Trajectory optimization using dose estimation and conflict detection

Abstract

The system and method (200) for radiation treatment planning may include a computational system (150) that determines (202) an estimate of the radiation dose distribution within an anatomical region (302) of a patient, and uses the estimate of the radiation dose distribution to determine (204) a cost matrix representing an objective function. The computational system (150) may project (206) the cost matrix onto each of a plurality of fluence planes. Each of the plurality of fluence planes may be associated with a corresponding gantry-treatment bed orientation among a plurality of gantry-treatment bed orientations of a medical linear accelerator. The computational system (150) may use the projection of the cost matrix onto each of the plurality of fluence planes to determine a sequence of gantry-treatment bed orientations representing treatment paths among the plurality of gantry-treatment bed orientations.

Description

Technical Field

[0001] This application generally relates to systems and methods for automated radiation therapy treatment planning. Specifically, this application relates to automated radiation therapy treatment planning using estimated radiation dose distribution and conflict detection, such as regarding organs at risk (OARs). Background Technology

[0002] Radiation therapy is a radiation-based therapy used to treat cancer. Specifically, high doses of radiation are used to kill or shrink tumors. The target area in a patient's body intended to receive radiation (such as a tumor) is called the planned target volume (PTV). The goal is to deliver enough radiation to the PTV to kill cancer cells. However, other organs or anatomical regions adjacent to or surrounding the PTV may be in the path of the radiation beam and may receive enough radiation to damage or harm these organs or anatomical regions. These organs or anatomical regions are called organs at risk (OARs). Typically, physicians or radiologists identify both the PTV and OAR before radiation therapy using images such as computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET), images obtained via some other imaging modality, or a combination thereof. For example, a physician or radiologist may manually mark the PTV and / or OAR on a patient's medical images.

[0003] Using the patient's medical images and identified PTV and OAR, a team of medical professionals (e.g., internists, radiologists, oncologists, radiation technicians, other medical professionals, or a combination thereof) (referred to herein as the treatment planner) determines the radiation parameters to be used during radiotherapy treatment. These radiation parameters include, for example, the type, angle, intensity, and / or shape of each radiation beam. In determining these parameters, the treatment planner attempts to achieve a radiation dose distribution to be delivered to the patient that meets, for example, predefined criteria set by the team. Such criteria typically include predefined radiation dose thresholds or ranges to be met for the PTV and OAR.

[0004] To optimize radiation parameters in a way that meets predefined criteria, treatment planners typically run multiple simulations with various radiation parameters and select a final set of parameters to use based on the simulation results. This process often involves adjusting the radiation parameters after each simulation. This type of approach is time-consuming and tedious, and may not provide optimal results. For example, a patient may have to wait days or weeks before a radiation therapy plan for that patient is ready. Summary of the Invention

[0005] In a first aspect of the invention, a method for planning radiation treatment according to claim 1 is provided.

[0006] In a second aspect of the invention, a radiation treatment planning system according to claim 11 is provided.

[0007] In a third aspect of the invention, a computer-readable medium according to claim 20 is provided.

[0008] Optional features are defined in the dependent claims.

[0009] The embodiments described herein relate to an automated trajectory planning method for radiation treatment planning. Using medical images of a patient's anatomical structures, a computing device estimates the expected radiation dose distribution or a typical achievable dose distribution within anatomical regions of the patient's body. Given an objective function defined based on the estimated distribution and dosimetric targets for PTV and OAR, the computing device computes a cost matrix representing the objective function defined based on the dose estimation and projects the cost matrix onto an available fluence plane. The computing device can use the projection of the cost matrix to detect conflicts with predefined medical targets (e.g., related to the amount of acceptable radiation dose in various organs of the patient) and identify radiation trajectories based on the identified conflicts. For each possible orientation of the radiator and treatment bed of the radiation machine, a corresponding value defined based on the corresponding projection of the cost matrix is ​​used as a metric to determine whether that orientation belongs to the final radiation trajectory.

