Radiotherapy equipment image guiding device, guiding method and related equipment

Abstract

The invention discloses a radiotherapy equipment image guiding device and method and related equipment, and relates to the technical field of radiotherapy, and the device comprises a ray generation module which is provided with a ray source and is arranged at one side of a treatment room isocenter; the radiation source is used for moving along a first preset path; the ray source is used for emitting rays, and the rays penetrate through a patient; the detection and acquisition module is provided with a detector, is arranged on the other side of the isocenter of the treatment room and is distributed opposite to the ray source along the isocenter; the detector is used for moving along a second preset path, collecting rays penetrating through the patient and converting the rays into corresponding projection images; the registration module is used for performing registration on the projection image and a planned CT image by executing a first registration mode or a second registration mode so as to calculate a positioning deviation; and the first preset path and the second preset path are straight lines, so that the adaptability, precision and stability of radiotherapy are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of radiotherapy technology, in particular to a radiotherapy device image guiding device, a guiding method and related devices. BACKGROUND

[0002] In proton radiotherapy, image guidance is a key link to ensure treatment accuracy, and its core is to realize patient positioning verification and target positioning through accurate imaging. The current mainstream image guidance methods mainly rely on X-ray and CT technology, but both have certain limitations: X-ray guidance can realize isocenter imaging and is relatively efficient, but its two-dimensional imaging characteristics make it difficult to clearly display three-dimensional anatomical structures, which is not conducive to anatomical analysis of tumors and surrounding tissues and accurate assessment of tumor size changes and registration results; CT guidance can provide three-dimensional images, support detailed analysis of tumors and surrounding tissues and high-precision registration, but its non-isocenter in-situ scanning mode requires the treatment bed to move the patient between the CT scanning position and the treatment positioning position. The mechanical movement error of the treatment bed and the autonomous / physiological movement of the patient (such as breathing, body movement) during this process may affect the efficiency and accuracy of image guidance.

[0003] To solve the above problems, some manufacturers have introduced an isocenter in-situ CBCT scanning scheme, which combines three-dimensional imaging with isocenter scanning, taking into account the stability of three-dimensional anatomical display and in-situ guidance; but the existing isocenter in-situ CBCT adopts a circular arc scanning mode, and its device structure needs to be arranged in a ring-shaped gantry around the isocenter of the treatment room. Limited by the spatial layout of some radiotherapy devices (such as room size, existing mechanical structure interference) and installation conditions, it cannot be adapted and applied on most devices, resulting in that these devices still have to rely on two-dimensional X-ray or non-isocenter CT guidance, making it difficult to balance the guidance accuracy, anatomical analysis requirements and device compatibility; and the existing isocenter in-situ CBCT cannot achieve a precise balance between imaging effect and radiation safety. SUMMARY

[0004] The purpose of the present application is to provide a radiotherapy device image guiding device, a guiding method and related devices, which can not only perform isocenter in-situ scanning, but also obtain three-dimensional imaging, greatly improving the accuracy and stability of radiotherapy; and can also realize X-ray two-dimensional imaging, efficiently realizing radiotherapy image guidance.

[0005] The purpose of the present application is achieved by adopting the following technical solutions:

[0006] In a first aspect, the present application provides a radiotherapy device image guiding device, which comprises:

[0007] A radiation generating module is configured with a radiation source and is arranged at one side of the isocenter of a treatment room; the radiation source is used to move along a first preset path; the radiation source is used to emit radiation, and the radiation passes through a patient;

[0008] A detection collecting module is configured with a detector and is arranged at the other side of the isocenter of the treatment room and is distributed opposite to the radiation source along the isocenter; the detector is used to move along a second preset path, is used to collect the radiation after passing through the patient, and is used to convert the radiation into a corresponding projection image;

[0009] A registration module is used to register the projection image with a planning CT image to calculate a positioning deviation by executing a first registration mode or a second registration mode;

[0010] The first preset path and the second preset path are both straight lines.

[0011] Preferably, the first preset path and the second preset path are both vertical straight lines along the direction of gravity; the detector moves in the opposite direction of the radiation source.

[0012] Preferably, the device acquires the projection image by the following way, which comprises:

[0013] The radiation source is moved along the first preset path until the end of the stroke, and at a preset position on the moving track, the projection image of the patient is collected by the detector;

[0014] When the radiation source moves along the first preset path, the central axis of the radiation source is always aligned with the isocenter of the treatment room; when the radiation source moves a first preset distance to reach the preset position, the detector is reversely moved along the second preset path by a second preset distance, so that the radiation beam is coplanarly aligned with the receiving surface of the detector to collect the projection image of the patient.

[0015] Preferably, the first registration mode comprises:

[0016] The radiation source is moved to a plurality of positions on the full stroke of the first preset path, and a plurality of projection images are collected;

[0017] The plurality of projection images are three-dimensionally reconstructed to generate a CBCT image;

[0018] The CBCT image is registered with the planning CT image, a positioning deviation is calculated, and parameters for correcting the treatment bed are outputted.

[0019] Optionally, the second registration mode comprises:

[0020] The projection image is collected at an arbitrary position on the moving track of the radiation source;

[0021] The projection image is registered with a DRR image generated from the planning CT image to obtain a first registration result;

[0022] According to the first registration result, a setup deviation is calculated and parameters for treatment bed correction are outputted.

[0023] Optionally, the performing the second registration step comprises:

[0024] acquiring projection images at any at least two positions on the moving trajectory of the ray source;

[0025] acquiring DRR images respectively corresponding to each of the projection images generated by the planning CT image;

[0026] respectively registering each of the projection images with the corresponding DRR image to obtain a plurality of first registration results,

[0027] According to the first registration result, a setup deviation is calculated and parameters for treatment bed correction are outputted.

[0028] In a second aspect, the present application provides an image-guided radiotherapy device, which comprises:

[0029] controlling a ray source arranged at one side of the isocenter of a treatment room to move along a first preset path, and emitting rays during the movement, wherein the rays pass through a patient;

[0030] controlling a detector arranged at the other side of the isocenter and opposite to the ray source along the isocenter to move along a second preset path, and collecting the rays after passing through the patient and converting the rays into corresponding projection images; wherein the first preset path and the second preset path are straight lines;

[0031] processing the projection images and registering the projection images with a planning CT image to calculate a setup deviation;

[0032] wherein the first preset path and the second preset path are straight lines.

[0033] Preferably, the first preset path and the second preset path are vertical straight lines along the direction of gravity; and the detector moves in the opposite direction to the ray source.

[0034] Preferably, the projection images are acquired by the following method, comprising:

[0035] moving the ray source along the first preset path until the end of the travel, and collecting the projection images of the patient by the detector at a preset position on the moving trajectory;

[0036] when the ray source moves along the first preset path, the central axis of the ray source is always aligned with the isocenter of the treatment room; when the ray source moves a first preset distance to reach the preset position, the detector is controlled to move in the opposite direction along the second preset path by a second preset distance, so that the ray beam is coplanarly aligned with the receiving surface of the detector to collect the projection images of the patient.

[0037] Optionally, the projection images are registered with the planning CT images by performing a first registration mode, including:

[0038] moving the radiation source to a plurality of positions along the first preset path, and acquiring a plurality of projection images;

[0039] performing three-dimensional reconstruction on the plurality of projection images to generate a CBCT image;

[0040] registering the CBCT image with the planning CT image, calculating a setup deviation, and outputting parameters for correction of the treatment bed.

[0041] Optionally, the projection images are registered with the planning CT images by performing a second registration mode, including:

[0042] acquiring a projection image at an arbitrary position along the trajectory of the radiation source;

[0043] registering the projection image with a DRR image generated from the planning CT image to obtain a first registration result;

[0044] calculating a setup deviation according to the first registration result, and outputting parameters for correction of the treatment bed.

