of voxel models of corresponding sizes; specifically: historical patient pelvic image data is obtained, and the upper and lower limits of the patient's pelvic size distribution are obtained by performing size analysis on the historical patient pelvic image data; according to the upper and lower limits of the patient's pelvic size distribution, a number of pelvic size models are constructed by preset incremental intervals; the pelvic size model is matched according to the patient image data to obtain a corresponding pelvic size model that meets the preset requirements; according to a first preset algorithm, the corresponding pelvic size model that meets the preset requirements is converted into a voxel model to obtain a number of voxel models of different sizes, wherein the number of voxel models of different sizes include a first model The method comprises the following steps: obtaining corresponding patient image data according to the plurality of pelvic size models, converting the corresponding pelvic size model that meets the preset requirements according to the first preset algorithm, and obtaining corresponding voxel models; obtaining the first model by assigning the HU value of the patient image data to the voxel model with a preset physical density; wherein the first model is a non-homogeneous model; obtaining the second model by assigning the voxel model with water density; wherein the second model is a water homogeneous model; wherein the first preset algorithm is a Monte Carlo program; According to the voxel models of the corresponding sizes, the planned position of the radiation source is updated by acquiring the real position of the radiation source to obtain an updated real position of the radiation source; According to the voxel models of several corresponding sizes, the type of radiation source at the current moment and the updated true position of the radiation source, a corresponding inhomogeneous correction factor is selected to correct the three-dimensional dose distribution calculation result of each radiation source residence point, so as to obtain the three-dimensional dose distribution correction result of each radiation source residence point; specifically, by using a preset algorithm and patient image data, the inhomogeneous correction factor is calculated for the first model and the second model respectively and a inhomogeneous correction factor database is constructed; according to the voxel models of several corresponding sizes, the type of radiation source at the current moment and the updated true position of the radiation source, the corresponding inhomogeneous correction factor is matched; according to the corresponding inhomogeneous correction factor, the three-dimensional dose distribution calculation result of each radiation source residence point is corrected by a GPU algorithm, so as to obtain the three-dimensional dose distribution correction result of each radiation source residence point. The three-dimensional dose distribution correction result of the point; the preset algorithm and the patient image data are used to calculate the inhomogeneous correction factor for the first model and the second model respectively and construct a inhomogeneous correction factor database, specifically: according to the first preset algorithm and the second preset algorithm, the three-dimensional voxel dose distribution calculation is performed on the first model of several sizes and the second model of several sizes respectively, and the first calculation result corresponding to each size and the second calculation result corresponding to each size are obtained; according to the first calculation result corresponding to each size and the second calculation result corresponding to each size, the corresponding inhomogeneous correction factor is obtained by calculation, and a inhomogeneous correction factor database is constructed; wherein, the second preset algorithm is: TG-43 dose formula, and the inhomogeneous correction factor database is HCF database; The true result of the three-dimensional dose distribution is obtained by accumulating the three-dimensional dose distribution correction results of each radiation source residence point.