[0010] According to one aspect, a method for radiation treatment planning may include one or more processors determining an estimate of the radiation dose distribution within an anatomical region of a patient. The method may include one or more processors using the estimate of the radiation dose distribution to determine a cost matrix representing an objective function. The objective function may be defined based on the estimate of the radiation dose distribution and patient-specific data. The method may include one or more processors projecting the cost matrix onto each of a plurality of fluence planes. Each of the plurality of fluence planes may be associated with a corresponding gantry-treatment bed orientation among a plurality of gantry-treatment bed orientations of a medical linear accelerator. The method may include one or more processors using the projection of the cost matrix onto each of the plurality of fluence planes to determine a sequence of gantry-treatment bed orientations representing treatment paths among the plurality of gantry-treatment bed orientations.

[0011] In some implementations, determining an estimate of the radiation dose distribution may include determining the estimate based on the planned target volume (PTV) from the anatomical region. The objective function may reflect one or more radiation constraints for the patient. The objective function may be defined as optimizing a radiation plan for intensity-modulated radiation therapy (IMRT). The objective function may be defined as optimizing a radiation plan for volume-modulated arc therapy (VMAT). The method may include determining multiple gantry-treatment bed orientations by discretizing the space of possible gantry-treatment bed orientations. Each point in the discrete space of possible gantry-treatment bed orientations may represent the corresponding gantry-treatment bed orientation among the multiple gantry-treatment bed orientations.

[0012] In some implementations, projecting the cost matrix onto each of a plurality of fluence planes may include applying a weighted projection. Applying a weighted projection may include weighting the projected values ​​of the cost matrix according to the depth of the planned target volume (PTV) relative to the anatomical region in the direction of the radiation beam. Determining the sequence of gantry-treatment bed orientations may include, for each gantry-treatment bed orientation, calculating a corresponding matrix sum representing the sum of projections of the cost matrix onto the target mask of the fluence plane associated with the gantry-treatment bed orientation, and using the matrix sums calculated for multiple gantry-treatment bed orientations to determine the sequence of gantry-treatment bed orientations. Determining the sequence of gantry-treatment bed orientations may include minimizing the total of the matrix sums along the processing path. The processing path may extend over a predefined range of gantry-treatment bed orientations.

[0013] According to another aspect, a radiation treatment planning system may include one or more processors and a memory for storing computer code instructions. When the computer code instructions are executed, the one or more processors may determine an estimate of the radiation dose distribution over the anatomical region of the patient. The one or more processors may use the estimate of the radiation dose distribution to determine a cost matrix representing an objective function. The objective function may be defined based on the estimate of the radiation dose distribution and patient-specific data. The one or more processors may project the cost matrix onto each of a plurality of fluence planes. Each of the plurality of fluence planes may be associated with a corresponding gantry-treatment bed orientation among a plurality of gantry-treatment bed orientations of a medical linear accelerator. The one or more processors may use the projection of the cost matrix onto each of the plurality of fluence planes to determine a sequence of gantry-treatment bed orientations representing treatment paths among the plurality of gantry-treatment bed orientations.

[0014] In some implementations, determining an estimate of the radiation dose distribution may include determining the estimate based on the distance to the planned target volume (PTV) from the anatomical region. The objective function may reflect one or more radiation constraints for the patient. The objective function may be defined as optimizing a radiation plan for intensity-modulated radiotherapy (IMRT) or optimizing a radiation plan for volume-modulated arc therapy (VMAT). One or more processors may also determine multiple gantry-treatment bed orientations by discretizing the space of possible gantry-treatment bed orientations. Each point in the discrete space of possible gantry-treatment bed orientations may represent the corresponding gantry-treatment bed orientation among the multiple gantry-treatment bed orientations.

[0015] In some implementations, when projecting the cost matrix onto each of a plurality of fluence planes, one or more processors may apply weighted projection. When applying weighted projection, one or more processors may weight the projected values ​​of the cost matrix according to the depth of the planned target volume (PTV) relative to the anatomical region in the direction of the radiation beam. When determining the sequence of gantry-treatment bed orientations, one or more processors may (i) compute a corresponding matrix sum for each gantry-treatment bed orientation, representing the sum of projections of the cost matrix onto the target mask of the fluence plane associated with the gantry-treatment bed orientation, and (ii) use the matrix sums computed for the plurality of gantry-treatment bed orientations to determine the sequence of gantry-treatment bed orientations. When determining the sequence of gantry-treatment bed orientations, one or more processors may minimize the sum of matrix sums along the processing path. The processing path may extend over a predefined range of gantry-treatment bed orientations.

