Three-dimensional lattice radiotherapy lattice arrangement method and system

Abstract

The application provides a three-dimensional lattice radiotherapy lattice arrangement method and system, comprising the following steps: obtaining a computed tomography (CT) image and a structure sequence. The application realizes a series of operations from image acquisition to lattice arrangement scheme planning through an automatic algorithm process, saves a large amount of time and effort, can quickly complete lattice arrangement, is suitable for radiotherapy plan making of large-volume tumors, greatly improves the design efficiency of the radiotherapy plan, and guarantees the consistency of lattice arrangement under different patients or different doctor operations, improves the quality stability of the radiotherapy plan, through targeted processing of the structure binary image sequence of GTV and OAR_PRV, and in combination with reasonable lattice small ball arrangement conditions, the lattice position can be more accurately planned, the damage to the organs at risk is reduced, the three-dimensional dose distribution is improved, the tumor treatment effect is improved, and the complication risk is reduced.

Description

Technical Field

[0001] This invention relates to the field of radiotherapy technology, and in particular to a method and system for arranging crystal lattices in three-dimensional lattice radiotherapy. Background Technology

[0002] External beam radiation therapy (EMB) has always aimed to deliver a uniform dose of radiation to tumors. However, despite progress in irradiation precision, organ-hazardous toxicity remains a key factor limiting the efficacy of most EMB therapies. For larger tumors, treatment outcomes are often poor due to difficulties in surgery, ineffective chemotherapy and interventional therapies, and resistance to external beam radiation.

[0003] Spatially Fractionated Radiation Therapy (SFRT) breaks with the traditional theory of uniform dose distribution within the target area, employing a highly non-uniform dose distribution. It has shown significant effectiveness in treating large-volume and refractory tumors resistant to conventional fractionated radiotherapy. Its potential mechanisms of action, besides the direct / indirect cell-killing effects of peak-dose radiation, involve multiple aspects such as vascular damage, systemic immune responses, and tumor microenvironment regulation; however, the specific mechanisms are not yet fully understood. Several hypotheses currently explain the strong killing effect on tumor tissue in the trough-dose zone of spatial fractionation, such as the bystander effect (where nearby unirradiated cells suffer radiation damage or death due to nearby irradiated / dead cells); the population effect (the interaction between peak-dose and trough-dose zones within a millimeter-scale region of the target area); the diffusion of reactive oxygen species (such as H2O2) between peak and trough dose zones; the damaging effect of peak dose on blood vessels; and the killing of distant lesions induced by tumor immune-related lymphatic infiltration and antigen release (the remote effect). Meanwhile, the activation of specific signal transduction pathways and the migration and proliferation of stem cells in peak-valley dose intervals may be related to the protective effect of spatial segmentation on normal tissues.

[0004] Lattice radiotherapy (LRT), as a crucial technique for achieving single-phase radiotherapy (SFRT), relies heavily on the lattice parameters—including diameter, spacing, and position—that significantly influence three-dimensional dose distribution. An ideal lattice spatial distribution can effectively improve patient outcomes. However, certain challenges remain in current clinical practice:

[0005] 1. The location of the lattice is usually drawn manually by the doctor. In order to obtain a more ideal three-dimensional dose distribution, it is often necessary to draw it multiple times to adjust the lattice parameters. This method is not only inefficient, but also the quality of the plan depends to a large extent on the doctor's drawing, which is highly subjective.

[0006] Second, existing radiotherapy planning systems do not consider the special requirements of SFRT planning. Professionals can only use conventional radiotherapy planning systems and simple geometric tools such as distance measurement, checkerboard display, and 3D display to manually place spheres as a lattice. When the tumor volume is small, this method is barely feasible, but when the tumor volume is large, the workload becomes unbearable, and it is impossible to guarantee that the placement of the spheres is optimal. This method has problems such as low efficiency, large workload, strong subjectivity, and poor consistency.

[0007] Therefore, a method and system for lattice arrangement in three-dimensional lattice radiotherapy are proposed. Summary of the Invention

[0008] In view of this, embodiments of the present invention aim to provide a method and system for arranging lattice for three-dimensional lattice radiotherapy to solve or alleviate the technical problems existing in the prior art, and at least provide a beneficial alternative.

[0009] To solve the above-mentioned technical problems, this application adopts a technical solution as follows: a three-dimensional lattice radiotherapy lattice arrangement method, comprising the following steps:

[0010] Step 1: Acquire a sequence of computed tomography (CT) images, which consists of multiple CT image planes arranged in spatial order and parallel to each other;

[0011] Step 2: Load contour lines for each structure based on the CT image plane. Each structure contains one or more contour lines. The structure includes human organs or tumor structures.

[0012] Step 3: A grid-based filling algorithm is used to generate a sequence of binary images of each structure, which is the same size as the CT image sequence.

[0013] Step 4: Process the structural binary image sequences of the tumor target volume (GTV) and the planned organ at risk (OAR_PRV) based on the erosion algorithm and the dilation algorithm respectively, and generate the processed image region;

[0014] Step 5: Determine the location of the lattice sphere center by layering and scanning the processed image region;

[0015] Step 6: Based on the determined lattice center position and lattice sphere arrangement conditions, plan the lattice arrangement scheme for three-dimensional lattice radiotherapy.

[0016] As a further preferred embodiment of this technical solution, in step four, the method for generating the processed image region includes the following steps:

[0017] Step 401: Based on the erosion algorithm, perform three-dimensional shrinking (m+d / 2) millimeters on the structural binary image sequence of GTV to obtain the shrunken image sequence V.L Where m is the distance from GTV to V L The inward distance, where d is the diameter of the lattice microsphere;

[0018] Step 402: Perform two-dimensional external expansion (r) on all OAR_PRV structural binary image sequences based on the dilation algorithm. P Processing the data by +d / 2) millimeters and performing a merging operation yields the merged image sequence V. PRV , where r P The distance between the lattice sphere and OAR_PRV;

[0019] Step 403, find V PRV Voxels with a median value of 1, and V L Middle and V PRV and V L The voxel values ​​in the intersection are set to 0.

[0020] As a further preferred embodiment of this technical solution, in step five, the method for determining the position of the lattice sphere center includes the following steps:

[0021] Step 501, find V L The middle layer S0 of the foreground region is scanned in the order of x, y, z, and foreground voxels are added to the candidate list L. possible middle;

[0022] Step 502, retrieve list L possible The first voxel point in the matrix is ​​added to the lattice sphere center list L0, and from L... possible Remove from;

[0023] Step 503: Remove candidate list L possible All voxel points in the list L0 whose distance from the last sphere center is less than r millimeters, where r is the minimum distance between the sphere centers;

[0024] Step 504: Repeat steps 502-503 above until list L is reached. possible Empty;

[0025] Step 505: Using the intermediate layer S0 as the center, find layers with a spacing of s millimeters in both the top and bottom directions, i.e., layers that satisfy S... i+1 With S i The distance between the two layers is s millimeters;

[0026] Step 506: Arrange S in the order of x, y, z. i Layer image scanning will be compared with the previous layer's lattice sphere center list L i-1 (when i>0) or L i+1 Foreground voxel points whose center-to-sphere distance is not less than d1 mm (when i < 0) are added to the candidate list L.possible In the diagram, d1 represents the two-dimensional distance between the center of the lattice microspheres in each layer and the center of the microspheres in the adjacent layer.

[0027] Step 507, retrieve list L possible The first voxel point in the array is added to the lattice sphere center list L. i In, and from L possible Remove from;

[0028] Step 508: Remove candidate list L possible List of all lattice sphere centers L i The last voxel point with a distance of less than r millimeters between its centers;

[0029] Step 509: Repeat steps 507-508 above until list L is reached. possible Empty, and process all layers, finally from the list L of lattice centers of each layer. i The positions of the centers of all crystal lattices are obtained from this.

[0030] As a further preferred embodiment of this technical solution, the arrangement conditions of the lattice microspheres include:

[0031] The diameter of the lattice microspheres is d millimeters;

[0032] GTV to V L The inward distance is m millimeters;

[0033] The minimum distance between the centers of the lattice spheres is r millimeters;

[0034] The distance between the lattice sphere and OAR_PRV is not less than r P millimeters;

[0035] The interlayer spacing is s millimeters;

[0036] The two-dimensional distance between the center of each lattice sphere and the center of the lattice sphere in the adjacent layer is no less than d1 mm. Since the center position of the first sphere may not be optimal, after the initial sphere arrangement is completed, the spheres are moved in three dimensions relative to the center of the first sphere by a fixed step size Δl (2.5 mm), where the movement range of x, y, and z is ±d1 / 2. After completing the cycle, the spheres that can accommodate the most lattice spheres are selected first, and the distance between the sphere and the OAR is r. P The most feasible layout will be the final plan.

[0037] As a further preferred embodiment of this technical solution, in step three, the method for generating a sequence of binary structural images of the same size as the CT image sequence for each structure includes the following steps:

[0038] Step 301: Set the initial value of all pixels in the structured binary image sequence to 0;

[0039] Step 302: For each contour line of each structure, start the grid-based fill algorithm. When executing the algorithm, scan the CT image pixels line by line, starting from the first line of the image.

[0040] Step 303: During the scanning process, for each row of pixels, the boundary of the area to be filled is determined according to rules; the rules are formulated based on the pixel features of the image and the definition of the contour lines;

[0041] Step 304: For each pixel currently scanned, determine its value based on the previously defined boundary conditions. If the pixel is inside the contour line, set its pixel value to 1; otherwise, keep its pixel value at 0.

[0042] Step 305: Continue the above scanning, judgment and assignment operations until the filling process of all contour lines is completed.