[0045] Optionally, the projection images are registered with the planning CT images by performing a second registration mode, including:

[0046] acquiring a projection image at an arbitrary position along the trajectory of the radiation source;

[0047] acquiring a DRR image corresponding to each of the projection images, which is generated from the planning CT image;

[0048] registering each of the projection images with the corresponding DRR image to obtain a plurality of first registration results,

[0049] calculating a setup deviation according to the plurality of first registration results, and outputting parameters for correction of the treatment bed.

[0050] In a third aspect, the present application provides a radiotherapy system, including:

[0051] the image-guided device of any of the embodiments of the present application.

[0052] In a fourth aspect, the present application provides an electronic device including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method of any of the present application or the functions of the device of the present application when executing the computer program.

[0053] In a fifth aspect, the present application provides a computer readable storage medium storing computer instructions, when the computer instructions are read by a computer, the computer executes the steps of any one of the methods of the present application.

[0054] Compared with the prior art, the present application has at least the following beneficial effects: using an isocenter in-situ straight line CBCT scanning design, without a circular arc track, solving the problem that some radiotherapy devices cannot install a circular arc CBCT system due to space or structural limitations, while ensuring the core advantage of isocenter imaging, more models are compatible, and the application range of precise image guidance is expanded. Both three-dimensional CBCT can be collected by straight line movement to reconstruct a three-dimensional CBCT, solving the problem that traditional X-ray two-dimensional images are not conducive to anatomical analysis and tumor size change evaluation, and providing complete three-dimensional anatomical information for doctors; and can be flexibly switched to a two-dimensional imaging mode to quickly complete guidance through two position X-ray images, balancing the demand for high precision and treatment efficiency. Through a projection number decision model and a step dynamic adjustment model, the acquisition scheme is optimized based on individualized parameters such as target region characteristics and physiological motion of the patient, while ensuring the coverage of key structures and controlling the radiation dose, meeting the precise guidance needs of complex cases, and complying with the radiation protection principle.

[0055] In summary, on the basis of breaking through the device limitations, through the innovative design of "in-situ scanning + three-dimensional / two-dimensional dual-mode imaging", the short board of traditional X-ray and non-in-situ CT guidance is made up for, more reliable and flexible image guidance support is provided for radiotherapy, and finally the treatment precision and safety are improved. BRIEF DESCRIPTION OF DRAWINGS

[0056] Figure 1 is a radiotherapy device image guidance device schematic diagram of an embodiment of the present application;

[0057] Figure 2 is a position and movement direction schematic diagram of a radiotherapy device image guidance device component of an embodiment of the present application;

[0058] Figure 3 is a radiotherapy device image guidance method schematic diagram of an embodiment of the present application;

[0059] Figure 4 is a schematic diagram of a radiotherapy system of an embodiment of the present application. DETAILED DESCRIPTION

[0060] Example implementations will now be described more fully with reference to the accompanying drawings. Example implementations can be implemented in any number of manners, and are not limited to the embodiments described herein; rather, the embodiments are provided as factual examples to convey the principles of the example implementations to one of ordinary skill in the art fully and entirely, and to enable one of ordinary skill in the art to embody and practice the example implementations. Like reference numerals refer to like or similar elements throughout the several views of the drawings, and a repeated description thereof will be omitted.

[0061] The words expressing position and direction described in the present application are illustrated by taking the drawings as an example, but changes can also be made as needed, and the changes made are included in the protection scope of the present application.

[0062] Referring to the drawings Figure 1 and the drawings Figure 2 , the radiotherapy equipment image guiding device, the device comprises:

[0063] The ray generating module is configured with a ray source and is arranged on one side of the isocenter; the ray source is used for moving along a first preset path; the ray source is used for emitting a ray, and the ray passes through the patient; the camera source is an X-ray source;

[0064] The detection acquisition module is configured with a detector and is arranged on the other side of the isocenter and is distributed opposite to the ray source along the isocenter; the detector is used for moving along a second preset path, is used for acquiring the ray after passing through the patient, and is used for converting it into a corresponding projection image, i.e. an X-ray image;

[0065] The registration module is used for registering the projection image with the planning CT image to calculate the positioning deviation by executing a first registration mode or a second registration mode;

[0066] The first preset path and the second preset path are straight lines.

[0067] The first preset path and the second preset path are not on the moving path of the treatment head and are not parallel to the moving path of the treatment head; the detector and the ray source are opposite in the moving direction, the central axis of the detector and the central axis of the ray source both pass through the isocenter, and the detector and the ray source move before the treatment head moves around the patient to perform treatment.

[0068] The working principle of the above technical solution is:

[0069] If the wall of the treatment room has the conditions of bearing and positioning, the ray source and the detector can be directly installed on the wall on both sides of the isocenter, and the stability of the moving track is ensured through the wall fixation. If the wall is not suitable for installation, an independent mounting rack (such as a floor column type mounting rack or a ceiling suspension type mounting rack) can also be used. The mounting rack needs to be rigidly connected with the floor or the ceiling of the treatment room to ensure that the ray source and the detector will not be displaced due to vibration or bearing during the movement.

[0070] From the spatial distance to the isocenter, the ray source and the isocenter need to maintain a preset working distance (usually determined according to the energy and penetration ability of the ray beam and the size of the detector). This distance needs to meet the condition that the ray beam can still be effectively received and imaged clearly by the detector after passing through the isocenter region (patient target area). If the distance is too close, it may result in insufficient coverage of the ray beam. If the distance is too far, it may affect the quality of the projection image due to the attenuation of the ray.

[0071] The ray source and the detector can adopt a symmetrical distribution relative to the isocenter: that is, the straight-line distance from the ray source to the isocenter is equal to the straight-line distance from the detector to the isocenter, and the three (ray source center-isocenter-detector center) are on the same straight line. The advantage of this symmetrical distribution is that when the ray source moves along the first preset path, the detector only needs to move in the opposite direction along the second preset path by the same distance, so as to keep the coplanar alignment with the ray beam, which simplifies the moving control logic. At the same time, the symmetrical layout can ensure that the divergence angle of the ray beam after passing through the isocenter is always matched with the receiving surface of the detector at different moving positions, reducing the geometric distortion of the projection image.

[0072] Of course, in actual application, if the treatment room space is limited (such as an obstacle on one side wall), an asymmetric distribution (that is, the distances from the ray source and the detector to the isocenter are different) can also be used, but the algorithm of the moving control module needs to be compensated to ensure that the moving distance of the detector and the moving distance of the ray source form an adaptive ratio, so as to keep the coplanar alignment of the ray beam and the receiving surface, thereby balancing the installation flexibility and imaging accuracy.

[0073] In the initial state, the central axis of the ray source is aligned with the isocenter, and the receiving surface of the detector is coplanarly aligned with the ray beam, ensuring that the ray beam passing through the isocenter region (patient target area) is effectively received by the detector.

[0074] The ray source moves along the first preset straight-line path, and the central axis is always aligned with the isocenter and continuously emits rays during the movement. The detector moves in the opposite direction along the second preset straight-line path with the ray source, and keeps the receiving surface coplanarly aligned with the ray beam at the corresponding position, collects the rays passing through the patient and converts them into projection images (multiple projections can be collected at multiple discrete positions on the moving track).

[0075] The registration module processes the collected projection images (such as three-dimensional reconstruction to generate CBCT images or directly using two-dimensional projections), and registers with the planning CT image, calculates the patient positioning deviation, and finally outputs the correction parameters for the treatment bed adjustment to ensure that the particle beam accurately irradiates the target area during treatment.

[0076] The technical scheme has the following effects:

[0077] The central axis of the ray source is aligned with the isocenter throughout the movement, the detector moves along the second preset straight line path in the opposite direction of the ray source, and the receiving surface is aligned with the ray beam at the corresponding position, so that the isocenter image is collected in situ, the treatment bed error caused by the back-and-forth movement of the patient in non-isocenter CT guidance and the interference of the patient's autonomous movement are avoided, and the accuracy and stability of image guidance are significantly improved.

[0078] The ray source and the detector move along a straight line path (rather than a circular arc scan), the mechanical structure is simple and occupies less space, and it can be adapted to radiotherapy equipment that cannot install a circular arc CBCT system due to space or structural limitations, thereby widening the application scenarios of isocenter three-dimensional image guidance technology.