[0016] According to another aspect, the computer-readable medium may include computer code instructions stored thereon. When executed, the computer code instructions cause one or more processors to determine an estimate of the radiation dose distribution over an anatomical region of the patient, and use the estimate of the radiation dose distribution to determine a cost matrix representing an objective function defined based on the estimate of the radiation dose distribution and patient-specific data. Execution of the computer code instructions causes one or more processors to project the cost matrix onto each of a plurality of fluence planes. Each of the plurality of fluence planes may be associated with a corresponding gantry-treatment bed orientation among a plurality of gantry-treatment bed orientations of a medical linear accelerator. One or more processors may use the projection of the cost matrix onto a target mask of each of the plurality of fluence planes to determine a sequence of gantry-treatment bed orientations representing a treatment path among the plurality of gantry-treatment bed orientations. Attached Figure Description

[0017] Figure 1AA block diagram illustrating an example computer environment for implementing the methods and processes described herein, according to one embodiment, is shown.

[0018] Figure 1B A block diagram illustrating one implementation of a system architecture according to one embodiment.

[0019] Figure 2 This is a flowchart illustrating an embodiment of a radiation treatment planning method according to one embodiment.

[0020] Figures 3A to 3C An illustration is shown with respect to one embodiment. Figure 2 The simulation results associated with each step of the method.

[0021] Figure 4A and 4B An image depicting a visual illustration of the projection of a cost matrix according to one embodiment is shown.

[0022] Figure 5A and 5B Images illustrating a processing path according to one embodiment are shown in two-dimensional (2-D) and three-dimensional (3-D) representations.

[0023] For illustrative purposes, some or all of the accompanying drawings are schematic representations. The foregoing information and the following detailed description include illustrative examples of various aspects and embodiments, and provide an overview or framework for understanding the nature and characteristics of the claimed aspects and embodiments. The accompanying drawings provide illustration and further understanding of various aspects and embodiments, and are incorporated into and constitute a part of this specification. Detailed Implementation

[0024] The following is a more detailed description of various concepts and their implementations related to methods, apparatus, and systems used for radiation treatment planning. Since the concepts described are not limited to any particular implementation, the various concepts introduced above and discussed in more detail below can be implemented in any of a variety of ways. Examples of specific implementations and applications are provided primarily for illustrative purposes.

[0025] Radiation therapy treatment planning is a complex and patient-specific optimization problem. Given a patient's anatomy, such as that shown in medical images, and the identification or masking of the PTV and OAR, the goal is to determine a treatment path (or trajectory) that meets predefined criteria or constraints, such as those predefined by a physician, radiologist, or other medical professional. During a radiation therapy session, the patient typically lies on a treatment bed in a radiation machine, and a gantry equipped with radiation sources rotates around the patient to deliver radiation from different angles at various intensities and / or shapes. Determining the treatment path or trajectory involves determining a sequence of positions of the radiation sources (e.g., relative to the patient) and corresponding radiation angles (e.g., in 3-D space), which defines the location and orientation of the radiation sources emitting the radiation beam toward the patient. The sequence of radiation source positions defines the rotational path or trajectory of the gantry around the patient. Determining the radiation path may also include determining the corresponding radiation intensity and / or beam shape for each radiation position and angle in the sequence of radiation positions and angles.

[0026] Optimizing radiation treatment trajectories or pathways can improve the dosimetric quality of treatment plans. Specifically, the goal of optimization is to minimize the radiation dose to the organ of origin (OAR) (or maintain it below a corresponding predefined upper limit) while maximizing the radiation dose to the organ of tissue vesicle (PTV) (or maintaining it above a corresponding predefined lower limit). In such cases, radiation therapy designed according to optimized radiation treatment trajectories can kill cancer cells without damaging or harming critical organs or OARs. Trajectory optimization methods based on manually selecting and prioritizing critical organs make the task difficult for treatment planners, time-consuming, requiring trial and error, and the results often depend on the experience and skill of the treatment planner.

[0027] In the present disclosure, systems and methods for improved automated radiation treatment planning begin with an estimation of the expected radiation dose distribution within an anatomical region of a patient's body, and the identification of conflicts between the estimated radiation dose distribution and the planned clinical objectives. The systems and methods described herein determine or optimize treatment trajectories or paths by considering spatial regions where conflicts are expected to occur. The severity of the conflict can be represented by an objective (or cost) function that can be evaluated at each voxel within the 3-D anatomical region. The systems and methods described herein can generate conflict-avoiding radiation treatment trajectories or paths based on the severity of the conflict (e.g., as expressed or described in the cost function).