[0043] As a further preferred embodiment of this technical solution, in step one, the CT image plane includes a set of sampling grids representing specific thicknesses, lengths, and widths in the three-dimensional human body space. The grids are voxels, and the voxel values ​​represent the density distribution sampling values ​​at the location of the voxel, i.e., the CT values.

[0044] As a further preferred embodiment of this technical solution, in step two, the contour line is drawn according to the grayscale distribution of human organs in the CT image and consists of multiple coordinate points arranged in sequence.

[0045] To solve the above-mentioned technical problems, another technical solution adopted in this application is: a three-dimensional lattice radiotherapy lattice arrangement system, the system comprising: an image data acquisition module, a contour loading module, a binary image generation module, an image transformation and processing module, a lattice sphere center determination module, and a lattice arrangement planning module;

[0046] The image data acquisition module is configured to acquire a CT image sequence, which consists of multiple CT image planes arranged in spatial order and parallel to each other.

[0047] The contour loading module is configured to load contour lines for each structure based on the CT image plane, and each structure contains one or more contour lines, including human organs or tumor structures.

[0048] The binary image generation module is configured with a grid-based filling algorithm to generate a binary image sequence of the same size as the CT image sequence for each structure.

[0049] The image transformation processing module is configured to process the structural binary image sequences of the tumor target volume (GTV) and the planned organ at risk volume (OAR_PRV) based on the erosion algorithm and the dilation algorithm, respectively, and generate the processed image region.

[0050] The lattice center determination module is configured to determine the position of the lattice center by performing layering and scanning on the processed image region;

[0051] The lattice layout planning module is configured to plan a three-dimensional lattice radiotherapy lattice layout scheme based on the determined lattice center position and lattice sphere layout conditions.

[0052] As a further preferred embodiment of this technical solution, the image transformation processing module is further configured with a data storage unit for storing intermediate data generated during the three-dimensional shrinking processing of the GTV structural binary image sequence and the two-dimensional expansion processing of the OAR_PRV structural binary image sequence, including the shrunk image sequence V. L Find the merged image sequence V. PRV .

[0053] As a further preferred embodiment of this technical solution, the lattice arrangement planning module includes a collision detection unit, which is configured to detect whether there is a collision or interference between adjacent lattice spheres when planning a three-dimensional lattice radiotherapy lattice arrangement scheme based on the determined lattice sphere center position and lattice sphere arrangement conditions.

[0054] The embodiments of the present invention have the following advantages due to the adoption of the above technical solutions:

[0055] 1. This invention realizes a series of operations from image acquisition to lattice layout planning through an automated algorithm process. Doctors do not need to manually adjust lattice parameters multiple times, saving a lot of time and effort. It can quickly complete the lattice layout, which is especially suitable for radiotherapy planning of large-volume tumors and greatly improves the design efficiency of radiotherapy plans.

[0056] 2. This invention reduces the differences caused by human factors by arranging the lattice according to a unified standard based on a predetermined algorithm and parameters, ensuring the consistency of lattice arrangement under different patients or different doctors, and improving the quality and stability of radiotherapy planning.

[0057] 3. By performing targeted processing on the structural binary image sequences of GTV and OAR_PRV, and combining it with reasonable lattice microsphere arrangement conditions, this invention can more accurately plan the lattice position, ensuring that while effectively irradiating the tumor, damage to organs at risk is minimized, improving the three-dimensional dose distribution, enhancing the tumor treatment effect, and reducing the risk of complications.

[0058] The above overview is for illustrative purposes only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features of the invention will become readily apparent from the accompanying drawings and the following detailed description. Attached Figure Description

[0059] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0060] Figure 1 This is a flowchart illustrating the three-dimensional lattice radiotherapy lattice arrangement method of the present invention.

[0061] Figure 2 This is a flowchart illustrating the method for generating processed image regions according to the present invention.

[0062] Figure 3 This is a flowchart illustrating the method for determining the location of the crystal sphere center in this invention.

[0063] Figure 4 This is a flowchart illustrating the method of generating a binary image sequence of the same size as the CT image sequence for each structure according to the present invention;

[0064] Figure 5 This is a schematic diagram of the functional modules of the three-dimensional lattice radiotherapy lattice arrangement system of the present invention. Detailed Implementation

[0065] The embodiments of this disclosure will now be described in detail with reference to the accompanying drawings.

[0066] It should be understood that the following specific examples illustrate the implementation of this disclosure, and those skilled in the art can easily understand other advantages and effects of this disclosure from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them. This disclosure can also be implemented or applied through other different specific implementation methods, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this disclosure. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0067] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this disclosure, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0068] It should also be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of this disclosure. The drawings only show the components related to this disclosure and are not drawn according to the number, shape and size of the components in actual implementation. In actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0069] Furthermore, specific details are provided in the following description to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that the described aspects can be practiced without these specific details.

[0070] Figure 1 This is a flowchart illustrating the three-dimensional lattice radiotherapy lattice arrangement method according to an embodiment of the present invention. It should be noted that if substantially the same result is achieved, the method of this application is not necessarily identical. Figure 1 The illustrated process sequence is limited. For example... Figures 1-4 As shown: A three-dimensional lattice radiotherapy lattice arrangement method, including the following steps:

[0071] Step 1: Acquire a sequence of computed tomography (CT) images, which consists of multiple CT image planes arranged in spatial order and parallel to each other;

[0072] Specifically, firstly, the patient is assisted to lie on the CT scanning table. The patient's position is adjusted according to the scanning area and radiotherapy plan to ensure comfort and stability. This avoids artifacts or inaccuracies in the images caused by changes in position during the scan. Positioning devices, such as vacuum pads or thermoplastic films, are typically used to immobilize the patient and ensure consistent positioning for each scan. For example, for patients undergoing head radiotherapy, thermoplastic films may be used to fix the head in a specific position; for patients undergoing chest radiotherapy, vacuum pads may be used to stabilize the body and minimize the impact of respiratory movements on the scan results.

[0073] After confirming that the patient's position and scanning parameters are set correctly, start the CT scan program. During the scan, closely observe the patient's condition and ensure that the patient remains still. At the same time, observe the real-time status of the scanned images through the monitoring system. If any abnormalities are found in the images, such as artifacts or blurring, the scan should be stopped immediately, the cause should be found, adjustments should be made, and the scan should be repeated.

[0074] After the scan is completed, the CT equipment will generate a series of image data. This data is usually transmitted to the image storage and processing system via network or storage medium. During the transmission process, to ensure the integrity and accuracy of the data and avoid data loss or damage, reliable transmission protocols and data verification mechanisms are generally used to verify and correct the transmitted data.

[0075] Finally, the acquired CT image sequences are stored in a dedicated image storage system to provide data support for subsequent radiotherapy planning. The storage system should have sufficient storage space and good data management functions to facilitate the effective organization, retrieval, and backup of large amounts of image data. At the same time, in order to ensure data security, corresponding data protection measures should be taken, such as setting access permissions and performing regular data backups to prevent data leakage or loss.

[0076] Step 2: Load contour lines for each structure based on the CT image plane. Each structure contains one or more contour lines. The structure includes human organs or tumor structures.

[0077] Specifically, first, the CT image sequence obtained in step one is imported into professional medical image processing software. These software programs typically have powerful image display, processing, and analysis capabilities, which can meet the requirements for loading contour lines. This ensures that the software can correctly identify the spatial order and inter-slice relationships of the CT images, so that operations can be performed accurately on each CT image plane.

[0078] Then, based on different scanning locations and imaging characteristics, the window width and window level of the image are adjusted. The window width determines the grayscale range of the image display, while the window level determines the central grayscale value of the display. Appropriate window width and window level settings can make human organs and tumor structures more clearly displayed in the image, which is convenient for subsequent contour drawing. For example, for lung CT images, the window width can usually be set to 1500-2000 HU and the window level can be set to -600-800 HU, which can better observe the fine structures of the lungs.

[0079] Next, professional doctors or medical imaging technicians observe and analyze the CT images to identify the human organs and tumor structures that need to be outlined. Different diseases and radiotherapy plans may focus on different structures. Common human organs include the heart, lungs, liver, and kidneys, while tumor structures are identified based on the specific condition. During the identification process, it is necessary to combine the patient's medical history, clinical symptoms, and other relevant examination results to ensure accurate location and differentiation of each structure. For complex structures or cases with multiple substructures, a reasonable delineation strategy is developed. For example, for the liver, it may be necessary to delineate different areas such as the left lobe and right lobe. For tumors, it may be necessary to distinguish between the main body of the tumor and the infiltrative edge. It is then determined whether to delineate the outline one by one on all CT image planes or to use a method of first delineating at key levels and then extending to other levels through interpolation or automatic segmentation algorithms.

[0080] Subsequently, starting from the first layer of the CT image sequence, the target structure is outlined sequentially on each image plane. During the outlining process, the boundary features of the structure, such as grayscale changes and morphological contours, should be carefully observed to ensure that the outline accurately depicts the actual boundary of the structure. For structures with unclear boundaries, information from multiple layers of images can be used for comprehensive judgment, or different viewing angles and image enhancement methods can be employed to assist in the outlining. When a structure contains multiple outlines, each outline is drawn separately, ensuring that the logical relationships between them are correct. For example, for a tumor with multiple nodules, the outline of each nodule needs to be drawn separately. For structures with regular shapes or obvious features, medical images can be used... Automatic segmentation algorithms, similar to those used in image processing software, are used to assist in contour generation. Common automatic segmentation algorithms include thresholding, region growing, and machine learning-based segmentation. For example, thresholding separates the target structure from the background by setting an appropriate grayscale threshold. Region growing algorithms start from one or more seed points and gradually expand the region based on the similarity of adjacent pixels until the boundary of the structure is reached. Automatic segmentation algorithms may not be able to completely and accurately segment the target structure, requiring manual correction of the segmentation results. It is necessary to check whether the automatically generated contour line is consistent with the actual boundary of the structure. For inaccurate parts, manual drawing tools can be used for adjustment and correction. At the same time, care should be taken to avoid over-correction, which may cause the contour line to lose its original features and accuracy.