[0079] Supporting multiple projection image acquisition and three-dimensional reconstruction (CBCT), providing three-dimensional anatomical structure information, solves the problem of difficult anatomical analysis in two-dimensional X-ray guidance; at the same time, it can be compatible with a two-dimensional projection mode of double positions, meet the guidance needs of different treatment scenarios, and has high flexibility.

[0080] The straight line scanning path can flexibly adjust the number and angle distribution of projection acquisition, which can not only ensure the accuracy of three-dimensional reconstruction through sufficient projection, but also avoid unnecessary redundant projection, reduce the radiation dose of the patient, and comply with the principle of minimum radiation in radiotherapy; the straight line movement mechanism is easy to integrate with existing radiotherapy equipment, thereby reducing the modification and maintenance costs.

[0081] In a possible implementation manner, the device further includes:

[0082] A movement control module configured to drive the movement of the ray source and the detector and the angular rotation of the ray source.

[0083] The movement control module is the core driving and coordinating unit of the device, and its functions can be refined into a three-layer architecture of accurate driving, real-time sensing, and intelligent adaptation, and in a possible implementation manner, specifically include:

[0084] A driving execution component: configured with a high-precision linear motor (to drive the movement of the ray source along the first preset path and the movement of the detector along the second preset path) and a servo rotary motor (to drive the rotation of the ray source around the central axis to adjust the emission angle), the motor control accuracy reaches the level of 0.1 mm (movement) and 0.1° (rotation), ensuring the accuracy of the motion response.

[0085] Sensing feedback components: integrated grating ruler (real-time monitoring of the actual displacement of the radiation source and the detector), angle encoder (recording the rotation angle of the radiation source) and laser positioner (assisting in calibrating the eccentricity alignment deviation), the feedback data sampling frequency is not less than 1 kHz, which provides real-time basis for closed-loop control.

[0086] Control algorithm unit: built-in motion coordination algorithm (ensuring the synchronization of the reverse movement of the radiation source and the detector, the time difference is less than or equal to 5 ms), angle compensation algorithm (dynamically calculating the required rotation angle according to the movement distance of the radiation source, ensuring that the central axis is always aligned with the eccentric center), and error correction algorithm (based on the real-time adjustment of the motor output based on the sensing feedback data, compensating for mechanical clearance or external disturbance).

[0087] The execution steps of the movement control module include:

[0088] Receiving the acquisition instructions (such as the number of projections, position distribution) of the registration module, combining the preset path parameters (length, direction) to plan the movement trajectory of the radiation source, the reverse trajectory of the detector and the rotation angle sequence of the radiation source, for example, when the radiation source moves 100mm along the vertical path, it needs to rotate θ angle to offset the central axis deviation caused by the displacement, θ is calculated by the geometric relationship θ = arctan (movement distance / source to eccentric center distance).

[0089] Driving execution: according to the planning results, sending driving signals to the linear motor and the rotary motor to control the radiation source to move along the first path, the detector to move reversely along the second path, and the radiation source to rotate to the target angle at the same time, ensuring that the central axis of the ray bundle always passes through the eccentric center.

[0090] Real-time feedback and correction: during the movement, the grating ruler and the angle encoder continuously collect actual displacement and angle data, if the deviation from the planned value exceeds the threshold (such as displacement deviation > 0.5mm, angle deviation > 0.5°), the control algorithm immediately adjusts the motor output (such as increasing / decreasing the driving torque), until the deviation returns to the allowed range.

[0091] Linkage triggers acquisition: when the radiation source and the detector reach the preset acquisition position and the movement deviation is stable within the threshold, send a trigger signal to the detection acquisition module to start the radiation emission and projection image acquisition simultaneously, ensuring the spatial position accuracy of each projection.

[0092] In one possible implementation, the first preset path and the second preset path are parallel to each other.

[0093] The parallelism of the two paths ensures that the moving directions of the ray source and the detector are always consistent (e.g., both along the vertical direction or both along a specific horizontal direction), providing a geometric basis for the reverse synchronous movement of the two. When the ray source moves along the first path, the detector moves in the opposite direction along the second path. Due to the parallelism of the paths, the relative displacement of the two is always in the same plane, and synchronization can be achieved through simple linear control (e.g., the ray source moves up by x distance, and the detector moves down by x distance), avoiding complex coordinate conversion calculations in the case of non-parallel paths, and significantly reducing the algorithm complexity of the movement control module.

[0094] From the perspective of ray beam alignment, the parallel paths, combined with the constraint that the central axis of the ray source is always aligned with the isocenter, can ensure that the divergence direction of the ray beam and the relative angle of the detector receiving surface remain stable. For example, when the ray source moves along a vertical path parallel to the direction of gravity, the central axis of the cone-shaped ray beam it emits is always directed at the isocenter, while the detector moves in the opposite direction along the parallel vertical path, and can always be coplanar with the divergence plane of the ray beam. This stability avoids the "tilt misalignment" of the ray beam and the detector receiving surface caused by non-parallel paths, reduces the geometric distortion of the projection image, and provides higher quality raw data for subsequent three-dimensional reconstruction (CBCT) or two-dimensional registration (X-ray).

[0095] In addition, the design of parallel straight paths makes the device more easily adaptable to the spatial layout of different treatment rooms. Compared with circular arc paths, the mechanical structure (e.g., guide rails, sliding blocks) of straight paths is simpler, and the path length can be flexibly adjusted (e.g., shortened or lengthened) according to the site size, making it particularly suitable for space-limited radiotherapy equipment. At the same time, the movement range of the ray source and the detector can be independently set (while maintaining the reverse synchronous relationship) in the case of parallel paths, which can not only meet the scanning needs of patients with large tumors (by increasing the path length to expand the imaging range), but also reduce unnecessary movement during scanning of small target areas (such as head and neck tumors) by shortening the path, further improving operational efficiency.

[0096] Referring to the drawings Figure 2 In one possible implementation, the first preset path and the second preset path are parallel to each other and are both vertical straight lines along the direction of gravity; the detector moves in the opposite direction to the ray source.

[0097] On the vertical movement trajectory of the ray source and the detector, projection acquisition is triggered at multiple discrete height positions (e.g., every 10 mm or according to clinical needs). Since the paths are along the direction of gravity, the patient's body naturally relaxes due to gravity, reducing the body position shaking that may occur during horizontal movement; at the same time, the multi-position acquisition in the vertical direction can cover different anatomical levels (e.g., the upper and lower lung lobes of the chest cavity, the liver area of the abdominal cavity, and the pelvic cavity), providing projection images of the full target area for three-dimensional reconstruction.

[0098] The vertical movement along the gravity direction eliminates the need for the patient to bear additional body position adjustment force, and the body naturally adheres to the treatment bed, reducing the self or non-self movement (such as the horizontal deviation of the body caused by breathing) that may occur during horizontal movement, especially suitable for patients with chest and abdominal tumors, and reducing the projection image error caused by body position change.

[0099] The vertical linear path can directly use the up-down space (such as the height from the ceiling to the floor) of the treatment room to arrange the guide rail, without reserving a large area of movement space in the horizontal direction, and the mechanical structure is more compact. For a space-limited radiotherapy room (such as a scene with sufficient layer height but limited lateral width), this design can maximize the use of existing space and reduce the cost of site modification during equipment installation.

[0100] The vertical movement along the gravity direction ensures that the relative positions of the ray source and the detector are always distributed along the longitudinal axis of the human body, and the path of the ray beam passing through the patient is more consistent with the natural trend of the anatomical structure (such as the vertical level from head to foot). Compared with the horizontal path, this layout can reduce the scattering interference when the ray beam passes through high-density tissues such as bones and cavities, and further reduce the projection geometry distortion caused by path deviation by cooperating with the reverse synchronous movement, thereby improving the target area clarity of three-dimensional reconstruction.