[0028] The embodiments described herein allow for automated trajectory planning. Thus, the user does not need to select or adjust the weights of critical organs when generating gantry-treatment-bed orientation quality-landscapes for processing pathfinding. Furthermore, the embodiments described herein provide finer spatial accuracy (e.g., more precise than per structure) because the weighting of the 3-D patient at each step of the method described herein can be applied at the voxel level. For example, in some cases, clinical goals may require avoiding only specific areas of critical organs, rather than avoiding the entire organ. This finer spatial accuracy results in an improved final processing trajectory relative to patient-specific clinical or dose-volume goals.

[0029] Figure 1A An example computer environment 100 is illustrated, which can be used to provide a network-based implementation of the methods described herein. Computer environment 100 may include a computer server 110a, a system database 110b, a user computing device 120, and electronic data sources 130a-e (collectively referred to as electronic data sources 130). These components may be interconnected via a network 140. Examples of network 140 may include, but are not limited to, private or public LANs, WLANs, MANs, WANs, and the Internet. Network 140 may include wired and wireless communications according to one or more standards and / or via one or more transmission media.

[0030] Communication on network 140 can be performed according to various communication protocols, such as Transmission Control Protocol and Internet Protocol (TCP / IP), User Datagram Protocol (UDP), and IEEE communication protocols. In one example, network 140 may include wireless communication according to the Bluetooth specification set or another standard or proprietary wireless communication protocol. In another example, network 140 may also include communication via cellular networks, including, for example, GSM (Global System for Mobile Communications), CDMA (Code Division Multiple Access), and EDGE (Evolution of Global System for Mobile Communications) networks.

[0031] The computer environment 100 is not necessarily limited to the components described herein, and may include additional or alternative components not shown for the sake of brevity, which are considered to be within the scope of the embodiments described herein.

[0032] In some implementations, computer server 110a may be configured to execute computer instructions to perform any of the methods or operations described herein. Computer server 110a may generate and display an electronic platform to display information indicating or relating to a radiation plan trajectory. The electronic platform may include a graphical user interface (GUI) displayed on user computing device 120. Examples of electronic platforms generated and hosted by computer server 110a may be web-based applications or websites configured to be displayed on various electronic devices, such as mobile devices, tablets, personal computers, etc. (e.g., user computing device 120).

[0033] Computer server 110a can host websites accessible to end users, where the content presented via various web pages can be controlled based on each specific user's role or viewing permissions. Computer server 110a can be any computing device including a processor and non-transitory machine-readable storage capable of performing the various tasks and processes described herein. Non-limiting examples of such computing devices include workstation computers, laptop computers, server computers, and the like. While computer environment 100 includes a single computer server 110a, in some configurations, computer server 110a may include any number of computing devices operating in a distributed computing environment.

[0034] Computer server 110a can execute software applications configured to display an electronic platform (e.g., a hosted website) that generates and provides various web pages to each user computing device 120. Different users operating the user computing devices(s) 120 can use the website to view the processing traces or paths of the output and / or interact with them.

[0035] In some implementations, computer server 110a may be configured to require user authentication based on a set of user authorization credentials (e.g., username, password, biometrics, encryption certificate, etc.). In such implementations, computer server 110a may access a system database 110b configured to store user credentials, and computer server 110a may be configured to refer to these user credentials to determine whether a set of input credentials (purely for authenticating a user) matches an appropriate set of credentials that identify and authenticate the user.

[0036] In some configurations, computer server 110a can generate and host web pages based on specific user roles (e.g., administrator, employee, and / or bidder). In such implementations, user roles can be defined by data fields and input fields in user records stored in system database 110b. Computer server 110a can authenticate users and identify user roles by executing access directory protocols (e.g., LDAP). Computer server 110a can generate web page content tailored to user roles defined by user records in system database 110b.

