Simulation method and device for radioactive particle source distribution in lumen

Abstract

Disclosed are a simulation method and device for radioactive particle source distribution in a lumen. The method comprises: marking a target area and a lumen target in an image space, and establishing a three-dimensional model containing the target area and the lumen target in a three-dimensional space; selecting the length and diameter of a stent on the basis of a luminal centerline of a narrowed section of the lumen target caused by the treatment target area, and establishing a stent model in the three-dimensional space; simulating implantation of the stent into the lumen, expanding the target area and the lumen, acquiring a stent-lumen-target area model, and extracting simulation results of the stent-lumen-target area model after stabilization of the stent-lumen-target area model; calculating a relationship matrix between the image space and the three-dimensional space, and inversely mapping the simulation results of the stent-lumen-target area to the image to obtain a simulated postoperative CT; rasterizing an expanded surface of the stent to obtain candidate particle coordinates; and establishing an objective function on the basis of the postoperative target area and the candidate particle coordinates, and performing plan optimization for dose and number of particles using a heuristic method.

Description

A method and device for simulating the distribution of radioactive particles in a cavity

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to Chinese application No. 2023115283407, filed on November 16, 2023, the entire contents of which are incorporated herein by reference. Technical Field

[0003] The present application relates to the field of simulation technology, and in particular to a method and device for simulating the distribution of radioactive particles in a cavity. Background Art

[0004] Seed implantation therapy is an important brachytherapy technique that can be used to treat targeted areas throughout the body. The treatment principle involves using a puncture needle to implant multiple radioactive seeds into the target area, where the gamma rays released by the seeds kill tumor cells. For the treatment of cavitary tumors, seed stents offer a "combined effect" of dual therapeutic benefits. On the one hand, the stent alleviates obstruction caused by the tumor; on the other hand, it carries the radioactive seeds, providing continuous brachytherapy to the target tumor area.

[0005] However, there is currently a lack of research on the integrated planning and design of particles and stents. When metal stents are placed to expand the cavity, the tumor and the cavity will deform accordingly. The current deployment of particle sources does not specifically and quantitatively consider the shape of the metal stent itself after placement and the deformation caused by the metal stent placement, resulting in dose field misalignment, affecting the treatment effect and causing serious complications.

[0006] Summary of the Invention

[0007] In view of this, the present application proposes a method and device for simulating the distribution of radioactive particles in a cavity. The simulation results provide a reference for clinical physicians to alleviate the shortcomings of the existing technology.

[0008] In a first aspect, the present application provides a method for simulating the distribution of radioactive particles in a cavity, comprising: ct Mark the treatment target area and cavity target in the three-dimensional model space O m Establish a three-dimensional model including the treatment target area and the lumen target area; select the stent length and diameter based on the lumen centerline of the lumen target stenosis segment caused by the treatment target area, and m Establish a stent model; simulate the stent implantation cavity target, expand the treatment target area and cavity target, obtain the stent-cavity-target area model, and extract the stent-cavity-target area model simulation results after the stent-cavity-target area model is stable; calculate the medical image space O ct and 3D model space O mBased on the relationship matrix T, the stent, lumen and target area in the simulation results of the stent-lumen-target model are reversely outlined to the medical image and the CT values ​​are reassigned to obtain the postoperative simulated CT image. The stent unfolded surface is rasterized based on the simulation results of the stent-lumen-target model to obtain the coordinates of the candidate particle positions. Based on the target area and candidate particle positions outlined on the postoperative simulated CT image, a heuristic optimization method is used to optimize the dose and particle number plan.

[0009] Optionally, the medical imaging space O ct Refers to the DICOM 3D image coordinate system, 3D model space O m Refers to the three-dimensional model coordinate system defined by mechanical simulation software.

[0010] Optionally, the method further includes: measuring the stenosis length using the lumen centerline of the target stenosis segment of the lumen, wherein the stent length is 5 mm longer at both ends of the stenosis, and selecting the stent diameter using the following formula, where the length unit is mm:

[0011] Where D is the estimated normal diameter of the target stenosis segment of the lumen, d is the narrowest diameter of the stenosis segment, T c is the length of the centerline of the target stenosis segment of the lumen, t is the straight-line distance between the two ends of the centerline of the target stenosis segment of the lumen, D stent is the decision variable for selecting the stent diameter; when D stent ≥50%, the stent diameter is selected as D+2, when D stent <50%, the stent diameter is selected as D+1.