[0101] For tumors in different parts of the head and neck (small vertical range) and the body (large vertical range), the vertical path movement range can be flexibly adapted by adjustment (such as shortening the movement distance for head and neck scanning and lengthening the distance for body scanning), which can ensure sufficient coverage of the key target area and avoid time waste caused by invalid movement, and both imaging efficiency and radiation safety are considered.

[0102] In one possible implementation, the device acquires the projection image by the following method, comprising:

[0103] The ray source is moved along the first preset path until the end of the stroke, and at the preset position on the movement track, the projection image (X-ray image) of the patient is acquired by the detector;

[0104] When the ray source moves along the first preset path, the central axis of the ray source is always aligned with the isocenter of the treatment room, and when the ray source moves a first preset distance to reach a first preset position, the detector is synchronously or asynchronously moved in a second preset path in a reverse direction by a second preset distance, so that the ray beam is coplanar with the receiving surface of the detector for alignment, and the projection image of the patient is acquired; the first preset distance is equal to or not equal to the second preset distance.

[0105] X-ray source and detector are distributed along first and second preset straight line paths respectively (such as vertical parallel paths or parallel paths with other angles), and initial positions satisfy that the central axis of the X-ray source is aligned with the isocenter of the treatment room (the central axis of the X-ray beam is always coincident with the isocenter, or the central axis of the X-ray beam is always deviated from the isocenter by not more than a deviation threshold), so as to ensure that the X-ray beam divergence range needs to cover the isocenter and the treatment area where the patient is located; the receiving surface of the detector is aligned with the X-ray beam, so as to ensure that the X-ray projection image passing through the patient object is completely acquired. According to the path space layout (such as whether the distances from the X-ray source and the detector to the isocenter are symmetrical), the adaptive relationship between the first preset distance (the movement distance of the X-ray source) and the second preset distance (the reverse movement distance of the detector) is set, if the distances from the X-ray source and the detector to the isocenter are equal (symmetrical layout), the distances are set to be equal; if the distances are not equal (asymmetrical layout), the distance ratio is set through geometric conversion. The part of the patient to be treated is located on the isocenter.

[0106] On the movement trajectory of the X-ray source, when reaching the preset discrete positions (such as the starting point, the midpoint, the end point of the path or the key height set according to clinical needs), the acquisition instruction is triggered: the X-ray source emits X-rays passing through the patient, and the detector synchronously receives the X-rays and converts them into projection images, that is, X-ray images; due to the distance coordination mechanism adapting to the path layout, even in an asymmetrical path, the X-ray beam at each acquisition position can accurately cover the isocenter area, and the detector receives without deviation, ensuring the consistency of multiple projections in the spatial coordinate system.

[0107] In a possible implementation, according to the distribution range of the target region and the key anatomical structure of the patient, the total travel of the first preset path is determined to ensure that the movement of the X-ray source along the movement direction from the starting position to the end position can make the X-ray beam completely cover the target region;

[0108] The position of the end of the travel of the X-ray source is set in coordination with the parameters of the radiotherapy plan, the characteristics of the patient and the target region and the physical constraints of the equipment;

[0109] The parameters of the radiotherapy plan: according to the three-dimensional coordinate range (such as the maximum / minimum coordinate value in the Z-axis direction) of the target region in the planning CT image, the basic travel boundary is extended outward by 5-10 mm;

[0110] The characteristics of the patient and the target region: in combination with the patient's body type (body thickness, height) and the target region movement range (such as the tumor displacement amplitude caused by respiration), the basic travel boundary is corrected, and the correction amount is 1-2 times the maximum displacement of the target region;

[0111] The physical constraints of the equipment: taking the mechanical travel limit of the equipment (such as the maximum range of vertical movement of the X-ray source) as the upper limit, to ensure that the set end of the travel does not exceed the safe running interval of the equipment.

[0112] In one possible implementation, the registration module performs a first registration mode (three-dimensional registration mode), reconstructing a three-dimensional CBCT image through multi-position projection and then registering it with the planned CT scan, including:

[0113] The X-ray source is moved to multiple locations along the entire first preset path to acquire multiple projection images;

[0114] CBCT images are generated by 3D reconstruction of multiple projection images;

[0115] The CBCT images are registered with the planned CT images, the positioning deviation is calculated, and parameters for treatment bed correction are output.

[0116] The working principle of the above technical solution is as follows:

[0117] The radiation source moves along a first preset path, acquiring projection images at multiple discrete locations (the number dynamically adjusted according to clinical needs, such as 10-30). These locations cover the entire path, ensuring that radiation attenuation information of the patient's target area and surrounding tissues is obtained from different angles (based on the perspective of the straight-line trajectory formed by the movement of the radiation source).

[0118] The acquired projection images are first preprocessed, including noise removal (through Gaussian filtering or median filtering), correction of radiation hardening effect (correction of high-energy radiation attenuation deviation based on a preset calibration curve), and unification of spatial coordinates (establishing a coordinate system with the center of the treatment room as the origin and marking the position and angle parameters of the radiation source corresponding to each projection), to provide high-quality raw data for subsequent reconstruction.

[0119] Three-dimensional reconstruction is performed on multiple preprocessed projection images, for example, using filtered back projection (FBP) or iterative reconstruction algorithms (such as ART and SIRT). During reconstruction, the projection images at different angles are back-projected onto a three-dimensional voxel grid, using the isocenter as a spatial reference. The density value of each voxel is calculated by weighting its corresponding ray attenuation value in all projection images, ultimately generating a three-dimensional CBCT image consistent with the patient's actual anatomical structure (resolution typically reaching 0.5-1 mm). Because the projection images originate from isocenter in-situ scanning, the reconstructed CBCT image is spatially aligned with the treatment location, avoiding positional deviations caused by non-isocenter scanning.

[0120] The registration process uses the planned CT images as a reference (including clinical markers such as the tumor target area and organs at risk), and can achieve accurate matching through the following steps:

[0121] Coarse registration is performed using image grayscale features (such as bone density differences and soft tissue contours) or anatomical markers (such as vertebral body edges and metal markers) to quickly align the approximate positions of CBCT and planning CT images, reducing the computational load for subsequent fine registration.

[0122] A rigid registration algorithm (for rigid structures such as the head and neck) or an elastic registration algorithm (for deformable regions such as the chest and abdomen) is employed. By iteratively optimizing the translation (X, Y, Z axes) and rotation (pitch, roll, rotation) parameters in three-dimensional space, the differences in anatomical structure edges and density distribution deviations between CBCT and planning CT images are minimized. Specifically, the normalized mutual information (measuring the correlation of gray-level distributions between the two images) reaches its maximum value, and the mean square error (measuring the density difference of corresponding voxels) reaches its minimum value. The final registration result must meet clinical acceptance criteria, such as a target area center overlap error ≤1mm and an organ contour overlap ≥95%, which serves as a quantitative indicator of the "optimal" registration state.

[0123] Deviation verification: The deviation of the target area center is manually checked by the doctor or automatically detected to ensure that the registration result meets clinical requirements (usually the error should be ≤1mm).

[0124] After registration, the module calculates the three-dimensional positioning deviation (including translation and rotation parameters) of the patient's current position relative to the planned CT image and converts it into adjustment commands for the treatment bed (such as moving +2mm along the X-axis and rotating -0.5° around the Y-axis), which are then output to the treatment control system. After the treatment bed completes the correction according to the commands, the patient's target area is precisely aligned with the irradiation position in the radiotherapy plan, ensuring accurate delivery of the particle beam during treatment.

[0125] The entire process relies on 3D data acquired in situ at the isocenter, achieving spatial consistency throughout the entire process from imaging to registration. This effectively eliminates positional errors from non-isocenter scanning and significantly improves the accuracy and efficiency of setup verification.

[0126] In one possible implementation, the registration module performs a second registration mode (i.e., a two-dimensional registration mode), directly registering the projected image with the DRR image generated by the planned CT scan, including:

[0127] Projected images are acquired at any position along the trajectory of the moving X-ray source;

[0128] The projected image is registered with the DRR image generated from the planned CT image to obtain the first registration result;

[0129] Based on the first registration result, calculate the positioning deviation and output the parameters for correction of the treatment bed.