[0081] After the outlines of all structures are drawn, the outline data is saved to the database of the medical image processing software. The outline data is usually stored in a specific file format, such as DICOM RT Structure Set format, which can contain detailed information such as the geometric information of the outline, the name and type of the structure to which it belongs. Another professional verifies the drawn outlines to check their accuracy and completeness. The verification process can be carried out by comparing with the original CT image, checking the continuity of the outlines at different levels, and measuring the size and position of the structures. If any problems are found with the outlines, they are modified and adjusted in a timely manner to ensure that the outlines can accurately reflect the actual situation of human organs and tumor structures, providing a reliable basis for the subsequent treatment plan.

[0082] In practice, when importing CT image sequences and RTSS into the Treatment Planning System (TPS), the import may fail due to the image sequence UID (Unique Identifier) ​​being the same. To solve this problem, an option to rewrite the image sequence UID has been added to the image export process.

[0083] The specific rewriting rule is as follows: add 1 to the last digit of the original CT and RTSS UID. When implementing this rewriting rule, the system first reads the original CT and RTSS UID, extracts the last digit through a programming algorithm. Taking Python as an example, assuming the original UID is stored in the string variable uid, the last digit can be obtained using uid[-1]. Then, it is converted to an integer type and incremented by 1. The result is then reassembled back into the UID string. During the processing, the case where the last digit of the original UID is 9 needs to be considered. In this case, adding 1 will result in a carry, and the number of digits and format of the entire UID need to be adjusted accordingly. For example, if the original UID is "123456789", adding 1 should result in "123456790".

[0084] After the rewriting is completed, the system will automatically perform a verification to check whether the rewritten UID is duplicated with an existing UID in the database. If a duplicate is found, the system will add 1 to the last digit of the UID and re-verify until a unique UID is obtained. This process can effectively avoid import failures caused by duplicate UIDs, ensuring that CT image sequences and RTSS can be successfully imported into the treatment planning system, providing a stable data foundation for subsequent radiotherapy planning and implementation.

[0085] Step 3: A grid-based filling algorithm is used to generate a sequence of binary images of each structure, which is the same size as the CT image sequence.

[0086] Specifically, first, a structural binary image sequence of the same size as the CT image sequence is created, and the initial value of all pixels in it is set to 0. This step provides a unified basis for subsequent filling operations. The value of 0 represents the background area, and the area filled with 1 will represent the corresponding human organ or tumor structure.

[0087] Then, for each contour line of each structure, a grid-based filling algorithm is launched. Starting from the first row of the image, the CT image pixels are scanned row by row. During the scanning process, for each row of pixels, the boundary of the area to be filled is determined according to the rules based on the pixel features of the image and the definition of the contour line. These rules may involve various factors such as the gray value of the pixel and the distance from the contour line, so as to accurately determine which pixels belong to the area that needs to be filled.

[0088] Next, for each pixel currently scanned, a judgment is made based on the previously determined boundary conditions. If the pixel is inside the contour line, its pixel value is set to 1; if it is not inside the contour line, its pixel value is kept at 0. In this way, each pixel is judged and assigned a value one by one, and a binary image representing human organs or tumor structures is gradually constructed.

[0089] The scanning, judgment, and assignment operations described above are continued until all contour lines are filled. At this point, in the resulting binary image sequence, pixels with a value of 1 represent the corresponding human organ or tumor structure region, and pixels with a value of 0 represent the background region. This achieves the conversion of each structure into a binary image representation of the same size as the CT image sequence, providing convenience for subsequent image analysis and processing.

[0090] Step 4: Process the structural binary image sequences of the tumor target volume (GTV) and the planned organ at risk (OAR_PRV) based on the erosion algorithm and the dilation algorithm respectively, and generate the processed image region;

[0091] Specifically, firstly, the binary image sequence of GTV structure is shrunk in three dimensions by (m+d / 2) millimeters based on the erosion algorithm to obtain the shrunk image sequence V. L Here, the erosion algorithm shrinks the GTV region by gradually eroding away the foreground pixels (pixels with a value of 1) in the image; where m is the distance from GTV to V. L The inward shrinkage distance, where d is the diameter of the lattice microsphere; in actual operation, according to the preset parameters, the erosion algorithm is used to process the binary image of each layer of GTV, causing the GTV region to shrink inward by (m+d / 2) millimeters in three-dimensional space.

[0092] Then, based on the dilation algorithm, two-dimensional expansion (r) is performed on all OAR_PRV structural binary image sequences. PProcessing the data by +d / 2) millimeters and performing a union operation yields the merged image sequence V. PRV The dilation algorithm, unlike the erosion algorithm, expands the foreground pixels in an image outwards. For each layer of the binary image in OAR_PRV, the dilation algorithm expands outwards according to the set parameters (r). P The image is expanded in two dimensions by +d / 2) millimeters, and then all the expanded images are merged to obtain a comprehensive V. PRV Image; where r P The distance between the lattice sphere and OAR_PRV;

[0093] Next, find V PRV Voxels with a median value of 1, and V L The voxel value at the corresponding position is set to 0. This step is to ensure that, during lattice arrangement, lattice spheres are not placed in areas that could cause high-dose radiation to organs at risk. L V corresponds to PRV Voxel positions with a median value of 1 are set to 0. These regions will be excluded when determining the location of the lattice sphere center and arranging the lattice spheres, thereby protecting the organs at risk.

[0094] Step 5: Determine the location of the lattice sphere center by layering and scanning the processed image region;

[0095] Specifically, first, find V in the processed image region. L In the (reduced tumor target area image sequence), the intermediate layer S0 of the foreground region typically refers to the area composed of pixels with a value of 1 in the image. This represents the effective portion of the tumor target area after processing, and can be calculated using V... L The range of all foreground regions along the axis is considered, and the layer corresponding to the middle position is taken as S0. The S0 layer image is scanned in the order of x, y, z, and foreground voxels (voxels with a value of 1) are added to the candidate list L. possible In this step, the goal is to identify all possible points in the intermediate layer that could serve as lattice centers.

[0096] Then, take list L possible The first voxel point in the array is added to the lattice sphere center list L0, and from L... possible Remove this point from the list; L0 will then be used to store the lattice center position of the intermediate layer S0; remove candidate list L... possible All voxel points in the list L0 whose distance from the last center of the lattice spheres is less than r millimeters, where r is the minimum distance between the centers of the lattice spheres. This is to ensure sufficient spacing between the lattice spheres to avoid overlap or excessive proximity, which could affect the radiotherapy effect. Repeat the above steps of selecting centers and screening candidate points until list L0 is reached. possibleIf it is empty, then L0 stores the positions of all lattice sphere centers in the intermediate layer S0 that meet the spacing requirements;

[0097] Next, taking the middle layer S0 as the center, search for layers with a spacing of s millimeters in both the top and bottom directions, that is, layers that satisfy S i+1 With S i The distance between the two layers is s millimeters; here, the interlayer spacing s is a preset parameter used to control the distribution of the lattice along the z-axis; s is arranged in the order of x, y, z. i The layer image is scanned and compared with the list of spherical centers of the previous layer L. i-1 Foreground voxel points whose center-to-sphere distance is not less than d1 mm (when i>0) or (when i<0) are added to the candidate list L. possible In the diagram, d1 represents the two-dimensional distance between the center of the lattice spheres in each layer and the center of the spheres in adjacent layers. This step ensures sufficient spacing between the lattice spheres in adjacent layers. (See list L.) possible The first voxel point in the array is added to the lattice sphere center list L. i In, and from list L possible Remove the point from the list, then remove the candidate list L. possible List of all lattice sphere centers L i Voxel points with a distance of less than r millimeters between the last sphere centers; repeat the above steps of scanning, selecting sphere centers, and screening candidate points until all layers have been processed, and finally select from the lattice sphere center list L of each layer. i The positions of the centers of all crystal lattices are obtained from this.

[0098] Step 6: Based on the determined lattice center position and lattice sphere arrangement conditions, plan the lattice arrangement scheme for three-dimensional lattice radiotherapy.

[0099] Specifically, firstly, based on the determined positions of the lattice sphere centers, a virtual model is constructed in three-dimensional space. This model covers all the determined lattice sphere centers, and their positions in three-dimensional space are determined according to the sphere center coordinates. At the same time, according to the requirements of the lattice sphere arrangement conditions, such as the interlayer spacing s mm and the two-dimensional distance between the sphere centers of each layer and the sphere centers of adjacent layers not less than d1 mm, the relative positional relationship of each sphere center in space is clarified, and the lattice framework structure is initially constructed.

[0100] Then, based on the diameter d mm in the arrangement conditions of the lattice spheres, the radius of the lattice spheres is calculated to be d / 2 mm. In the three-dimensional model constructed above, spheres are drawn with the center of each lattice sphere as the center and the radius as d / 2 mm to represent the actual lattice spheres, ensuring that the size of each lattice sphere meets the set standard, and providing accurate geometric shape basis for subsequent precise planning of radiotherapy dose;

[0101] Next, the constructed lattice arrangement is comprehensively evaluated against the lattice sphere arrangement conditions. This includes checking whether the distance between the lattice spheres meets the requirement that the minimum distance between the centers of the lattice spheres is r millimeters, and whether the distance between the lattice spheres and the OAR_PRV is not less than r. P If there are situations where the conditions are not met, such as some lattice microspheres being too close together or not close enough to OAR_PRV, the position of the lattice center will be fine-tuned. When adjusting, priority will be given to areas that have less impact on the radiotherapy effect. The lattice arrangement will be optimized by changing the coordinates of the center to ensure that all arrangement conditions are met.