[0037] In some embodiments, computer server 110a receives medical images, masks, and / or medical data indicating medical objectives from a user (or retrieved from a data repository); computer server 110a processes the data and displays indications of the processing trajectory on an electronic platform. For example, in a non-limiting example, a user operating computing device 130a uploads a series of CT scan images or other medical images using an electronic platform. Computer server 110a may determine the processing trajectory based on the input data and display the results via an electronic platform on user computing device 120 or computing device 130a. User computing device 120 and / or computing device 130a may be any computing device including a processor and a non-transitory machine-readable storage medium capable of performing the various tasks and processes described herein. Non-limiting examples of network nodes may be workstation computers, laptop computers, tablet computers, and server computers. In operation, various users may use user computing device 120 and / or computing device 130a to access a GUI operatively managed by computer server 110a.

[0038] Electronic data source 130 may represent various electronic data sources that contain and / or retrieve patient medical images. For example, database 130b and third-party server 130c may represent data sources that provide computer server 110a with a collection of data (e.g., medical images, masks, or other medical data) required to determine the processing trajectory. Computer server 110a may also retrieve data directly from medical scanner 130e and / or medical imaging device 130d (e.g., CT scanner).

[0039] In some implementations, the methods or operations described herein may be implemented by user equipment 120, any electronic device 130, or a combination thereof.

[0040] Although Figure 1A A network-based implementation is shown, but it should be noted that the methods described herein can be implemented by a single computing device that receives the patient’s medical images and medical data and determines the radiation treatment trajectory or path according to the methods described herein.

[0041] refer to Figure 1B Based on the currently disclosed inventive concept, a block diagram is shown depicting one embodiment of a system architecture for a computing system 150 that can be used to implement the methods described herein. The computing system 150 may include a computing device 152. The computing device 152 may represent... Figure 1A Example implementations of any of devices 110a, 120, and / or 130a-e. For example, computing device 152 may include, but is not limited to, a computed tomography (CT) scanner, a medical linear accelerator device, a desktop computer, a laptop computer, a hardware computer server, a workstation, a personal digital assistant, a mobile computing device, a smartphone, a tablet computer, or other types of computing devices. Computing device 152 may include one or more processors 154 that execute computer code instructions, a memory 156, and a bus 158 communicatively coupling the processors 154 and the memory 156.

[0042] One or more processors 154 may include a microprocessor, a general-purpose processor, a multi-core processor, a digital signal processor (DSP) or a field-programmable gate array (FPGA), an application-specific integrated circuit (ASIC), or other types of processor. One or more processors 154 may be communicatively coupled to a bus 158 for processing information. Memory 156 may include a main memory device 160, such as random access memory (RAM) or other dynamic storage device, coupled to the bus 158, for storing information and instructions to be executed by the processor 154. The main memory device 160 may be used to store temporary variables or other intermediate information during the execution of instructions by the processor 154 (e.g., in relation to methods such as method 200 described herein). Computing device 152 may include a read-only memory (ROM) 162 or other static storage device coupled to the bus 158 for storing static information and instructions for the processor 154. For example, ROM 162 may store, for example, medical images of a patient received as input. ROM 162 may store computer code instructions related to or representing implementations of the methods described herein. Storage device 164, such as a solid-state device, disk, or optical disk, may be coupled to bus 158 for storing (or providing as input) information and / or instructions.

[0043] Computing device 152 may be communicatively coupled to input device 166 and / or output device 168, or may include input device 166 and / or output device 168. Computing device 102 may be coupled to output device 168 via bus 158. Output device 168 may include a display device for displaying information to a user, such as a liquid crystal display (LCD), thin-film transistor LCD (TFT), organic light-emitting diode (OLED) display, LED display, electronic paper display, plasma display panel (PDP), or other display. Output device 168 may include a communication interface for transmitting information to other external devices. Input device 166, such as a keyboard including alphanumeric keys and other keys, may be coupled to bus 158 for transmitting information and command selection to processor 154. In another embodiment, input device 166 may be integrated into a display device, such as a touchscreen display. Input device 166 may include cursor controls, such as a mouse, trackball, or cursor arrow keys, for transmitting directional information and command selection to processor 154, and for controlling cursor movement on the display device.