[0130] In one possible implementation, the first registration result includes the similarity between the projected image and the DRR image (such as mutual information, normalized cross-correlation coefficient, etc.);

[0131] The patient's positioning deviation is determined based on the similarity, and parameters for correction of the treatment bed are output; the positioning deviation includes 6 degrees of freedom deviation (including translation of the X, Y, and Z axes and rotation around the three axes).

[0132] In one possible implementation, determining the patient's positioning deviation based on the similarity includes:

[0133] Establish a mapping relationship between the combination of placement parameters and the similarity, wherein the combination of placement parameters includes translation parameters and / or rotation parameters;

[0134] Search for the optimal combination of placement parameters in the mapping relationship and determine it as the placement deviation.

[0135] In this process, a mapping relationship is established by iteratively adjusting the combination of placement parameters (translation and rotation parameters) and calculating the corresponding similarity.

[0136] The above technical solution offers the following advantages: it supports the acquisition of projection images at any position along the trajectory of the radiation source, eliminating the need for fixed angles and adapting to various radiotherapy scenarios (such as emergency positioning or special target locations); by iteratively adjusting positioning parameters and establishing a parameter-similarity mapping, it can accurately locate the parameter combination that best matches the projection image with the DRR, ensuring the accuracy of the 6-DOF positioning deviation calculation; the two-dimensional registration process based on a single projection image is simple and faster than three-dimensional reconstruction (such as CBCT), reducing patient positioning time and lowering the risk of radiation exposure; the output 6-DOF deviation can directly drive mainstream treatment beds to complete multi-dimensional position corrections, adapting to common clinical needs for precise radiotherapy (such as head and neck, thoracic and abdominal tumors).

[0137] In one possible implementation, the registration module performs a second registration mode, including:

[0138] Projected images are acquired at at least two locations along the trajectory of the radiation source.

[0139] Obtain DRR images corresponding to each of the projected images generated from the planned CT images;

[0140] Each of the projected images is registered with its corresponding DRR image to obtain multiple first registration results.

[0141] Based on multiple first registration results, the setup deviation is calculated and parameters for treatment bed correction are output.

[0142] By using projected images from at least two different locations, a richer spatial perspective (such as frontal, lateral, and oblique views) is covered, overcoming the limitations of single-view projection in three-dimensional spatial information and reducing directional ambiguity in positioning deviation calculations. Multiple sets of registration results mutually constrain each other, reducing registration errors caused by noise, artifacts, or overlapping anatomical structures in a single image, thus improving the reliability of 6-DOF positioning deviations. Through multi-view consistency verification (such as eliminating abnormal registration results), misjudgments from a single view are avoided, making it particularly suitable for scenarios with complex anatomical structures around the target area (such as pelvic and skull base tumors). It supports flexible selection of radiation source positions and allows for customized acquisition schemes based on target area characteristics (such as long strip target areas requiring multi-directional verification), adapting to the precise radiotherapy needs of tumors in multiple locations such as the head and neck, chest and abdomen.

[0143] In one possible implementation, each of the first registration results includes the similarity between the corresponding projected image and the DRR image;

[0144] The multiple similarities are fused to determine the final similarity.

[0145] The patient's positioning deviation is determined based on the final similarity, and parameters for treatment bed correction are output.

[0146] The fusion process for the multiple similarities includes at least one of averaging, weighted fusion, peak-first fusion, or gradient fusion.

[0147] In the weighted fusion, the weight of each similarity is determined based on at least one of the spatial coverage of the corresponding projected image, noise level, or clinical target area relevance.

[0148] In one possible implementation, the spatial coverage of the image is determined by the distance between the radiation source and the treatment isocenter, as well as the relative relationship between the projection direction and the tomographic direction of the planned CT image. The closer the radiation source is to the isocenter, the higher the spatial coverage of the corresponding projected image and the greater its weight. The more parallel (rather than orthogonal) the projection direction is to the tomographic direction of the planned CT image, the higher the spatial coverage and the greater the weight of the projected image. The relevance of the clinical target area is determined by the principle of priority projection of the target area boundary and coverage of high-risk organs.

[0149] In practical applications, the treatment isocenter of radiotherapy is the core reference point for focusing X-rays and planning CT images (such as the tumor center). When the X-ray source moves to a projection position close to the isocenter, the spatial overlap between the anatomical structures (such as tumors and bones) in the projected image and the planned CT image is the highest.

[0150] For example, in a planned CT image, the tumor in a lung cancer patient is located in the middle of the chest cavity (where the isocenter is set). The X-ray source moves along a vertical path. When it reaches the isocenter height, the projected coronal image of the chest clearly shows the positional relationship between the tumor and the left and right hilum and ribs, with almost the same viewpoint as the planned CT image, resulting in high spatial coverage. Conversely, if the X-ray source moves to the top of the path (away from the isocenter, such as above the head), the projected image will look down on the chest cavity, compressing the tumor and deforming the ribs, resulting in a significant difference in spatial relationship from the planned CT image and low spatial coverage.

[0151] The planned CT image is a tomographic image of the patient in a supine position (e.g., a transverse axial view). If the projected image is a coronal / sagittal projection acquired by the X-ray source in a vertical direction, it will form an orthogonal complement to the transverse axial view of the planned CT image, which can verify the left-right / up-down offset of the positioning. If the projection direction is completely parallel to the planned CT image (e.g., the X-ray source moves horizontally to acquire a transverse axial projection), it will be closer to the spatial coverage of the planned CT image and have a higher degree of overlap with the anatomical structures.

[0152] Clinical target relevance describes the ability of projected images to display the target area (tumor) and organs at risk (normal tissue). It can directly determine whether positioning deviations lead to incorrect radiotherapy doses. Clinical target relevance is determined by prioritizing target boundary projection and covering high-risk organs.

[0153] Target boundary projection priority:

[0154] If the projected image happens to pinpoint the critical location between the target area and the organ at risk, this type of projection has extremely high clinical relevance.

[0155] For example, in radiotherapy for nasopharyngeal carcinoma, the target area is close to the brainstem (an organ at risk). When the radiation source is moved to a lateral projection position (e.g., the radiation source is on the left side of the patient, and a sagittal projection is acquired), the projection image can clearly show the lateral boundary between the nasopharyngeal tumor (target area) and the brainstem. Once the positioning is off, the change in the distance between the tumor and the brainstem in the projection directly corresponds to whether the brainstem will be over-irradiated → clinical relevance is maximized.

[0156] Conversely, if the projection only shows the center of the tumor (far from organs at risk), even if the positioning is off, it is difficult to directly determine whether normal tissue is damaged → low clinical relevance.

[0157] High-risk organ coverage:

[0158] During radiotherapy, certain organs are extremely sensitive to radiation (such as the spinal cord, eyeballs, and kidneys); if the projection image can completely cover these high-risk organs and the target area, it will have higher clinical relevance.

[0159] For example, in breast cancer radiotherapy, the target area is in the chest, and the organ at risk is the heart. When the radiation source is moved to a forward and lateral projection position, the projected image can simultaneously show the anterior-posterior positional relationship between the breast tumor (target area) and the heart (organ at risk). → When the position shifts, it can be directly observed whether the tumor has moved or whether the heart is covered by radiation. → This is something that clinicians must pay attention to, as it has a high correlation.

[0160] For example: the outline boundary of the target area (such as a tumor) is identified by the planned CT image. When the overlap length between the target area outline and the planned CT image outline in the projected image is greater than or equal to a preset ratio (e.g., 80%) of the target area perimeter, it is determined to be the target area boundary projection and assigned a first weight (e.g., 1.5, and a normal weight is, for example, 1).

[0161] For example, the location of organs at risk (such as the spinal cord and eyeballs) in the planned CT images is pre-marked. When the minimum distance between the organ at risk and the target area in the projected image is less than or equal to the distance threshold (e.g., ≤5mm clinical high-risk threshold), it is judged as a high-risk coverage projection and assigned a second weight (e.g., 1.3). If the projection simultaneously satisfies the target area boundary + high-risk coverage, it is assigned a third weight (e.g., 1.6).