[0102] After evaluation and adjustments, a three-dimensional lattice radiotherapy lattice arrangement that meets all lattice sphere placement conditions is determined. Detailed information such as the final lattice sphere center coordinates, lattice sphere radii, and spatial relationships between each lattice sphere is compiled into a complete protocol document. This document should include clear text descriptions and intuitive three-dimensional diagrams to facilitate understanding and implementation by radiotherapy team members (such as doctors and physicists), providing accurate guidance for subsequent radiotherapy planning and implementation.

[0103] In one embodiment, specifically, in step four, the method for generating the processed image region includes the following steps:

[0104] Step 401: Based on the erosion algorithm, perform three-dimensional shrinking (m+d / 2) millimeters on the structural binary image sequence of GTV to obtain the shrunken image sequence V. L Where m is the distance from GTV to V L The inward distance, where d is the diameter of the lattice microsphere;

[0105] The determination of m requires comprehensive consideration of the tumor's biological characteristics, radiotherapy dose distribution requirements, and clinical experience. For actively growing and highly invasive tumors, the m value will be appropriately increased to ensure thorough radiotherapy; for relatively slow-growing and well-defined tumors, the m value can be moderately decreased. Clinicians determine the appropriate m value based on their long-term accumulated treatment experience and relevant research data, combined with the specific tumor condition of each patient. d is determined based on the performance of the radiotherapy equipment, the size and location of the tumor, and the precision requirements of the radiotherapy plan. For high-precision radiotherapy equipment that enables more precise dose control, the d value can be relatively reduced; for smaller tumors in special locations with high radiotherapy precision requirements, the d value also needs to be adjusted accordingly. In practice, doctors and physicists discuss and determine the d value based on the specific radiotherapy plan.

[0106] Specifically, firstly, among the many erosion algorithms, it is important to select one suitable for medical images. Common erosion algorithms include morphology-based algorithms, such as using structuring elements (e.g., squares, circles, etc.) to perform convolution operations on the image. In medical image processing, considering the irregularity of tumor shapes, algorithms with flexibly adjustable structuring elements are more suitable. Disk-shaped structuring elements can better adapt to the circular or near-circular regions of tumors, while cross-shaped structuring elements may be more advantageous when processing directional tumor boundaries. Distance-transform-based erosion algorithms can also be selected, which perform erosion operations by calculating the distance from each pixel to the background pixel, allowing for more precise control over the depth and range of erosion. When selecting an algorithm, its efficiency and accuracy should be considered to ensure that processing is completed within a reasonable time and meets accuracy requirements.

[0107] Then, the structural binary image sequence of GTV is imported into medical image processing software. The software must have powerful image processing capabilities, supporting 3D image processing and the execution of the selected erosion algorithm. Using the software's algorithm implementation module, the structural binary image sequence of GTV is subjected to 3D shrinkage of (m+d / 2) millimeters according to the selected erosion algorithm. During processing, each layer and each pixel of the image is operated on. Based on the algorithm rules, and according to the structural elements and shrinkage distance, it is determined whether each pixel should be eroded. For 3D images, not only the pixel relationships in the plane but also the pixel relationships between layers must be considered to ensure that the shrinkage operation is performed accurately in 3D space. After processing all images, the shrunken image sequence V is obtained. L ;

[0108] Finally, the processed image sequence V L Visual inspection is performed, comparing the original binary image sequence of the GTV structure to observe whether the tumor region has shrunk uniformly, whether the boundary after shrinkage is smooth and continuous, and whether there are any abnormal cavities or protrusions. Image analysis tools are used to measure parameters such as the tumor volume and surface area after shrinkage and compare them with theoretical calculation values. The theoretical value of the tumor volume after shrinkage is calculated according to the formula. The measured value is compared with the theoretical value. If the error is within a reasonable range (e.g., within 5%), it indicates that the processing result is relatively accurate. If the error is large, it is necessary to check whether there are problems with parameter settings, algorithm selection, and execution process, and adjust and reprocess in time.

[0109] Step 402: Perform two-dimensional external expansion (r) on all OAR_PRV structural binary image sequences based on the dilation algorithm. P Processing the data by +d / 2) millimeters and performing a merging operation yields the merged image sequence V. PRV , where r P The distance between the lattice sphere and OAR_PRV;

[0110] Wherein, parameter r PDetermining the appropriate level of radiation therapy requires comprehensive consideration of multiple factors. Doctors must take into account the patient's specific condition, the sensitivity of the organs at risk, and the expected outcomes of the radiotherapy plan. For highly sensitive organs, such as the eyes and spinal cord, [the specific treatment plan needs to be considered]. P The value of r is usually set relatively large to give these organs a wider safety margin; while for relatively less sensitive organs, r P The value can be appropriately reduced, but it is still necessary to ensure that the organs are effectively protected from excessive radiation doses; d is the diameter of the lattice microspheres, and its size will also affect the degree of outward expansion. When determining d, factors such as the performance of the radiotherapy equipment, the size and location of the tumor should be taken into account to ensure the accuracy and effectiveness of radiotherapy.

[0111] Specifically, firstly, a dilation algorithm must be selected. Common dilation algorithms include morphological dilation algorithms, which use structuring elements (such as squares, circles, crosses, etc.) to perform convolution operations on the image to achieve image dilation. When processing structured binary image sequences of OAR_PRV, a suitable structuring element must be selected based on the image characteristics and processing requirements. If the shape of the OAR_PRV is relatively regular, a circular structuring element may result in a more uniform expansion effect. If the OAR_PRV has a clear directionality, a cross-shaped structuring element may be more suitable, enabling more effective expansion in a specific direction. Alternatively, a distance-transform-based dilation algorithm can be chosen. This algorithm determines the degree of dilation by calculating the distance from each pixel to the background pixel, allowing for more precise control of the expansion range. In practical applications, the efficiency and accuracy of the algorithm must be considered comprehensively to ensure that image processing is completed within a reasonable time and meets accuracy requirements.

[0112] Then, all OAR_PRV structural binary image sequences are imported into software or systems with image processing capabilities. Using the selected dilation algorithm, each OAR_PRV structural binary image is subjected to two-dimensional external expansion (r... P The process involves processing (+d / 2) millimeters. During the processing, each pixel in the image is judged and processed according to the algorithm rules. Taking the morphology-based dilation algorithm as an example, the structuring element slides on the image. If there is a foreground pixel (a pixel with a value of 1) in the area covered by the structuring element, the current center pixel is set as the foreground pixel, thereby realizing the outward expansion of the image. In the two-dimensional outward expansion process, the operation is only performed in the horizontal and vertical directions, without involving changes in the depth direction.

[0113] After completing the two-dimensional expansion processing of the structural binary images of all OAR_PRVs, a merging operation is performed on these expanded images. The purpose of the merging operation is to combine the expanded regions of each OAR_PRV into a single merged region, resulting in the merged image sequence V. PRVDuring the operation, for each pixel location, if the value of that pixel is 1 in any of the expanded OAR_PRV images, then in V... PRV The value of this pixel in the image is also set to 1; V only becomes true when the value at this pixel location is 0 in all expanded images. PRV The value of that pixel in the image is 0; in this way, all the risk areas that endanger organs are integrated together, providing a basis for avoiding these areas when determining the location of the lattice center.

[0114] Step 403, find V PRV Voxels with a median value of 1, and V L Middle and V PRV and V L The voxel values ​​in the intersection are set to 0;

[0115] Specifically, firstly, medical image processing software or programming tools are used to analyze V... PRV Each image in the image sequence is traversed pixel by pixel, starting from the top-left pixel and checking the value of each pixel sequentially from left to right and top to bottom. Since V PRV It is a binary image sequence, where the pixel values ​​are only 0 and 1. Therefore, during the traversal, we focus on the pixels with a value of 1.

[0116] When in V PRV When a voxel with a value of 1 is found (a pixel in a 2D image and a voxel in a 3D image, referred to as a voxel here), the location of that voxel in V is determined according to the image's spatial coordinate system. L The corresponding position in the image, because V L and V PRV They are obtained by processing the same CT image sequence, and their spatial coordinates have a corresponding relationship; for example, if V PRV If the coordinates of a certain voxel are (x, y, z), then in V L The corresponding voxel coordinates are also (x, y, z);

[0117] Find V L After selecting the voxel at the corresponding position, its value is set from its original value (which may be 0 or 1) to 0. This step is implemented programmatically, using assignment statements in the code to set the voxel value. L Modify the pixel value of the corresponding voxel in the image to 0; for example, when processing images using the NumPy library in Python, the value of the corresponding element in the array can be directly modified through index operations; repeat the above traversal, confirmation, and modification operations until V has been traversed. PRV All images in the image sequence, ensuring V PRV V of all voxels with a value of 1 LThe voxel values ​​at the corresponding positions were all set to 0.

[0118] In one embodiment, specifically, in step five, the method for determining the location of the lattice sphere center includes the following steps:

[0119] Step 501, find V L The middle layer S0 of the foreground region is scanned in the order of x, y, z, and foreground voxels are added to the candidate list L. possible middle;

[0120] Specifically, firstly, in a medical image processing software or programming environment, the compressed image sequence V is read. L Because of V L It is a binary image; the voxel value of the foreground region (representing the portion of the tumor target area after inward contraction) is typically 1, while the voxel value of the background region is 0. This is achieved by traversing V... L For each voxel in the array, the regions with a voxel value of 1 are counted to determine the extent of the foreground region.