[0012] Optionally, it also includes: deleting voxels in the medical image used to construct a three-dimensional model of the treatment target area and the cavity target, and reversely outlining the stent, cavity and target area in the stent-cavity-target area model simulation results to the medical image through the relationship matrix T, assigning the CT values ​​of the cavity and target area in the reverse outlined area according to the original CT values, assigning the CT value of the stent area according to the actual postoperative CT value, assigning the CT value of the stent area according to the stent CT value in the actual postoperative CT, assigning the CT value of the overlapping part of the reverse outlined area and the medical image according to the newly outlined area, and assigning the CT value of the vacant part according to other surrounding close tissues.

[0013] Optionally, the method further comprises: dividing the unfolded surface of the support into grids at preset intervals according to Cartesian coordinates, wherein the coordinates of the grid points represent positions of candidate particles.

[0014] Optionally, the optimization objective function is related to the dose and the number of particles, and its form is based on the following formula:

[0015] Where P is the arrangement of particles on the expanded support, d tv,j(P) represents the dose received by the jth voxel in the treatment target volume (TV); is the lower bound of the dose to the treatment target. N(P) is the number of particles used; H(·) is the step function; and w n are the weights of the dose constraint term and the particle number constraint term, respectively.

[0016] Optionally, the method further includes: initializing the algorithm starting point, setting the shortest distance d to the treatment target area TH All candidate particle positions are arranged to initialize the particle arrangement P0. Set the iteration variable C and the iteration variable lower limit C min , process decay rate q and number of iterations N and other parameters, calculate the initial target value f(P0); change the current state, specifically, discard or add a deployed or undeployed particle with a probability of 50%, and obtain a new particle arrangement P i ; Calculate the target value f(P i ); If the target value obtained by the new particle arrangement is less than that obtained in the previous step, that is, f(P i ) <f(P i-1 ), accept the new particle arrangement P i Otherwise, with a certain probability Accept the new particle arrangement, where tanh is the hyperbolic tangent function; linearly reduce the iteration variable value, Q i+1 =Q i -q; Repeat the particle arrangement modification and iteration variable reduction operation several times, or the iteration variable reaches the lower limit Q min When , the iterative optimization operation is stopped and the optimized particle arrangement P is obtained. * .

[0017] Optionally, the prefabricated spacing between adjacent grid points is set to 10 mm.

[0018] Optionally, and w n Based on experience, the values ​​are set to 1 and 10.

[0019] Optionally, set the closest distance d TH is 3mm, iteration variable C=1000, C min =1,q=1,N=1000.

[0020] In a second aspect, the present application provides a radioactive particle source simulation device, comprising: a stent implantation simulation module, which is used to simulate the radioactive particle source in a medical imaging space. ct Mark the treatment target area and cavity target in the three-dimensional model space O mEstablish a three-dimensional model including the treatment target area and the lumen target area; select the stent length and diameter based on the lumen centerline of the lumen target stenosis segment caused by the treatment target area, and m Establish a stent model; simulate the stent implantation cavity target, expand the treatment target area and cavity target, obtain the stent-cavity-target area model, and extract the stent-cavity-target area model simulation results after the stent-cavity-target area model is stable; calculate the medical image space O ct and 3D model space O m Based on the relationship matrix T, the stent, cavity and target area in the simulation results of the stent-cavity-target area model are reversely outlined to the medical image and the CT values ​​are reassigned to obtain a postoperative simulated CT image; a particle source distribution module is used to rasterize the stent expansion surface based on the simulation results of the stent-cavity-target area model to obtain the coordinates of the candidate particle positions; based on the target area and candidate particle positions outlined by the postoperative CT, a heuristic optimization method is used to optimize the dose and particle number plan.

[0021] The beneficial effects of this application are as follows:

[0022] The technical solution provided by the present application may include the following beneficial effects: a method and device for simulating the placement of radioactive particle sources in a cavity are proposed, which simulates the placement of a metal stent into the cavity through finite element mechanics simulation, obtains the shape of the metal stent itself after placement, and the shape of the target area and the cavity after placement of the metal stent, and arranges the radioactive particle source according to the shape of the expanded stent, target area and cavity to ensure that the particle source dose field and the target area are precisely conformal to each other, thereby realizing automated simulation of the placement of radioactive particle sources in the cavity, providing auxiliary reference for clinical physicians, and alleviating the technical problems of low conformality of the dose field obtained by traditional stent particle placement technology and high dependence on physician experience.