[0162] The effects of the above technical solution are as follows:

[0163] By averaging and weighted fusion of multiple similarities, spatial information from multi-view projection images can be integrated, avoiding registration errors caused by noise, artifacts, or viewpoint limitations in single projections. For example, the complementary information between orthogonal and parallel projections, combined with the rule that "the closer the X-ray source is to the isocenter, the higher the weight," can prioritize the use of accurate information from high-coverage projections, reducing positioning errors caused by misjudgments from a single viewpoint, and making the 6-DOF deviation calculation closer to the patient's actual position.

[0164] Determining the relevance weight of clinical target volumes based on the principles of prioritizing target boundary projection and covering high-risk organs can significantly enhance attention to the tumor-at-risk organ boundary. For example, in nasopharyngeal carcinoma radiotherapy, a high weighting of lateral projection can directly warn of potential brainstem damage caused by positioning deviations; in breast cancer radiotherapy, a weighting bias towards anterior lateral projection can accurately verify whether the heart has been over-irradiated, thus mechanistically reducing the risk of excessive dose to normal tissues.

[0165] Spatial coverage weighting combines the distance from the isocenter and the parallelism of the projection direction with the planned CT image. This ensures the core position of high-overlap projections near the isocenter and enhances the deviation verification dimension through supplementary information from orthogonal projection. Clinical relevance weighting is standardized through quantitative thresholds (such as 80% overlap of target area contours and ≤5mm of high-risk distance) to avoid subjective judgment errors and make weight allocation more scientific and reproducible.

[0166] Multiple fusion methods (such as peak-priority fusion to remove outliers and gradient fusion to highlight sensitive parameters) can address the characteristics of different radiotherapy sites: for pelvic tumors with complex anatomical structures, multi-view weighted fusion can balance bone interference and target area display; for skull base tumors adjacent to high-risk organs, projection with high clinical relevance weights can prioritize locking key boundaries and ensure that the positioning correction focuses on the deviation direction with the highest clinical risk.

[0167] The relative relationship between the projection direction and the tomographic direction of the planned CT image is incorporated into the weight calculation, making the registration process more consistent with the original imaging logic of the planned CT image—parallel projection ensures the overlap of the basic structure, and orthogonal projection supplements the three-dimensional space verification, reducing the systematic error of cross-modal registration.

[0168] In one possible implementation, determining the patient's positioning deviation based on the final similarity includes:

[0169] By combining the ray source angles corresponding to each projection image, a mapping relationship between the combination of placement parameters and the final similarity is established;

[0170] The optimal combination of placement parameters corresponding to the final similarity in the mapping relationship is searched using a multidimensional optimization algorithm, and this combination is determined as the placement deviation.

[0171] The multidimensional optimization algorithm includes at least one of particle swarm optimization algorithm or simulated annealing algorithm.

[0172] In one possible implementation, before processing the plurality of similarities, a consistency check is performed on the plurality of similarities to remove outliers.

[0173] Combining the X-ray source angles (i.e., projection viewpoints) corresponding to each projection image, the 6-DOF positioning parameter combination (X, Y, Z axis translation and rotation around the three axes) is used as input. By calculating and fusing the similarities between multiple projection images and corresponding DRR images under this parameter combination, the final similarity is obtained, forming a mapping relationship between positioning parameter combination and final similarity (parameter space with multi-angle constraints). Using multi-dimensional optimization algorithms such as particle swarm optimization or simulated annealing, the positioning parameter combination that results in the optimal (e.g., maximum) final similarity is searched in the above mapping relationship and identified as the patient's positioning deviation. Before fusing multiple similarities, outliers (e.g., similarities deviating from the mean by more than a preset threshold) are identified and removed through statistical analysis (e.g., calculating the mean and standard deviation) to ensure the reliability and consistency of the similarities involved in the fusion. Multidimensional optimization algorithms can efficiently handle the coordinated optimization of 6-DOF positioning parameters and solve the complex problems caused by the coupling of translation and rotation parameters (such as rotational deviation may be accompanied by translational deviation in head and neck tumor positioning). At the same time, the algorithm's flexibility can adapt to the parameter accuracy requirements of different target areas (such as intracranial tumors being more sensitive to rotational deviation), and personalized optimization can be achieved by adjusting the search accuracy.

[0174] In one possible implementation, the number of projected images (number of location points) is determined through dynamic adaptation of multi-dimensional clinical and individual characteristics; including:

[0175] A projection quantity decision model is constructed. The inputs of the projection quantity decision model include tumor anatomical location (such as head and neck, body), planned target volume, historical registration error trend, patient physiological movement amplitude and registration pattern (such as breathing, voluntary movement);

[0176] The projection quantity decision model dynamically allocates projection density based on a preset anatomical structure coverage priority (e.g., target area > organs at risk > normal tissue). The result satisfies the following conditions: the minimum projection density required for coverage of key anatomical structures, and the patient's radiation dose is limited to a safe threshold.

[0177] In one possible implementation, the first preset distance is determined in the following manner:

[0178] Based on the number of projection position points output by the projection number decision model, the total journey of the first preset path is divided into intervals that match the number of projection position points in the direction of movement, and the length of each interval is the basic step size.

[0179] The basic step size is adapted and optimized using a dynamic step size adjustment model. The inputs of the dynamic step size adjustment model include:

[0180] The patient's structural-dose characteristics and physiological-motor characteristics; the structural-dose characteristics include the volume and shape complexity of the target area, its spatial proximity to organs at risk, and the radiation sensitivity level of normal tissues outside the target area; the physiological-motor characteristics include respiratory amplitude, voluntary movement frequency, and range.

[0181] The output of the step size dynamic adjustment model is the specific step size value of each interval on the first preset path, wherein: for the interval corresponding to the target area, the model outputs the reduction magnitude according to the structure-dose characteristics (e.g., the smaller the target area volume, the more complex the shape, or the closer it is to the organ at risk, the greater the reduction magnitude).

[0182] Based on the structural-dose characteristics, physiological motion characteristics, and corresponding optimal step size allocation cases of a large number of patients, a dynamic step size adjustment model is constructed using machine learning and other methods, enabling it to generalize and adapt to different patient characteristics. For the current patient, individualized structural-dose characteristics (such as self-target volume, normal tissue sensitivity, etc.) and real-time physiological motion characteristics are extracted, input into the pre-trained model, and the output is a step size adjustment scheme specific to the patient.

[0183] The step size dynamic adjustment model and the projection quantity decision model are related but independent models. The former is based on the total number of position points output by the latter to achieve fine allocation of step size in each interval. The first preset distance formed in the end satisfies the minimum projection density of key anatomical structures, and the patient's radiation dose is limited within a safe threshold, while balancing registration accuracy and acquisition efficiency.

[0184] The effects of the above technical solution are as follows:

[0185] By integrating individualized structural-dose characteristics and physiological motion characteristics of patients through a step-size dynamic adjustment model, the step-size allocation of the radiation source movement is highly adapted to the specific situation of the patient. This ensures high-resolution projection coverage of key structures such as the target area while avoiding redundant acquisition of non-critical areas, thereby improving the accuracy of image guidance.

[0186] The step size is dynamically adjusted based on the radiation sensitivity level of normal tissues outside the target area (maintaining a smaller step size increment when sensitivity is high and increasing the step size when sensitivity is low). This ensures that the overall radiation dose to the patient is strictly controlled within a safe threshold while meeting image quality requirements, reducing the risk of damage to normal tissues. The step size is also adjusted in real time based on the patient's physiological movement characteristics (such as respiratory amplitude and range of voluntary movement). When the movement amplitude is large, the step size is reduced to ensure registration accuracy; when the movement amplitude is small, the step size is increased to shorten acquisition time. This achieves an optimal balance between accuracy and efficiency, improving the smoothness of the treatment process.