[0044] According to various embodiments, the methods or corresponding operations described herein can be implemented as an arrangement of computer code instructions executed by the processor(s)154(s) of computing system 150. The arrangement of computer code instructions may be read into main memory device 160 from another computer-readable medium, such as ROM 162 or storage device 164. Execution of the arrangement of computer code instructions stored in main memory device 160 may cause computing system 150 to perform the methods or operations described herein. In some embodiments, one or more processors 154 in a multiprocessor arrangement may be used to execute computer code instructions representing embodiments of the methods or processes described herein. In some other embodiments, hardwired circuitry may be used in place of or in combination with software instructions to implement illustrative embodiments of the methods or operations described herein. Generally, the implementation is not limited to any particular combination of hardware circuitry and software. The functional operations described herein may be implemented in other types of digital electronic circuitry, computer software, firmware, hardware, or combinations thereof.

[0045] Figure 2A flowchart illustrating an embodiment of a radiation treatment planning method 200 according to the inventive concept of this disclosure is shown. Method 200 may include a computing system 150 or device determining an estimate of the radiation dose distribution within a patient (step 202), and determining a cost matrix representing an objective function (step 204). Method 200 may include the computing system 150 or computing device 152 projecting the cost matrix onto each of a plurality of fluence planes or corresponding masks (step 206), and determining a processing path based on the projection of the cost matrix onto the plurality of fluence planes or corresponding masks (step 208).

[0046] Return to reference Figure 1B and Figure 2 Method 200 may include a computing system 150 or computing device 152 determining an estimate of the radiation dose distribution within the patient's body (step 202). The computing system 150 may acquire medical images of anatomical regions of the patient, one or more structural masks of the PTV and OAR, information indicating clinical goals, or a combination thereof. A CT scanner, MRI device, ultrasound imaging device, other types of medical imaging devices, or a combination thereof may generate medical images of the patient. Medical images may include 3D images, 2D images, or a combination thereof. Acquiring one or more masks may include receiving masks from another computing device. In some other embodiments, the computing system 150 may segment one or more medical images of the patient and generate masks(s) using the segmented images. A user may label segmented regions of the medical images as corresponding to a PTV or OAR. For each of the PTV and OAR, information indicating clinical goals may include a corresponding radiation dose threshold, a corresponding radiation dose range, or a corresponding expected radiation dose value. In some embodiments, one or more structural masks may be combined to define the radiation dose threshold, range, or expected value. The computing system 150 may receive information indicating clinical goals as input via input device 166.

[0047] An estimate of the radiation dose distribution can represent the expected radiation dose distribution in response to the radiation therapy to be performed within an anatomical region of the patient's body, or a typical achievable dose distribution. The dose distribution estimate does not necessarily have to be an optimal radiation dose distribution. In some embodiments, the computing system 150 or processor 154 can generate an estimate of the radiation dose distribution based on the distance from the PTV to model the typical attenuation of the radiation dose around the PTV. This estimate can be isotropic for all directions originating from the PTV. In some embodiments, the computing system 200 can generate an estimate of the radiation dose distribution as follows:

[0048]

[0049] Where d0 is a constant, and d represents the distance from the PTV surface. The variable x represents a point or voxel in 3-D space, and α represents a coefficient that can be equal to or defined relative to the prescription dose of the PTV. The computational system 150 can use some other functions defined based on the distance d(x) to generate an estimate of the radiation dose distribution.

[0050] refer to Figure 3A The image shows a 2-D slice of an example estimate of the radiation dose distribution along with anatomical region 302. Anatomical region 302 includes PTV 304 and two OARs 306 and 308. Figure 3A The estimate of the radiation dose distribution shown is defined as D(x), as shown in equation (1). Figure 3A As shown, the radiation dose function D(x) is significantly reduced outside of PTV 304.

[0051] When a PTV comprises multiple non-overlapping regions (e.g., multiple tumors or abnormalities), the calculation system 150 can generate or define an estimate of the radiation dose distribution based on the distance from each PTV region. For example, at each voxel x, the calculation system 150 can evaluate expressions for various distances from different PTV regions. (or some other distance function), and if voxel x is outside any PTV area, the maximum value is used as the radiation dose D(x). If voxel x is within the PTV area, the calculation system 150 or processor 154 may use the maximum value of the distance evaluation function as the radiation dose D(x).