[0121] Then, after determining the foreground region, calculate the number of layers in the foreground region along the z-axis; assuming the foreground region extends from layer z1 to layer z2 along the z-axis, the z-coordinate of the intermediate layer S0 can be obtained using the formula:

[0122] or To calculate, find V in this way L The middle layer S0 in the foreground region;

[0123] Next, the intermediate layer S0 is scanned in the order of x, y, z. On the two-dimensional plane (xy plane), starting from the top left corner of the image (i.e., the position with the smallest x and y coordinates), each pixel is scanned line by line (for a two-dimensional layer in a three-dimensional image, a pixel is equivalent to the projection of a voxel into that layer). During the scanning process, it is determined whether the value of each voxel is 1. If it is, the voxel point (i.e., its coordinate information in the image) is added to the candidate list L. possible In the middle; this candidate list L possible This will be used to subsequently select suitable lattice center positions.

[0124] Step 502, retrieve list L possible The first voxel point in the matrix is ​​added to the lattice sphere center list L0, and from L... possible Remove from;

[0125] Specifically, firstly, in the programming environment, the list L storing candidate voxel points is accessed through the corresponding data structure operation functions. possibleGenerally, lists are ordered data structures, and most programming languages ​​have built-in methods to access their elements; for example, in Python, if L... possible It is a list, and index operations can be used. possible Use [0] to get the first element in the list;

[0126] Then, in obtaining L possible After the first voxel point in the array, it is added to a list L0 specifically used to store lattice sphere centers; in Python, this can be done using the list's append() method, such as L0.append(L... possible [0]), the selected voxel point is added to the end of L0, which means that this voxel point is used as a potential lattice center for subsequent analysis and processing;

[0127] To avoid repeatedly selecting the same voxel point as the lattice sphere center, it is necessary to start from L... possible Remove the voxel point that has been added to L0; in Python, this can be done using the list's pop() method, specifying index 0, i.e., L. possible .pop(0), thus popping from L possible The first element was deleted; through this step, L... possible The list was updated, retaining only the remaining candidate voxel points, so that suitable lattice center locations could be further selected later.

[0128] Step 503: Remove candidate list L possible All voxel points in the list L0 whose distance from the last sphere center is less than r millimeters, where r is the minimum distance between the sphere centers;

[0129] Specifically, firstly, in the program, the list of lattice centers L0 is accessed. Since the list is an ordered data structure, the last determined coordinates of the lattice center are obtained by accessing the last element of the list. In many programming languages, specific indexing methods can be used to obtain the coordinates. For example, in Python, the last element of the list L0 can be obtained by using L0[-1]. This element contains the coordinate information representing the position of the center (assuming it is a three-dimensional coordinate (x, y, z)).

[0130] Then, for the candidate list L possible For each voxel point in L0, calculate its distance to the center of the last sphere in L0. The distance is typically calculated using the Euclidean distance formula. For the voxel coordinates (x1, y1, z1) and the sphere center coordinates (x2, y2, z2), the formula for calculating the distance d is:

[0131] ;

[0132] Next, the calculated distance is compared with the minimum distance r between the centers of the lattice spheres. If the distance is less than r, it indicates that the voxel point is too close to the determined center, which does not meet the lattice arrangement requirements, and it needs to be moved from L. possible Remove from the middle; if the distance is greater than or equal to r, retain the voxel point;

[0133] Traversing L possible After completing distance calculation and filtering for all voxel points, L possible The list will change, retaining only voxel points that are at least r millimeters away from the center of the last sphere in L0. This will update L... possible The list provides a more reasonable set of candidate points for further determination of lattice centers, ensuring that the final determined lattice centers maintain an appropriate distance from each other, meeting the requirements for lattice distribution in the radiotherapy plan.

[0134] Step 504: Repeat steps 502-503 above until list L is reached. possible Empty;

[0135] Specifically, before performing this step, it must be ensured that step 501, which yielded the candidate list L, has already been completed. possible And L has already been possible The first voxel point in the matrix is ​​added to the lattice sphere center list L0 (i.e., step 502 is completed), and at the same time, L... possible The initial screening was conducted (i.e., step 503 was completed);

[0136] As long as the candidate list L possible If it is not empty, continue to repeat steps 502 and 503;

[0137] When L possible When the list becomes empty, it means that all lattice centers that meet the conditions have been selected and added to L0, and the loop ends. At this time, L0 stores the positions of all lattice centers on the intermediate layer S0 that meet the minimum distance requirement of the centers.

[0138] Step 505: Using the intermediate layer S0 as the center, find layers with a spacing of s millimeters in both the top and bottom directions, i.e., layers that satisfy S... i+1 With S i The distance between the two layers is s millimeters;

[0139] Specifically, first, the search is performed with the intermediate layer S0 as the center, moving towards the positive (upward) and negative (downward) directions of the z-axis. Simultaneously, the interlayer spacing parameter s is determined. The value of s is typically determined based on factors such as the accuracy requirements of the radiotherapy plan, the size and shape of the tumor, and the uniformity requirements of the lattice distribution. In practice, the value of s is generally between a few millimeters and tens of millimeters. For example, for some larger tumors with relatively low radiotherapy accuracy requirements, s may be set to 5-10 millimeters; while for smaller tumors with surrounding vital organs, s may be set to 2-5 millimeters to ensure radiotherapy effectiveness and safety.

[0140] Then, starting from the middle layer S0, layers that meet the conditions are searched sequentially along the positive z-axis at intervals of s. During the search, the target layer is determined by whether the distance between adjacent layers is equal to s. In medical image data, each layer usually has corresponding coordinate information (such as z-coordinate), and the distance between adjacent layers can be calculated based on this coordinate information. Assuming the z-coordinate of the middle layer S0 is z0, then the z-coordinate of the first layer s1 searched upwards should be z0+s. By traversing the image sequence, the layer with the z-coordinate closest to z0+s is found as s1. If no layer with a distance of exactly s is found during the search, an interpolation algorithm can be used to process the image to simulate a layer that meets the interlayer spacing requirements; or the layer closest to s and meeting a certain error range (such as within ±5 mm) can be selected as the approximate target layer.

[0141] Next, using a method similar to the upward search, starting from the middle layer S0, searching along the negative z-axis at intervals of interlayer spacing s for layers that meet the conditions, the first layer s in the downward search is... -1 The z-coordinate should be z0-s; similarly, by traversing the image sequence, the layer with coordinates closest to z0-s is found as s. -1 If a layer with a distance of exactly 's' cannot be found, it can be handled in the same way as when searching upwards to deal with errors.

[0142] Finally, after finding the layers that meet the interlayer spacing requirements (including layers found by searching upwards and downwards), record the relevant information of these layers, such as the layer number (e.g., s1, s2, s3, s4, s5, s6, s7, s8, s9, s1 ... -1 The information, such as the position of the layer in the image sequence and the image data of the layer, will be used in subsequent steps to determine the position of the lattice center of the layer.

[0143] Step 506: Arrange S in the order of x, y, z. i Layer image scanning will be compared with the previous layer's lattice sphere center list L i-1 (when i>0) or L i+1 Foreground voxel points whose center-to-sphere distance is not less than d1 mm (when i < 0) are added to the candidate list L. possibleIn the diagram, d1 represents the two-dimensional distance between the center of the lattice microspheres in each layer and the center of the microspheres in the adjacent layer.

[0144] Specifically, first, sort S in the order of x, y, z. i The layer image is scanned in two dimensions S i On the image plane, starting from the top left corner (where the x and y coordinates are smallest), scan each pixel row by row from left to right (in a 3D image, these pixels are actually the projections of voxels onto that layer). After completing one row, move to the next row and continue scanning until the entire S layer has been traversed. i Layered images;

[0145] Then, the list of reference lattice sphere centers is determined based on the value of i; when i>0, the previous layer s is referenced. i-1 List of lattice sphere centers L i-1 When i < 0, refer to the next layer s. i+1 List of lattice sphere centers L i+1 These lists store the locations of the lattice centers that have been determined in the previous or next layer (assuming that each center location is represented by three-dimensional coordinates (x, y, z)).

[0146] For S i For each foreground voxel point scanned in the layer image (foreground voxels generally refer to voxel points with a value of 1, indicating that the region may be suitable for placing the lattice sphere center), calculate its relationship with the reference list (L). i-1 or L i+1 The two-dimensional distance (distance in the xy-plane) between the centers of each sphere in the xy-plane is calculated using the Euclidean distance formula. For the voxel coordinates (x1, y1) and the sphere center coordinates (x2, y2), the formula for calculating the two-dimensional distance d is:

[0147] ;

[0148] The calculated distance d1 is compared with the two-dimensional distance between the center of the lattice sphere in each layer and the center of the sphere in the adjacent layer. If the distance is not less than d1 mm, it means that the distance between the voxel point and the center of the sphere in the adjacent layer meets the requirements, and it is added to the candidate list L. possible If the distance is less than d1 mm, the voxel point does not meet the requirements and no addition operation will be performed.

[0149] After scanning S i After calculating distances and filtering all voxel points in the layer image, the candidate list L is generated. possible Updated; L at this point possible Includes S iForeground voxel points in the layer image that meet the distance requirements from the center of the adjacent layer will be used as candidate points for the lattice center of that layer, and will be further screened to determine the final lattice center position.