[0187] The step length dynamic adjustment model, trained based on a large amount of patient data, has generalization and adaptation capabilities, covering patients with different anatomical features and movement states. At the same time, it outputs a personalized step length plan for the current patient, taking into account both group patterns and individual differences, thereby improving the clinical applicability of the device.

[0188] The interconnected yet independent division of labor (the former determines the total amount, while the latter provides fine-grained allocation) ensures that the overall layout of the projection points meets the minimum density requirements for covering key structures, while also achieving optimal resource allocation through step size optimization, thereby enhancing the intelligence and efficiency of the image guidance device.

[0189] See attached document Figure 3 This invention provides an image-guided method for radiotherapy equipment, the method comprising:

[0190] A radiation source positioned on one side of the center of the treatment room is controlled to move along a first preset path, emitting radiation during the movement, and the radiation passes through the patient;

[0191] The detector, which is located on the other side of the center of the treatment room and is distributed relative to the radiation source along the center, moves along a second preset path to collect the radiation after passing through the patient and convert it into a corresponding projection image; wherein, both the first preset path and the second preset path are straight lines.

[0192] The projected image is processed and registered with the planned CT image to calculate the positioning deviation;

[0193] Both the first preset path and the second preset path are straight lines.

[0194] In one possible implementation, the movement of the X-ray source and detector, as well as the angular rotation of the X-ray source, are driven by a motion control module.

[0195] In one possible implementation, the first preset path and the second preset path are parallel to each other.

[0196] In one possible implementation, both the first and second preset paths are vertical straight lines along the direction of gravity; the detector moves in the opposite direction to the radiation source.

[0197] In one possible implementation, the projected image is obtained in the following ways:

[0198] The X-ray source is moved along the first preset path until the end of the journey, and at the preset position on the moving trajectory, the patient's projected image is acquired by the detector.

[0199] When the radiation source moves along the first preset path, the central axis of the radiation source is always aligned with the center of the treatment room. When the radiation source moves a first preset distance to the first preset position, the detector is moved in the opposite direction along the second preset path by a second preset distance, either synchronously or asynchronously, so that the radiation beam and the receiving surface of the detector are aligned together to collect the patient's projected image.

[0200] In one possible implementation, the total travel of the first preset path is determined based on the distribution range of the patient's target area and key anatomical structures, ensuring that the movement of the radiation source from the starting position to the ending position along the direction of movement enables the radiation beam to completely cover the target area; the position of the end of the radiation source travel is set collaboratively by radiotherapy planning parameters, patient and target area characteristics, and equipment physical constraints.

[0201] In one possible implementation, the registration of the projected image with the planned CT image is achieved by performing a first registration mode, including:

[0202] The X-ray source is moved to multiple locations along the entire first preset path to acquire multiple projection images;

[0203] CBCT images are generated by 3D reconstruction of multiple projection images;

[0204] The CBCT images are registered with the planned CT images, the positioning deviation is calculated, and parameters for treatment bed correction are output.

[0205] In one possible implementation, the registration of the projected image with the planned CT image is achieved by performing a second registration mode, including:

[0206] Projected images are acquired at any position along the trajectory of the moving X-ray source;

[0207] The projected image is registered with the DRR image generated from the planned CT image to obtain the first registration result;

[0208] Based on the first registration result, calculate the positioning deviation and output the parameters for correction of the treatment bed.

[0209] In one possible implementation, the first registration result includes the similarity between the projected image and the DRR image (such as mutual information, normalized cross-correlation coefficient, etc.);

[0210] The patient's positioning deviation is determined based on the similarity, and parameters for correction of the treatment bed are output; the positioning deviation includes 6 degrees of freedom deviation (including translation of the X, Y, and Z axes and rotation around the three axes).

[0211] In one possible implementation, determining the patient's positioning deviation based on the similarity includes:

[0212] Establish a mapping relationship between the combination of placement parameters and the similarity, wherein the combination of placement parameters includes translation parameters and / or rotation parameters;

[0213] Search for the optimal combination of placement parameters in the mapping relationship and determine it as the placement deviation.

[0214] In this process, a mapping relationship is established by iteratively adjusting the combination of placement parameters (translation and rotation parameters) and calculating the corresponding similarity.

[0215] In one possible implementation, the registration of the projected image with the planned CT image is achieved by performing a second registration mode, including:

[0216] Projected images are acquired at at least two locations along the trajectory of the radiation source.

[0217] Obtain DRR images corresponding to each of the projected images generated from the planned CT images;

[0218] Each of the projected images is registered with its corresponding DRR image to obtain multiple first registration results.

[0219] Based on multiple first registration results, the setup deviation is calculated and parameters for treatment bed correction are output.

[0220] In one possible implementation, each of the first registration results includes the similarity between the corresponding projected image and the DRR image;

[0221] The multiple similarities are fused to determine the final similarity.

[0222] The patient's positioning deviation is determined based on the final similarity, and parameters for treatment bed correction are output.

[0223] The fusion process for the multiple similarities includes at least one of averaging, weighted fusion, peak-first fusion, or gradient fusion.

[0224] In the weighted fusion, the weight of each similarity is determined based on at least one of the spatial coverage of the corresponding projected image, noise level, or clinical target area relevance.

[0225] In one possible implementation, the spatial coverage of the image is determined by the distance between the radiation source and the treatment isocenter, as well as the relative relationship between the projection direction and the tomographic direction of the planned CT image. The closer the radiation source is to the isocenter, the higher the spatial coverage of the corresponding projected image and the greater its weight. The more parallel (rather than orthogonal) the projection direction is to the tomographic direction of the planned CT image, the higher the spatial coverage and the greater the weight of the projected image. The relevance of the clinical target area is determined by the principle of priority projection of the target area boundary and coverage of high-risk organs.

[0226] In one possible implementation, each of the first registration results includes the similarity between the corresponding projected image and the DRR image;

[0227] The multiple similarities are fused to determine the final similarity.

[0228] The patient's positioning deviation is determined based on the final similarity, and parameters for treatment bed correction are output.

[0229] The fusion process for the multiple similarities includes at least one of averaging, weighted fusion, peak-first fusion, or gradient fusion.

[0230] In the weighted fusion, the weight of each similarity is determined based on at least one of the spatial coverage of the corresponding projected image, noise level, or clinical target area relevance.

[0231] In one possible implementation, the spatial coverage of the image is determined by the distance between the radiation source and the treatment isocenter, as well as the relative relationship between the projection direction and the tomographic direction of the planned CT image. The closer the radiation source is to the isocenter, the higher the spatial coverage of the corresponding projected image and the greater its weight. The more parallel (rather than orthogonal) the projection direction is to the tomographic direction of the planned CT image, the higher the spatial coverage and the greater the weight of the projected image. The relevance of the clinical target area is determined by the principle of priority projection of the target area boundary and coverage of high-risk organs.

[0232] In one possible implementation, the number of projected images (number of location points) is determined through dynamic adaptation of multi-dimensional clinical and individual characteristics; including:

[0233] A projection quantity decision model is constructed. The inputs of the projection quantity decision model include tumor anatomical location (such as head and neck, body), planned target volume, historical registration error trend, patient physiological movement amplitude and registration pattern (such as breathing, voluntary movement);

[0234] The projection quantity decision model dynamically allocates projection density based on a preset anatomical structure coverage priority (e.g., target area > organs at risk > normal tissue). The result satisfies the following conditions: the minimum projection density required for coverage of key anatomical structures, and the patient's radiation dose is limited to a safe threshold.

[0235] In one possible implementation, the first preset distance is determined in the following manner:

[0236] Based on the number of projection position points output by the projection number decision model, the total journey of the first preset path is divided into intervals that match the number of projection position points in the direction of movement, and the length of each interval is the basic step size.

[0237] The basic step size is adapted and optimized using a dynamic step size adjustment model. The inputs of the dynamic step size adjustment model include:

[0238] The patient's structural-dose characteristics and physiological-motor characteristics; the structural-dose characteristics include the volume and shape complexity of the target area, its spatial proximity to organs at risk, and the radiation sensitivity level of normal tissues outside the target area; the physiological-motor characteristics include respiratory amplitude, voluntary movement frequency, and range.