[0023] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific implementation methods or the description of the prior art. Obviously, the drawings described below are an implementation method of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0025] FIG1 is a schematic flow chart of a method for simulating the placement of radioactive particles in a cavity according to the first embodiment of the present application;

[0026] FIG2 is a schematic structural diagram of a radioactive particle source simulation device in a cavity according to the second embodiment of the present application. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions of the present application will be clearly and completely described below in conjunction with the accompanying drawings. The described embodiments are only part of the embodiments of the present application, not all of the embodiments.

[0028] First embodiment:

[0029] FIG1 is a flow chart of a method for simulating the placement of radioactive particles in a cavity according to the first embodiment of the present application. As shown in FIG1 , the method includes steps S1 , S2 , S3 , S4 and S5 .

[0030] Step S1: In the medical image space O ct Mark the treatment target area and cavity target in the three-dimensional model space O m Establishing a three-dimensional model including the treatment target area and the cavity target area;

[0031] It should be noted that the medical imaging space O ct Refers to the DICOM 3D image coordinate system, 3D model space O m Refers to the three-dimensional model coordinate system defined by mechanical simulation software.

[0032] Optionally, detection can be performed based on the time-domain or frequency-domain characteristics of the treatment target and cavity target regions, i.e., by color and frequency characteristics. Alternatively, the treatment target and cavity target regions in the medical image can be manually marked by delineation. For example, CT medical images can be constructed into a three-dimensional model using multi-slice reconstruction methods, converting a two-dimensional CT image sequence into a three-dimensional model.

[0033] Step S2: Select the stent length and diameter based on the lumen centerline of the target stenosis segment of the lumen caused by the treatment target area, and m Establish a stent model. Simulate stent implantation into the lumen target, expand the treatment target volume and lumen target, and obtain a stent-lumen-target model. Once the stent-lumen-target model is stable, extract the simulation results of the stent-lumen-target model.

[0034] For example, the stent length and diameter are selected based on the lumen centerline of the target stenosis segment of the lumen caused by the treatment target volume. The stenosis length is measured using the lumen centerline, and the stent length is 5 mm longer at both ends of the stenosis. The stent diameter is selected using the following formula, with the length unit being mm:

[0035] Where D is the estimated normal diameter of the target stenosis segment of the lumen, d is the narrowest diameter of the stenosis segment, T cis the length of the centerline of the target stenosis segment of the lumen, t is the straight-line distance between the two ends of the centerline of the target stenosis segment of the lumen, D stent is the decision variable for selecting the stent diameter. stent ≥50%, the stent diameter is selected as D+2, when D stent <50%, the stent diameter is selected as D+1.

[0036] Optionally, the stent is a self-expanding stent made of nickel-titanium alloy, and the stent model is a commercial stent SMART Control mesh stent.

[0037] Exemplarily, the simulation result is an OBJ file.

[0038] Alternatively, simulations were performed using the finite element software Abaqus.

[0039] Step S3: Calculate the medical image space O ct and 3D model space O m Based on the relationship matrix T, the stent, lumen and target area in the simulation results of the stent-lumen-target model are reversely outlined to the medical image and the CT values ​​are reassigned to obtain the postoperative simulated CT image.

[0040] For example, the origins of the medical image space and the three-dimensional model space are aligned to construct a relationship matrix T between the medical image and the three-dimensional model surface. The relationship matrix T can be used to extract the three-dimensional model outline and map it onto the two-dimensional slice of the medical image.

[0041] Step S4: Rasterize the expanded surface of the bracket to obtain the coordinates of the candidate particle positions. Specifically, based on the bracket's bottom radius and height, a rectangular surface is expanded. The candidate particle positions are rasterized based on the particle placement interval.

[0042] In an optional embodiment, the interval between adjacent grid points is set to 10 mm.

[0043] In an optional embodiment, the stent deployment surface is divided into grids at preset intervals according to Cartesian coordinates, and the grid point coordinates represent candidate particle positions.

[0044] Step S5: Optimize the stent seed source surgical plan. Specifically, establish an objective function and perform a heuristic optimization operation on the candidate particle positions to optimize the dose distribution and number of radiation particles.

[0045] Optionally, the optimization objective function is related to the dose and the number of particles, and its form is based on the following formula:

[0046] Where P is the arrangement of particles on the expanded support, d tv,j(P) represents the dose received by the jth voxel in the treatment target volume (TV). is the lower bound of the dose to the treatment target. N(P) is the number of particles used. H(·) is the step function. and w n are the weights of the dose constraint term and the particle number constraint term, respectively.