[0150] Step 507, retrieve list L possible The first voxel point in the array is added to the lattice sphere center list L. i In, and from L possible Remove from;

[0151] Specifically, access the candidate list L possible In most programming languages, lists are ordered data structures, and the first element can be retrieved using a specific indexing method; for example, in Python, if L... possible It is a list, using L possible [0] will give you the first voxel point, which is usually represented by three-dimensional coordinates (x, y, z) to indicate its position in the image;

[0152] Add the first voxel point obtained to the current layer S i The corresponding list of lattice centers L i In Python, the list's append() method can be used, such as Li.append(L_{possible}[0]), to add the voxel point to L. i At the end, it means that this voxel point has been identified as a candidate location for the center of a lattice in the current layer;

[0153] To avoid repeatedly selecting this voxel point in the future, it needs to be removed from L. possible Remove from L; in Python, use the list's pop() method, specifying the index as L, i.e., L_{possible}.pop(0), thus removing from L. possible The first element was deleted from L, and the updated L possible It only contains the remaining candidate voxel points, which can be used to further screen suitable lattice center locations.

[0154] Step 508: Remove candidate list L possible List of all lattice sphere centers L i The last voxel point with a distance of less than r millimeters between its centers;

[0155] Specifically, firstly, it is necessary to start from the current layer S i The corresponding list of lattice centers L l To obtain the coordinates of the last sphere's center, in most programming languages, since lists are ordered data structures, the last element can be retrieved using a specific indexing method; for example, in Python, using L... l [-1] can be used to obtain L lThe last element of the list, which is usually represented by three-dimensional coordinates (x, y, z) indicating the position of the sphere's center in the image;

[0156] For candidate list L possible Traverse each voxel in L, for L possible For each voxel point in L, calculate its relationship with L. l The distance between the last center of the sphere; in three-dimensional space, the Euclidean distance formula is usually used to calculate the distance between two points. Let the coordinates of the voxel point be (x1, y1, z1), and the coordinates of the last center of the sphere be (x1, y1, z1). Then the formula for calculating the distance d between them is:

[0157] ;

[0158] The calculated distance d is compared with the preset minimum distance r between the centers of the lattice spheres; if d < r, it means that the voxel point is at L l The distance to the center of the last sphere is too close, which does not meet the requirements. The voxel point needs to be moved from L. possible Remove from the list; if d ≥ r, retain the voxel; in Python, you can create a new empty list and iterate through L. possible When the conditions are met, add the voxel points that meet the criteria to a new list, and finally replace L with the new list. possible Complete the update of the candidate list.

[0159] Step 509: Repeat steps 507-508 above until list L is reached. possible Empty, and process all layers, finally from the list L of lattice centers of each layer. i The positions of the centers of all lattices are obtained from this.

[0160] Specifically, regarding a certain layer S i During processing, ensure that the candidate list L possible Not empty, as long as L possible If there are still voxel points, continue executing steps 507 and 508;

[0161] When L possible When the list becomes empty, it means that all lattice centers that meet the criteria in that layer have been filtered out and added to L. l In the middle, the single-layer loop ends;

[0162] After processing the intermediate layer S0, a series of layers S have been determined in the upward and downward directions. i (i is a positive or negative integer), process these layers in order, from the bottom layer to the top layer or vice versa;

[0163] For each layer S iThe above single-level loop operation is repeated, that is, continuously starting from L in that level. possible Select lattice spheres from the middle and add them to L i In, until L possible Empty;

[0164] After all layers have been processed, list the lattice centers L for each layer. i When combined, this results in a comprehensive list containing the center positions of all lattices in all layers. This comprehensive list records the center position information of all lattices in the entire three-dimensional space.

[0165] In one embodiment, specifically, the arrangement conditions of the lattice spheres include:

[0166] The diameter of the lattice microspheres is d millimeters. The size of the lattice microspheres is determined, which will affect the dose distribution during radiotherapy and the interaction with surrounding tissues. Smaller diameters may allow for more precise dose delivery, but may require more lattice microspheres to cover the tumor target area; larger diameters, on the other hand, may result in a relatively wider dose distribution.

[0167] GTV to V L The inward distance is m millimeters; GTV is obtained by inward processing to obtain V. L The m value determines the degree of inward retraction. The inward retraction operation is to leave a certain safe boundary around the tumor target area, and at the same time prepare for the subsequent determination of the lattice sphere center position, so as to avoid the lattice spheres being too close to the tumor edge, affecting the radiotherapy effect or causing unnecessary radiation to the surrounding normal tissues.

[0168] The minimum distance between the centers of the lattice microspheres is r millimeters; the minimum spacing between the lattice microspheres is specified to ensure that there is enough distance between them, which can avoid excessive dose superposition between the microspheres, ensure that the radiotherapy dose is more evenly distributed in the tumor target area, and also facilitate physical arrangement and operation.

[0169] The distance between the lattice sphere and OAR_PRV is not less than r P millimeters; OAR_PRV is the area related to the endangered organ that needs protection, r P This ensures that the lattice microspheres maintain a sufficient safe distance from these sensitive areas to reduce potential damage to organs at risk from radiotherapy and protect the patient's normal physiological functions;

[0170] The interlayer spacing was set to s millimeters, determining the distance between different layers of lattice microspheres in three-dimensional space. A suitable interlayer spacing helps to rationally distribute the lattice microspheres in the vertical direction, enabling more uniform coverage of the radiotherapy dose within the three-dimensional space of the tumor target area, while also taking into account the anatomical structure of human tissues and the operational requirements of radiotherapy equipment.

[0171] The two-dimensional distance between the center of each lattice microsphere and the center of the lattice microsphere in the adjacent layer is not less than d1 mm. The positional relationship between the lattice microspheres in the adjacent layers is further constrained. In addition to the interlayer spacing, a certain distance is also guaranteed in the two-dimensional plane. This helps to optimize the distribution of lattice microspheres in three-dimensional space and avoid the lattice microspheres in the adjacent layers from being too close, which may lead to uneven dose distribution or adverse effects on the surrounding tissues.

[0172] In one embodiment, specifically, in step three, the method for generating a sequence of binary structural images of the same size as the CT image sequence for each structure includes the following steps:

[0173] Step 301: Set the initial value of all pixels in the structured binary image sequence to 0;

[0174] Specifically, in the image preprocessing stage of radiotherapy, structural binary image sequences are used to represent the distribution of structures such as GTV and OAR_PRV. Setting the initial value of all pixels in the structural binary image sequence to 0 lays the foundation for accurate labeling and processing of these structures in the future.

[0175] In medical image processing, binary images have only two pixel values: 0 represents the background region and 1 represents the foreground region (i.e., the region of interest). Initializing all pixels to 0 is equivalent to first setting the entire image region as the background, and then modifying the pixel values ​​of the corresponding regions to 1 according to the actual structural information, so as to highlight the tumor target area and structures such as organs at risk.

[0176] For example, for a sequence of binary structural images containing multiple tomographic images, where each image represents a cross-section of the human body, during initialization, regardless of whether the cross-section contains a tumor target area or endangered organs, all pixels are first marked as 0, meaning that it is assumed that there is no structure of interest at present.

[0177] Step 302: For each contour line of each structure, start the grid-based fill algorithm. When executing the algorithm, scan the CT image pixels line by line, starting from the first line of the image.

[0178] Step 303: During the scanning process, for each row of pixels, the boundary of the area to be filled is determined according to rules; the rules are formulated based on the pixel features of the image and the definition of the contour lines;

[0179] Specifically, first, we need to identify the structures to be processed, such as GTV and OAR_PRV. For each structure, we need to obtain its corresponding contour information. These contours are usually obtained in the previous image segmentation steps. They delineate the boundaries of the structure. Each contour contains a series of coordinate points, which define the shape of the structure in the image.

[0180] For each contour line of each structure, a grid-based filling algorithm is initiated, starting from the first row of the image (usually the first row of the image, i.e. the row where y=0, assuming that the axis in the image coordinate system represents the row direction) and scanning line by line. During the scanning process, each pixel in the CT image is checked in order from left to right.

[0181] During the scanning process, for each pixel, it is determined whether it is located within the area enclosed by the current contour line. Several methods can be used for this determination, such as the ray casting method. The ray casting method involves emitting a ray from the pixel towards one side of the image (e.g., the right side) and counting the number of intersections between this ray and the contour line. If the number of intersections is odd, the pixel is located within the area enclosed by the contour line, and its corresponding pixel value in the structured binary image sequence is set to 1 (representing the foreground region). If the number of intersections is even, the pixel is located outside the contour line, and its corresponding pixel value remains 0 (representing the background region).

[0182] After completing the scanning and judgment of a row of pixels, move to the next row and repeat the above steps until all rows of the entire CT image have been scanned. For each contour line of each structure, repeat the above filling algorithm execution process until the contour lines of all structures have been processed, thereby accurately marking the region of each structure in the binary image sequence of the structure.

[0183] Step 304: For each pixel currently scanned, determine its value based on the previously defined boundary conditions. If the pixel is inside the contour line, set its pixel value to 1; otherwise, keep its pixel value at 0.

[0184] Specifically, when performing this step, the previously determined contour information must first be clarified. These contours constitute the boundary conditions for determining whether a pixel is within a structural region. At the same time, the coordinate position of the currently scanned pixel in the image is obtained (usually represented by (x, y), where x is the column coordinate and y is the row coordinate). For example, when scanning a two-dimensional CT image line by line, the coordinate information of each pixel is obtained sequentially as the scan progresses.