[0239] The output of the step size dynamic adjustment model is the specific step size value of each interval on the first preset path.

[0240] The working principle and effect of the above technical solution are the same as those in the embodiments of the present invention, and will not be repeated here.

[0241] This invention also provides a radiotherapy system, including an image guidance device for radiotherapy equipment as described in any of the embodiments of this invention. The radiotherapy system includes a rotating gantry 605 installed in a treatment room 612, a particle accelerator 604 installed on the rotating gantry 605 for providing a particle beam, a treatment bed 606 for supporting a patient 607 and moving the patient to an isocentric position, and a treatment head 621 for treating the patient's treatment area. The treatment head 621 rotates along a rotation direction 615 on a movement path or track 611, and a particle beam is emitted from the end of the treatment head to treat the patient. The image guidance device for radiotherapy equipment is installed on a pair of spaced-apart vertical walls 613 in the treatment room 612, and the particle accelerator may be a proton accelerator.

[0242] This invention also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of any of the methods or the functions of the devices described in this invention.

[0243] This invention also provides a computer-readable storage medium for storing a computer program. When the computer program is executed, it implements the steps of the method in this invention. The specific implementation method is consistent with the implementation method and the technical effect achieved in the above method embodiments, and some contents will not be repeated.

[0244] In this invention, a readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. The program product can take the form of any combination of one or more readable media. A readable medium can be a readable signal medium or a readable storage medium. A readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0245] Computer-readable storage media may include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The readable storage medium may also be any readable medium capable of sending, propagating, or transmitting a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, or any suitable combination thereof. Program code for performing operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar programming languages. The program code may be executed entirely on a user computing device, partially on an associated device, as a standalone software package, partially on a user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing devices can be connected to user computing devices via any type of network, including local area networks (LANs) or wide area networks (WANs), or they can be connected to external computing devices (e.g., via the Internet using an Internet service provider).

[0246] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the invention without departing from the principles and spirit of the invention, and all such changes should fall within the protection scope of the claims of the present invention.

Claims

1. An image guidance device for radiotherapy equipment, characterized in that, The device includes: A radiation generating module, equipped with a radiation source, is positioned on one side of the center of a treatment room, etc.; the radiation source is used to move along a first preset path; the radiation source is used to emit radiation, and the radiation passes through the patient; The detection and acquisition module is equipped with a detector, which is located on the other side of the center of the treatment room and is distributed relative to the radiation source along the center; the detector is used to move along a second preset path to collect the radiation after passing through the patient and convert it into a corresponding projection image. The registration module is used to register the projected image with the planned CT image by executing a first registration mode or a second registration mode to calculate the positioning deviation. The first and second preset paths are both straight lines and parallel to each other. The first and second preset paths are not on the movement path of the treatment head and are not parallel to the movement path of the treatment head. The detector moves in the opposite direction to the radiation source. The central axis of the detector and the central axis of the radiation source both pass through the center of the treatment room. The detector and the radiation source move before the treatment head moves around the patient to perform treatment.

2. The image guidance device for radiotherapy equipment according to claim 1, characterized in that, Both the first preset path and the second preset path are vertical straight lines along the direction of gravity.

3. The image guidance device for radiotherapy equipment according to claim 1, characterized in that, The device acquires the projected image in the following ways: The X-ray source is moved along the first preset path until the end of the journey, and at the preset position on the moving trajectory, the patient's projected image is acquired by the detector. When the radiation source moves along the first preset path, the central axis of the radiation source is always aligned with the center of the treatment room. When the radiation source moves a first preset distance to its preset position, the detector is moved in the opposite direction along the second preset path by a second preset distance so that the radiation beam and the receiving surface of the detector are aligned so as to collect the patient's projected image.

4. The image guidance device for radiotherapy equipment according to claim 3, characterized in that, The first registration mode includes: The X-ray source is moved to multiple locations along the entire first preset path to acquire multiple projection images; CBCT images are generated by 3D reconstruction of multiple projection images; The CBCT images are registered with the planned CT images, the positioning deviation is calculated, and parameters for treatment bed correction are output.

5. The image guidance device for radiotherapy equipment according to claim 3, characterized in that, The second registration mode includes: Projected images are acquired at any position along the trajectory of the moving X-ray source; The projected image is registered with the DRR image generated from the planned CT image to obtain the first registration result; Based on the first registration result, calculate the positioning deviation and output the parameters for correction of the treatment bed.

6. The image guidance device for radiotherapy equipment according to claim 3, characterized in that, The second registration mode includes: Projected images are acquired at at least two locations along the trajectory of the radiation source. Obtain DRR images corresponding to each of the projected images generated from the planned CT images; Each of the projected images is registered with its corresponding DRR image to obtain multiple first registration results. Based on multiple first registration results, the setup deviation is calculated and parameters for treatment bed correction are output.

7. An image-guided method for radiotherapy equipment, characterized in that, The method includes: A radiation source positioned on one side of the center of the treatment room is controlled to move along a first preset path, emitting radiation during the movement, and the radiation passes through the patient; The detector, which is located on the other side of the center of the treatment room and is distributed relative to the radiation source along the center, moves along a second preset path to collect the radiation after passing through the patient and convert it into a corresponding projection image; wherein, both the first preset path and the second preset path are straight lines. The projected image is processed and registered with the planned CT image to calculate the positioning deviation; Both the first preset path and the second preset path are straight lines.

8. The image-guided method for radiotherapy equipment according to claim 7, characterized in that, Both the first and second preset paths are vertical straight lines along the direction of gravity; the detector moves in the opposite direction to the radiation source.

9. The image-guided method for radiotherapy equipment according to claim 7, characterized in that, Projected images can be obtained in the following ways: The X-ray source is moved along the first preset path until the end of the journey, and at the preset position on the moving trajectory, the patient's projected image is acquired by the detector. When the radiation source moves along the first preset path, the central axis of the radiation source is always aligned with the center of the treatment room. When the radiation source moves a first preset distance to its preset position, the detector is moved in the opposite direction along the second preset path by a second preset distance so that the radiation beam and the receiving surface of the detector are aligned so as to collect the patient's projected image.

10. The image-guided method for radiotherapy equipment according to claim 7, characterized in that, The method includes: The X-ray source is moved to multiple locations along the entire first preset path to acquire multiple projection images; CBCT images are generated by 3D reconstruction of multiple projection images; The CBCT image is registered with the planning CT image, the positioning deviation is calculated and the parameters for treatment bed correction are output, thereby executing the first registration mode to achieve the registration of the projected image with the planning CT image.

11. The image-guided method for radiotherapy equipment according to claim 7, characterized in that, The method includes: Projected images are acquired at any position along the trajectory of the moving X-ray source; The projected image is registered with the DRR image generated from the planned CT image to obtain the first registration result; Based on the first registration result, the positioning deviation is calculated and parameters for treatment bed correction are output, thereby executing the second registration mode to achieve registration of the projected image with the planned CT image.

12. The image-guided method for radiotherapy equipment according to claim 7, characterized in that, The method includes: Projected images are acquired at at least two locations along the trajectory of the radiation source. Obtain DRR images corresponding to each of the projected images generated from the planned CT images; Each of the projected images is registered with its corresponding DRR image to obtain multiple first registration results, thereby executing a second registration mode to achieve registration between the projected images and the planned CT images. Based on multiple first registration results, the setup deviation is calculated and parameters for treatment bed correction are output.

13. A radiotherapy system, characterized in that, include: The image guidance device for radiotherapy equipment according to any one of claims 1-6.

14. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform the functions of the device according to any one of claims 1-6 or the steps of the method according to any one of claims 7-12.

15. A computer-readable storage medium, characterized in that, The storage medium stores computer instructions, and when the computer reads the computer instructions, the computer performs the steps of the method as described in any one of claims 7-12.

Images