[0047] In a specific embodiment, a heuristic optimization method is used to optimize the dose and particle number plan, including: initializing the algorithm starting point, setting the closest distance d to the treatment target area to TH All candidate particle positions are arranged to initialize the particle arrangement P0. Set the iteration variable C and the iteration variable lower limit C min , process decay rate q and number of iterations N and other parameters, calculate the initial target value f(P0); change the current state, specifically, discard or add a deployed or undeployed particle with a probability of 50%, and obtain a new particle arrangement P i . Calculate the target value f(P i ). If the target value obtained by the new particle arrangement is less than that obtained in the previous step, that is, f(P i ) <f(P i-1 ), accept the new particle arrangement P i Otherwise, with a certain probability Accept the new particle arrangement, where tanh is the hyperbolic tangent function; linearly reduce the iterative variable value, Q i+1 =Q i -q. Repeat the particle arrangement modification and iteration variable reduction operation several times, or the iteration variable reaches the lower limit Q min When , the iterative optimization operation is stopped and the optimized particle arrangement P is obtained. * .

[0048] Second embodiment:

[0049] FIG2 is a schematic structural diagram of an intraluminal radioactive particle source simulation device according to an embodiment of the present application. As shown in FIG2 , the intraluminal radioactive particle source simulation device 200 includes a stent implantation simulation module 201 and a particle source module 202 .

[0050] Stent implantation simulation module 201 is used to simulate the implantation of a stent in a medical imaging space. ct Mark the treatment target area and cavity target in the three-dimensional model space O m Establish a three-dimensional model including the treatment target area and the lumen target area; select the stent length and diameter based on the lumen centerline of the lumen target stenosis segment caused by the treatment target area, and mEstablish a stent model; simulate the stent implantation cavity target, expand the treatment target area and cavity target, obtain the stent-cavity-target area model, and extract the stent-cavity-target area model simulation results after the stent-cavity-target area model is stable; calculate the medical image space O ct and 3D model space O m Based on the relationship matrix T, the stent, lumen and target area in the simulation results of the stent-lumen-target model are reversely outlined to the medical image and the CT values ​​are reassigned to obtain the postoperative simulated CT image.

[0051] The particle source module 202 is used to rasterize the stent expansion surface based on the postoperative stent-cavity-target model to obtain the candidate particle position coordinates; based on the target area and candidate particle positions outlined by the postoperative simulated CT image, the heuristic optimization method is used to optimize the dose and particle number plan.

[0052] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The scope of protection of the present application is not limited thereto. Although the present application has been described in detail with reference to the above-described embodiments, those skilled in the art should understand that any person skilled in the art can modify or easily conceive of changes to the technical solutions described in the above-described embodiments within the technical scope disclosed in the present application, or perform equivalent replacements for some of the technical features thereof. These modifications, changes, or replacements do not deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application shall be subject to the scope of protection of the claims.

Claims

1. A method for simulating the distribution of radioactive particles in a cavity target, characterized in that: include: In the medical imaging space ct The treatment target area and cavity target are marked in the 3D model space O m Establishing a three-dimensional model including the treatment target area and the cavity target area; The length and diameter of the stent are selected based on the lumen centerline of the target stenosis segment of the lumen caused by the treatment target area, and the stent is placed in the three-dimensional model space O m Establishing the scaffold model; Simulate the stent implantation cavity target, expand the treatment target area and cavity target, and obtain the stent-cavity-target area model; after the stent-cavity-target area model is stable, extract the simulation results of the stent-cavity-target area model; Computational Medical Image Space ct and the three-dimensional model space O m Based on the relationship matrix T, the stent, cavity and target area in the simulation result of the stent-cavity-target area model are reversely outlined to the medical image and the CT values ​​are reassigned to obtain a postoperative simulated CT image; Based on the simulation results of the stent-cavity-target model, the unfolded surface of the cylindrical stent is rasterized to obtain the position coordinates of the candidate particles; Based on the target area and candidate particle positions delineated by postoperative simulated CT images, an optimization objective function was established and a heuristic optimization method was used to optimize the dose and particle number plan.

2. The method according to claim 1, characterized in that The medical imaging space O ct Refers to the DICOM 3D image coordinate system, 3D model space O m Refers to the three-dimensional model coordinate system defined by the mechanical simulation software.