[0052] Refer again Figure 1B and Figure 2 Method 200 may include a computing system 150 or processor(s) 154 that determines a cost matrix representing an objective function (step 204). The objective function may be defined based on an estimate D(x) of the radiation dose distribution and patient-specific data such as dose targets for PTV and OAR. In some embodiments, the computing system 150 or processor(s) 154 may define the objective function as:

[0053] Φ(x)=W(x)(D(x)-C(x)) 2 · (2)

[0054] The objective function Φ(x) is defined at each voxel x as the squared difference between the estimated radiation dose D(x) and the desired or reference radiation dose C(x), multiplied by a weighting value W(x). The reference radiation dose function C(x) can be defined within each structure (e.g., PTV or OAR) of an anatomical region as equal to a corresponding constant dose value or threshold. The function C(x) can reflect the dose objectives specific to the patient's PTV and OARs. For example, the function C(x) can be equal to a first radiation dose value within PTV 304, equal to a second radiation dose value within OAR 306, and equal to a third radiation dose value within OAR 308. The first, second, and third radiation dose values can be defined based on patient-specific clinical or dose objectives. The weighting function W can reflect the severity of deviation from the radiation function C(x). In an OAR, Φ(x) can be defined as zero where D(x) < C(x). The computing system 150 or the processor(s) 154 can determine the value of voxel x of the cost matrix as Φ(x). Reference Figure 3B , showing a 2-D slice of an example cost matrix. Based on the objective function and Figure 3A the estimated radiation dose distribution D(x) to calculate the cost matrix voxel values or the corresponding objective function values.

[0055] In some embodiments, the cost matrix can be defined to represent the derivative of the objective function Φ(x) with respect to the radiation dose or to represent the absolute value of the derivative The computing system 150 or the processor(s) 154 can define or calculate the voxel value at each voxel x of the cost matrix as the derivative of the objective function Φ(x) with respect to the radiation dose or the corresponding absolute value In some embodiments, the cost matrix can be defined differently. For example, the computing system 150 or the processor(s) 154 can define or calculate the cost matrix as another function of the objective function Φ(x), e.g., different from the absolute value of the derivative.

[0056] Method 200 may include a computing system 150 or processor(s) 154 projecting a cost matrix onto each of a plurality of injection planes (step 206). The computing system 150 or processor(s) 154 may discretize the space of possible gantry-treatment bed orientations. Each point in the discretized space of possible gantry-treatment bed orientations may represent a corresponding gantry-treatment bed orientation among a plurality of gantry-treatment bed orientations, such as (gantry angle, treatment bed angle) pairs. For example, the space of possible gantry-treatment bed orientations may be a 2-D space, where the x-axis represents the available gantry angle and the y-axis represents the available treatment bed angle, and vice versa. That is, assuming both the gantry and the treatment bed are movable or rotatable, each relative orientation or position of the gantry and the treatment bed may be represented according to the corresponding gantry angle and the corresponding treatment bed angle. Each of the gantry angle and the treatment bed angle may be defined in 3-D space relative to a corresponding reference direction. Each (stand angle, treatment bed angle) pair can define the corresponding position and / or orientation of the treatment bed or patient, as well as the corresponding position and / or orientation of the stand or the corresponding direction of the radiation beam.

[0057] For each pair of (stand angle, treatment bed angle), the computing system 150 or processor(s) 154 can calculate the projection of the cost matrix onto the corresponding fluence plane. The computing system 150 or processor(s) 154 can project the voxels of the cost matrix onto the corresponding fluence plane along the direction of the corresponding radiation beam. The voxels of the cost matrix are projected by determining the pixels on the fluence plane that intersect with the rays passing through the voxels in the direction of the radiation beam. The cost value of this voxel is added to the pixel value on the fluence plane.

[0058] In some implementations, computing system 150 or processor(s) 154 may apply weights to each projection of the cost matrix. For each projection of the cost matrix, computing system 150 or processor(s) 104 may apply a corresponding weighting function defined based on the depth of the planned target volume (PTV) within the anatomical region in the direction of the radiation beam. Applying weighted projections may include weighting the projected values ​​of the cost matrix according to the depth of the planned target volume (PTV) within the anatomical region in the direction of the radiation beam. The weight of projections from within the PTV may be assumed to be zero so that only cost contributions from the OAR are included in the projection. For example, (considering the direction of the corresponding radiation beam) computing system 150 or processor(s) 154 may apply a higher weight to volumes preceding the PTV than to volumes following or after the PTV. The weight of the PTV may be assumed to be zero so that only cost contributions from the OAR or from all normal tissues including the OAR are included in the projection.