[0185] To determine whether a pixel is inside the contour line, a suitable algorithm can be used. A common method is the ray casting method, which involves casting a ray from the pixel to one side of the image (e.g., the right side) and counting the number of intersections between the ray and the contour line. If the number of intersections is odd, the pixel is located within the area enclosed by the contour line; if the number of intersections is even (including 0), the pixel is outside the contour line. Other methods can also be used, such as the corner method, which determines whether a pixel is inside the contour line by calculating the corner of the pixel relative to the vertex of the contour line.

[0186] The pixel value is set according to the judgment result; if it is determined that the pixel is inside the contour line, its pixel value in the structural binary image is set to 1, indicating that the pixel belongs to the structural region (foreground); if it is determined that the pixel is not inside the contour line, its pixel value is kept at 0, that is, the pixel belongs to the background region; during the process of scanning the entire image line by line, the above judgment and pixel value setting operation is repeated for each pixel until the processing of the entire image is completed.

[0187] Step 305: Continue the above scanning, judgment and assignment operations until the filling process of all contour lines is completed;

[0188] Specifically, before starting this step, ensure that all contour information that needs to be processed has been acquired. These contours may come from different structures (such as tumor target areas, organs at risk, etc.). At the same time, determine the starting position of the scan, generally starting from the first row of the image (such as the first row), and scan the CT image pixels row by row in order from left to right.

[0189] The image pixels are scanned line by line. Starting from the first line of the image, for each pixel in each line, a judgment operation is performed; that is, based on the previously determined boundary conditions (defined by the contour line), an appropriate algorithm (such as ray casting, corner turning, etc.) is used to determine whether the pixel is inside the contour line.

[0190] The assignment operation is performed based on the judgment result; if the pixel is inside the contour line, its pixel value in the structured binary image is set to 1; if it is not inside the contour line, its pixel value is kept at 0.

[0191] After completing the scanning, judgment and assignment of a row, move to the next row and repeat the above operation until all rows involved in the current outline have been scanned;

[0192] For all contour lines, repeat the scanning, judgment, and assignment process described above in sequence. After processing each contour line, proceed to the next one, until all contour lines have been processed;

[0193] Once all contour lines have been filled, the regions of each structure have been accurately marked in the binary image. The pixel value of the foreground region (the region where the structure is located) is , and the pixel value of the background region is , thus completing the generation of the binary image.

[0194] In one embodiment, specifically in step one, the CT image plane represents a set of sampling grids with specific thickness, length and width in three-dimensional human body space. The grid is a voxel, and the voxel value represents the density distribution sampling value at the location of the voxel, i.e., the CT value.

[0195] CT images are two-dimensional images obtained by tomographic scanning of the three-dimensional human body. The CT image plane is actually a cross-section of the three-dimensional human body space, representing a sample of the human body within a specific range of thickness, length, and width. This specific thickness is determined by the parameters of the CT scanning equipment. For example, common CT scan slice thicknesses may have different specifications such as 1mm, 2mm, and 5mm. The length and width correspond to the range of the CT image on the plane, covering the size of the corresponding part of the human body.

[0196] Taking a chest CT scan as an example, the CT image plane obtained from the scan is like slicing the chest at a certain height, showing the structural information of the human tissue at that slice. Through a series of such CT image planes (slices at different heights), a three-dimensional human chest model can be reconstructed in a computer.

[0197] CT images are composed of discrete sampling points, which form a set of sampling grids. Each sampling point in a CT image is called a voxel (short for volume pixel), which can be understood as a pixel in three-dimensional space. A voxel is the basic building block of a CT image. It has a certain spatial size and corresponds to a tiny volume in the three-dimensional human body space.

[0198] The size of a voxel depends on factors such as the resolution of the CT scanning device. High-resolution CT scanning devices can produce smaller voxels, thus providing more detailed image information. For example, in some high-end CT devices, the size of voxels can be very small, allowing doctors to observe the fine structure of human tissues more clearly.

[0199] Voxel value represents the density distribution sample value at the location of the voxel, also known as CT value. CT value is a quantitative value used to measure the degree of absorption of X-rays by human tissue. Different human tissues have different CT value ranges. For example, the CT value of air is close to -1000 HU (Hounsfield Unit, which is the unit of CT value), the CT value of water is about 0 HU, and the CT value of bone is usually around 1000 HU.

[0200] By measuring the CT value of voxels, doctors can distinguish different human tissues and determine whether there are lesions. The CT value is also very important in radiotherapy planning because it is related to the absorption and scattering characteristics of radiation by tissues, affecting the calculation and distribution of radiotherapy dose. For example, the CT values ​​of tumor tissue and normal tissue may differ. The radiotherapy planning system can adjust the radiation dose based on this CT value information to achieve better treatment results while minimizing damage to normal tissues.

[0201] In one embodiment, specifically in step two, the outline is drawn based on the grayscale distribution of human organs in the CT image and consists of multiple coordinate points arranged in sequence.

[0202] CT images reveal the structure of human organs by showing the differences in the absorption of X-rays by different tissues, which are represented by different gray values. For example, bone tissue absorbs more X-rays and appears as a higher gray value (usually brighter) on CT images; while soft tissue absorbs less X-rays and has a lower gray value (usually darker). Doctors or professionals can distinguish the boundaries of different human organs based on these gray distributions and then outline their contours. For example, the gray value of a tumor target area may differ from that of the surrounding normal tissue. By carefully observing the changes in gray value, the general outline of the tumor can be delineated.

[0203] The outlined lines are stored and represented in the computer as multiple coordinate points arranged in sequence. These coordinate points determine the shape and position of the outline. Adjacent coordinate points are connected in sequence to form continuous lines, thus outlining the boundaries of human organs. The orderly arrangement of coordinate points is important because it determines the direction and closure of the outline, which is crucial for subsequent image processing operations based on the outline (such as filling algorithms to determine structural regions). If the order of coordinate points is disordered, it may lead to incorrect outline shapes, which in turn affects the accurate representation of human organs and the formulation of subsequent radiotherapy plans.

[0204] In summary, the three-dimensional lattice radiotherapy lattice arrangement method provided in this embodiment of the invention acquires CT image sequences and processes them in multiple steps. First, contour lines are loaded for each structure. Then, a grid-based filling algorithm is used to generate binary image sequences of the structures. Next, the image sequences of the tumor target volume (GTV) and the planned organ at risk (OAR_PRV) are processed based on erosion and dilation algorithms. After that, the positions of the lattice sphere centers are determined by layering and scanning. Finally, a three-dimensional lattice radiotherapy lattice arrangement scheme is planned based on the lattice sphere arrangement conditions. This method automates the process from image acquisition to lattice arrangement scheme planning, saving time and effort and improving the efficiency of radiotherapy planning. The lattice arrangement is carried out according to a unified standard, ensuring the consistency of lattice arrangement under different patients or doctors, and improving the quality and stability of radiotherapy planning. Through targeted processing of GTV and OAR_PRV and reasonable lattice sphere arrangement conditions, the lattice position is accurately planned, reducing damage to organs at risk, improving three-dimensional dose distribution, improving tumor treatment effect, and reducing the risk of complications.

[0205] Figure 5 This is a functional module diagram of the three-dimensional lattice radiotherapy lattice arrangement system according to an embodiment of this application, as shown below. Figure 5As shown, a three-dimensional lattice radiotherapy lattice arrangement system includes: an image data acquisition module, a contour loading module, a binary image generation module, an image transformation and processing module, a lattice sphere center determination module, and a lattice arrangement planning module.

[0206] The image data acquisition module is configured to acquire a CT image sequence, which consists of multiple CT image planes arranged in spatial order and parallel to each other.

[0207] The contour loading module is configured to load contour lines for each structure based on the CT image plane, and each structure contains one or more contour lines, including human organs or tumor structures.

[0208] The binary image generation module is configured with a grid-based filling algorithm to generate a binary image sequence of the same size as the CT image sequence for each structure.

[0209] The image transformation processing module is configured to process the structural binary image sequences of the tumor target volume (GTV) and the planned organ at risk volume (OAR_PRV) based on the erosion algorithm and the dilation algorithm, respectively, and generate the processed image region.

[0210] The lattice center determination module is configured to determine the position of the lattice center by performing layering and scanning on the processed image region;

[0211] The lattice layout planning module is configured to plan a three-dimensional lattice radiotherapy lattice layout scheme based on the determined lattice center position and lattice sphere layout conditions.

[0212] In one embodiment, the image transformation processing module is further configured with a data storage unit for storing intermediate data generated during the three-dimensional shrinking processing of the GTV structural binary image sequence and the two-dimensional expansion processing of the OAR_PRV structural binary image sequence, including the shrunk image sequence V. L Find the merged image sequence V. PRV .

[0213] In one embodiment, the lattice arrangement planning module includes a collision detection unit, which is configured to detect whether there is a collision or interference between adjacent lattice spheres when planning a three-dimensional lattice radiotherapy lattice arrangement scheme based on the determined lattice sphere center position and lattice sphere arrangement conditions.

[0214] In summary, the three-dimensional lattice radiotherapy lattice arrangement system provided in this embodiment of the invention acquires CT image sequences through an image data acquisition module, loads contour lines for each structure based on the CT images through a contour line loading module, generates binary image sequences of structures using a raster filling algorithm through a binary image generation module, processes relevant image sequences using erosion and dilation algorithms through an image transformation processing module and stores intermediate data through a data storage unit, determines the position of the lattice sphere center through layered scanning of the processed image region through a lattice sphere center determination module, and plans the lattice arrangement scheme based on the lattice sphere center position and arrangement conditions through a lattice arrangement planning module. The collision detection unit can also detect collision interference between lattice spheres. This system automates the process from image data processing to lattice arrangement scheme planning, improving the efficiency of radiotherapy planning; ensures the consistency of lattice arrangement, improving the quality and stability of radiotherapy plans; accurately plans lattice positions, reduces damage to organs at risk, improves three-dimensional dose distribution, enhances tumor treatment efficacy, reduces the risk of complications, and provides strong support for three-dimensional lattice radiotherapy.