2. A three-dimensional dose reconstruction system, characterized in that: include: Pelvic model matching module, radiation source position updating module, dose distribution result correction module and real result acquisition module; The pelvic model matching module is used to match a pelvic size model that meets the preset requirements according to the patient image data, and convert the pelvic size model that meets the preset requirements into a number of voxel models of corresponding sizes; it also includes: a pelvic size upper and lower limit acquisition unit, a pelvic size model acquisition unit, a pelvic size model matching unit and a voxel model conversion unit; the pelvic size upper and lower limit acquisition unit is used to acquire historical patient pelvic image data, and obtain the upper and lower limits of the patient's pelvic size distribution by performing size analysis on the historical patient pelvic image data; the pelvic size model acquisition unit is used to construct a number of pelvic size models according to the upper and lower limits of the patient's pelvic size distribution through a preset incremental interval; the pelvic size model matching unit is used to match the pelvic size model according to the patient image data to obtain a corresponding pelvic size model that meets the preset requirements; the voxel model conversion unit is used to perform voxel model conversion on the corresponding pelvic size model that meets the preset requirements according to a first preset algorithm to obtain a number of voxel models of different sizes, wherein the The voxel models of different sizes include a first model and a second model; the voxel model conversion unit is used to perform voxel model conversion on the corresponding pelvic size model that meets the preset requirements according to a first preset algorithm to obtain a number of voxel models of different sizes, wherein the several voxel models of different sizes include the first model and the second model, and also include: a voxel model construction subunit, a first model construction subunit and a second model construction subunit; the voxel model construction subunit is used to obtain corresponding patient image data according to several pelvic size models, and convert the corresponding pelvic size model that meets the preset requirements according to the first preset algorithm to obtain the corresponding voxel model; the first model construction subunit is used to obtain the first model by assigning the HU value of the patient image data to the voxel model with a preset physical density; wherein the first model is a non-homogeneous model; the second model construction subunit is used to obtain the second model by assigning water density to the voxel model; wherein the second model is a water homogeneous model; wherein the first preset algorithm is a Monte Carlo program; The radiation source position updating module is used to update the planned position of the radiation source by acquiring the real position of the radiation source according to the voxel models of the corresponding sizes, so as to obtain the updated real position of the radiation source; The dose distribution result correction module is used to select a corresponding inhomogeneous correction factor to correct the three-dimensional dose distribution calculation result of each radiation source residence point according to the voxel models of the corresponding sizes, the type of the radiation source at the current moment and the updated true position of the radiation source, so as to obtain the three-dimensional dose distribution correction result of each radiation source residence point; it also includes: a inhomogeneous correction factor database construction unit, a inhomogeneous correction factor matching unit and a result correction unit; the inhomogeneous correction factor database construction unit is used to calculate and obtain a number of inhomogeneous correction factors and construct a inhomogeneous correction factor database through a preset algorithm and patient image data; the inhomogeneous correction factor matching unit is used to match the corresponding inhomogeneous correction factor according to the voxel models of the corresponding sizes, the type of the radiation source at the current moment and the updated true position of the radiation source; the result correction unit is used to correct the three-dimensional dose distribution calculation result of each radiation source residence point through a GPU algorithm according to the corresponding inhomogeneous correction factor, The three-dimensional dose distribution correction result of each radiation source residence point is obtained; the inhomogeneous correction factor database construction unit is used to calculate a number of inhomogeneous correction factors and construct a inhomogeneous correction factor database through a preset algorithm and patient image data, and also includes: a three-dimensional voxel dose distribution calculation subunit and a inhomogeneous correction factor calculation subunit; the three-dimensional voxel dose distribution calculation subunit is used to perform three-dimensional voxel dose distribution calculation on a first model of a number of sizes and a second model of a number of sizes according to the first preset algorithm and the second preset algorithm, respectively, to obtain a first calculation result corresponding to each size and a second calculation result corresponding to each size; the inhomogeneous correction factor calculation subunit is used to obtain the corresponding inhomogeneous correction factor by calculation according to the first calculation result corresponding to each size and the second calculation result corresponding to each size, and construct a inhomogeneous correction factor database; wherein the second preset algorithm is: TG-43 dose formula, and the inhomogeneous correction factor database is an HCF database; The real result acquisition module is used to obtain the real result of the three-dimensional dose distribution by accumulating the three-dimensional dose distribution correction results of each radiation source residence point.
Citation Information
Patent Citations
Dose verification die body model manufacturing method, device thereof and system
CN110975174A
Radiotherapy dose prediction method and device
CN113599728A
Three-dimensional in-vivo dose reconstruction method and device for heavy ion radiotherapy
CN115957453A
After-loading three-dimensional dose calculation method based on Monte Carlo, medium and terminal
CN117839101A