3. The method according to claim 1, characterized in that The length and diameter of the stent are selected based on the lumen centerline of the target stenosis segment of the lumen caused by the treatment target area, including: using the lumen centerline of the target stenosis segment of the lumen to measure the stenosis length, the stent length is 5 mm longer at both ends of the stenosis, and the stent diameter is selected using the following formula, and the length unit is mm: Where D is the estimated normal diameter of the target stenosis segment of the lumen, d is the narrowest diameter of the stenosis segment, and T c is the length of the centerline of the target stenosis segment of the lumen caused by the treatment target area, t is the straight-line distance between the two ends of the centerline of the target stenosis segment of the lumen caused by the treatment target area, and D stent is the decision variable for selecting the stent diameter; when D stent ≥50%, the stent diameter is selected as D+2, when D stent <50%, the stent diameter is selected as D+1.

4. The method according to claim 1, characterized in that: The computational medical image space O ct and the three-dimensional model space O m The method comprises: deleting voxels used to construct the three-dimensional model of the treatment target area and the cavity target in the medical image, and reversely outlining the stent, cavity and target area in the simulation result of the stent-cavity-target area model to the medical image through the relationship matrix T, and assigning CT values ​​to the stent, cavity and target area in the simulation result of the stent-cavity-target area model according to the original CT values, assigning CT values ​​of the cavity and target area in the reverse outlined area according to the stent CT value in the real postoperative CT, assigning CT values ​​of the stent area according to the stent CT value in the real postoperative CT, assigning CT values ​​of the overlapping parts of the reverse outlined area and the medical image according to the newly outlined area, and assigning CT values ​​of the missing parts according to other surrounding close tissues.

5. The method according to claim 1, characterized in that The method of rasterizing the unfolded surface of the cylindrical stent based on the postoperative stent, cavity and target area model to obtain the candidate particle position coordinates also includes: dividing the unfolded surface of the stent into grids at preset intervals according to Cartesian coordinates, and the grid point coordinates represent the candidate particle positions.

6. The method according to claim 1, characterized in that The establishing of the optimization objective function includes: the optimization objective function is related to the dose and the number of particles, and its form is based on the following formula: Where P is the arrangement of particles on the unfolded support, d tv,j (P) represents the jth individual in the treatment target area The dose of hormone received; is the lower limit of the dose in the treatment target area; N(P) is the number of particles used; H(·) is the step function; and w n are the weights of the dose constraint and the particle number constraint, respectively.

7. The method according to claim 1, characterized in that The dose and particle number planning optimization using a heuristic optimization method includes: initializing the algorithm starting point, setting the shortest distance d to the treatment target area TH All candidate particle positions are arranged to initialize the particle arrangement P0; set the iteration variable C and the iteration variable lower limit C min , process decay rate q and number of iterations N, calculate the initial target value f(P0); change the current state, discard or add a deployed or undeployed particle with a probability of 50%, and obtain a new particle arrangement P i ; Calculate the target value f(P i ); if f(P i ) <f(P i-1 ), accept the new particle arrangement P i Otherwise, with a certain probability Accept the new particle arrangement, where tanh is the hyperbolic tangent function; linearly reduce the iteration variable value, Q i+1 =Q i -q; Repeat the particle arrangement modification and iteration variable reduction operation several times, or the iteration variable reaches the lower limit Q min When , the iterative optimization operation is stopped and the optimized particle arrangement P is obtained. * .

8. The method according to claim 5, characterized in that The prefabrication spacing between adjacent grid points is set to 10 mm.

9. The method according to claim 6, characterized in that is 1, w n is 10.

10. The method according to claim 7, characterized in that Set the minimum distance d TH is 3mm, iteration variable C = 1000, C min =1, q=1, N=1000.

11. A device for simulating the distribution of radioactive particles in a cavity, characterized in that: include: Stent implantation simulation module, which is used in medical imaging space ct The treatment target area and cavity target are marked in the 3D model space O m Establish a three-dimensional model including the treatment target area and the cavity target area; select the stent length and diameter based on the cavity centerline of the cavity target stenosis segment caused by the treatment target area, and m Establish a stent model; simulate the stent implantation cavity target, expand the treatment target area and cavity target, obtain the stent-cavity-target area model, and extract the simulation results of the stent-cavity-target area model after the stent-cavity-target area model is stable; calculate the medical image space O ct and the three-dimensional model space O m Based on the relationship matrix T, the stent, cavity and target area in the simulation result of the stent-cavity-target area model are reversely outlined to the medical image and the CT values ​​are reassigned to obtain a postoperative simulated CT image; The particle source distribution module is used to rasterize the unfolded surface of the cylindrical stent based on the simulation results of the stent-cavity-target model to obtain the coordinates of the candidate particle positions; based on the target area and candidate particle positions outlined by the post-operative simulated CT image, a heuristic optimization method is used to optimize the particle dose and particle number plan.

Images