[0215] For further details regarding the implementation techniques of each module in the three-dimensional lattice radiotherapy lattice arrangement system of the above embodiments, please refer to the description in the three-dimensional lattice radiotherapy lattice arrangement method of the above embodiments, which will not be repeated here.

[0216] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For system-type embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0217] The basic principles of this disclosure have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this disclosure are merely examples and not limitations, and should not be considered as essential features of each embodiment of this disclosure. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the scope of this disclosure to the necessity of employing the aforementioned specific details for implementation.

[0218] In this disclosure, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. The block diagrams of devices, apparatuses, devices, and systems involved in this disclosure are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as "comprising," "including," "having," etc., are open-ended terms meaning "including but not limited to," and are used interchangeably with them. The terms "or" and "and" as used herein refer to the terms "and / or," and are used interchangeably with them unless the context clearly indicates otherwise. The term "such as" as used herein refers to the phrase "such as but not limited to," and is used interchangeably with it.

[0219] Additionally, as used herein, the "or" used in a list of items beginning with "at least one" indicates a separate list, such that a list of, for example, "at least one of A, B, or C" means A or B or C, or AB or AC or BC, or ABC (i.e., A and B and C). Furthermore, the word "exemplary" does not imply that the described example is preferred or better than other examples.

[0220] It should also be noted that in the systems and methods of this disclosure, the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered as equivalent solutions to this disclosure.

[0221] Various changes, substitutions, and modifications can be made to the technology described herein without departing from the teachings defined by the appended claims. Furthermore, the scope of the claims of this disclosure is not limited to the specific aspects of the processes, machines, manufactures, events, means, methods, and actions described above. Currently existing or later-developed processes, machines, manufactures, events, means, methods, or actions that perform substantially the same function or achieve substantially the same result as the corresponding aspects described herein can be utilized. Therefore, the appended claims include such processes, machines, manufactures, events, means, methods, or actions within their scope.

[0222] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of this disclosure. Therefore, this disclosure is not intended to be limited to the aspects shown herein, but rather to be carried out within the widest scope consistent with the principles and novel features disclosed herein.

[0223] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this disclosure to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations therein.

Claims

1. A method for arranging crystal lattice in three-dimensional lattice radiotherapy, characterized in that, Includes the following steps: A computed tomography (CT) image sequence is acquired, the CT image sequence consisting of multiple CT image planes arranged in spatial order and parallel to each other; contour lines are loaded for each structure according to the CT image planes, each structure containing one or more contour lines, the structure including human organs or tumor structures; a grid-based filling algorithm is used to generate a structural binary image sequence of the same size as the CT image sequence for each structure; the structural binary image sequences of the tumor target volume (GTV) and the planned organ at risk volume (OAR_PRV) are processed using erosion and dilation algorithms respectively, and processed image regions are generated; the location of the lattice sphere center is determined by layering and scanning the processed image regions; Based on the determined lattice center position and lattice sphere arrangement conditions, a three-dimensional lattice radiotherapy lattice arrangement scheme is planned. The method for determining the lattice center position includes the following steps: Step 501, find V L The middle layer S0 of the foreground region is scanned in the order of x, y, z, and foreground voxels are added to the candidate list L. possible Middle; Step 502, retrieve list L possible The first voxel point in the matrix is ​​added to the lattice sphere center list L0, and from L... possible Remove from the list; Step 503, Remove candidate list L possible For all voxel points in the list L0 whose distance from the last center of the lattice spheres is less than r millimeters, where r is the minimum distance between the centers of the lattice spheres; Step 504: Repeat steps 502-503 above until list L0 is reached. possible Empty; Step 505: Using the middle layer S0 as the center, find layers with a spacing of s millimeters in both the top and bottom directions, that is, layers that satisfy S i+1 With S i The distance between the two layers is s millimeters; Step 506: Arrange S in the order of x, y, z. i Layer image scanning will be compared with the previous layer's lattice sphere center list L i-1 or L i+1 Foreground voxels with a center distance of not less than d1 mm are added to the candidate list L. possible In the context of d1, d1 represents the two-dimensional distance between the center of the lattice microspheres in each layer and the center of the microspheres in the adjacent layer; Step 507: Take list L possible The first voxel point in the array is added to the lattice sphere center list L. i In, and from L possible Remove from the list; Step 508: Remove candidate list L possible List of all lattice sphere centers L i The last voxel point with a distance of less than r millimeters between its centers; Step 509: Repeat steps 507-508 above until list L. possible Empty, and process all layers, finally from the list L of lattice centers of each layer. i The positions of the centers of all crystal lattices are obtained from this.

2. The three-dimensional lattice radiotherapy lattice arrangement method according to claim 1, characterized in that, The method for generating the processed image region includes the following steps: The structure of the GTV binary image sequence is shrunk in three dimensions by (m+d / 2) millimeters based on the erosion algorithm to obtain the shrunk image sequence V. L Where m is the distance from GTV to V L The inward distance, where d is the diameter of the lattice microsphere; Based on the dilation algorithm, two-dimensional expansion is performed on all OAR_PRV structured binary image sequences (r). P Processing the data by +d / 2) millimeters and performing a merging operation yields the merged image sequence V. PRV , where r P The distance between the lattice sphere and OAR_PRV; Find V PRV Voxels with a median value of 1, and V L Middle and V PRV and V L The voxel values ​​in the intersection are set to 0.

3. The three-dimensional lattice radiotherapy lattice arrangement method according to claim 1, characterized in that, The lattice microsphere arrangement conditions include: The diameter of the lattice microspheres is d millimeters; GTV to V L The inward distance is m millimeters; The minimum distance between the centers of the lattice spheres is r millimeters; The distance between the lattice sphere and OAR_PRV is not less than r P millimeters; The interlayer spacing is s millimeters; The two-dimensional distance between the center of the lattice microspheres in each layer and the center of the lattice microspheres in the adjacent layer is not less than d1 millimeters.

4. The three-dimensional lattice radiotherapy lattice arrangement method according to claim 1, characterized in that, The method for generating a binary image sequence of each structure with the same size as the CT image sequence includes the following steps: Step 301: Set the initial value of all pixels in the structured binary image sequence to 0; Step 302: For each contour line of each structure, start the grid-based fill algorithm. When executing the algorithm, scan the CT image pixels line by line, starting from the first line of the image. Step 303: During the scanning process, for each row of pixels, the boundary of the area to be filled is determined according to rules; the rules are formulated based on the pixel features of the image and the definition of the contour lines; Step 304: For each pixel currently scanned, determine its value based on the previously defined boundary conditions. If the pixel is inside the contour line, set its pixel value to 1; otherwise, keep its pixel value at 0. Step 305: Continue the above scanning, judgment and assignment operations until the filling process of all contour lines is completed.

5. The three-dimensional lattice radiotherapy lattice arrangement method according to claim 1, characterized in that: The CT image plane includes a set of sampling grids representing specific thicknesses, lengths, and widths in the three-dimensional human body space. Each grid is a voxel, and the voxel value represents the density distribution sampling value at the location of the voxel.

6. The three-dimensional lattice radiotherapy lattice arrangement method according to claim 1, characterized in that: The process involves loading contour lines for each structure based on the plane of the CT image. Each structure contains one or more contour lines, including human organs or tumor structures. The contour lines are drawn based on the grayscale distribution of the human organs in the CT image and consist of multiple coordinate points arranged in sequence.

7. A three-dimensional lattice radiotherapy lattice arrangement system, applied to the three-dimensional lattice radiotherapy lattice arrangement method according to any one of claims 1-6, characterized in that, The system includes: an image data acquisition module, a contour loading module, a binary image generation module, an image transformation and processing module, a lattice sphere center determination module, and a lattice arrangement planning module; The image data acquisition module is configured to acquire a CT image sequence, which consists of multiple CT image planes arranged in spatial order and parallel to each other. The contour loading module is configured to load contour lines for each structure based on the plane of the CT image, the structure including human organs or tumor structures. The binary image generation module is configured with a grid-based filling algorithm to generate a binary image sequence of the same size as the CT image sequence for each structure. The image transformation processing module is configured to process the structural binary image sequences of the tumor target volume (GTV) and the planned organ at risk volume (OAR_PRV) based on the erosion algorithm and the dilation algorithm, respectively, and generate the processed image region. The lattice center determination module is configured to determine the position of the lattice center by performing layering and scanning on the processed image region; The lattice layout planning module is configured to plan a three-dimensional lattice radiotherapy lattice layout scheme based on the determined lattice center position and lattice sphere layout conditions.

8. The three-dimensional lattice radiotherapy lattice arrangement system according to claim 7, characterized in that: The image transformation and processing module is also equipped with a data storage unit for storing intermediate data generated during the three-dimensional shrinking processing of the GTV structural binary image sequence and the two-dimensional expansion processing of the OAR_PRV structural binary image sequence, including the shrunk image sequence V. L Find the merged image sequence V. PRV .

9. The three-dimensional lattice radiotherapy lattice arrangement system according to claim 8, characterized in that: The lattice arrangement planning module includes a collision detection unit, which is configured to detect whether there is a collision or interference between adjacent lattice spheres when planning a three-dimensional lattice radiotherapy lattice arrangement scheme based on the determined lattice sphere center position and lattice sphere arrangement conditions.

Images