[0059] Method 200 may include a computing system 150 or processor(s) 154 determining a processing path (step 208) based on the projection of the cost matrix onto a target mask in multiple fluence planes. The computing system 150 or processor(s) 154 may calculate a corresponding aggregated projection value for each (gantry angle, treatment bed angle) pair, the aggregated projection value representing the sum of entries in the corresponding projection matrix. That is, for each (gantry angle, treatment bed angle) pair, the computing system 150 or processor(s) 154 may calculate the sum of entries in the corresponding projection of the cost matrix to determine the corresponding aggregated projection value. In some embodiments, the sum of entries on a target mask in the fluence plane is calculated. The target mask can be formed by projecting voxels of the PTV onto the fluence plane. Pixels receiving any projection are included in the target mask. Some blank space around the target projection may be included in the mask. That is, the projection of the cost matrix may be over the entire fluence plane, but only the portion of the projection hitting the target mask is relevant to determining the aggregated sum and therefore to determining the processing path. The aggregate projection value corresponding to the (stand angle, treatment bed angle) pair represents a measure of the severity of the conflict with medical or clinical standards or constraints. For a given (stand angle, treatment bed angle) pair, the corresponding aggregate projection value indicates whether the beam radiated by the stand at the stand angle, when the treatment bed is oriented according to the treatment bed angle, violates any clinical or medical standards set by the healthcare professional caring for the patient. The larger the aggregate projection value, the more severe the conflict associated with the corresponding (stand angle, treatment bed angle) pair.

[0060] The computing system 150 or processor(s) 154 can use the calculated aggregate projection values ​​to determine the optimal treatment path or trajectory. Specifically, the computing system 150 or processor(s) 154 can apply a path or trajectory search to a matrix of aggregate projection values ​​to determine the optimal treatment path or trajectory. The columns of the aggregate projection value matrix may correspond to different gantry angles, and the rows may correspond to different treatment bed angles, and vice versa. When performing a path search, the computing system 150 or processor(s) 154 can start with an initial entry in the matrix of aggregate projection values ​​and iteratively determine the sequence of entries until a final entry is reached. The computing system 150 or processor(s) 154 can apply the path search in a manner that minimizes the corresponding total severity or the corresponding sum of aggregate projection values. For example, the computing system 150 or processor(s) 154 can apply a path search algorithm, such as the A* algorithm, to determine the path or trajectory with the minimum sum of aggregate projection values.

[0061] Each entry in the determined sequence of entries of the matrix of aggregated projection values ​​represents a corresponding (stand angle, treatment bed angle) pair. Thus, the sequence of entries in the matrix determining the aggregated projection values ​​signifies a sequence of (stand angle, treatment bed angle) pairs that form or represent a treatment path or trajectory. The input to the path search algorithm may include the start and end points of the path. In some embodiments, the start and end points may be the same, such that the path or trajectory forms a complete loop around the patient. The computing system 150 or (multiple) processors 154 may select the starting point as the smallest (stand angle, treatment bed angle) pair corresponding to the aggregated projection values ​​in the matrix. In some embodiments, the computing system 150 or (multiple) processors 154 may select different starting points.

[0062] refer to Figure 3C An image of an example matrix of aggregated projection values ​​310, representing the severity of conflict associated with possible (gantry angle, treatment bed angle) pairs, is shown, along with a processing path 312 determined based on the matrix of aggregated projection values ​​310. Processing path 312 represents a sequence of (gantry angle, treatment bed angle) pairs that define a loop around the patient. The gantry angles are equally spaced. The processing path shown is associated with the minimum sum of corresponding entries in the matrix of aggregated projection values ​​310. Flat gray areas are blocked by those (gantry angle, treatment bed angle) directions that cause conflict, or by beams entering the patient volume through the clipping plane of the CT image. Points a and b in the aggregated projection matrix correspond to... Figure 3B The projection directions are a and b.

[0063] Figure 4A Images depicting example arrangements of various hypothetical anatomical structures are shown. Structure 402 represents the PTV, while structures 404 and 406 represent two different OARs. Figure 4B It shows the depiction in the corresponding Figure 4A An image of a visual representation of the projection of the cost matrix onto the beam orientation. In this case, it is assumed that the OAR structure 406 has clinical targets that highly conflict with the estimated radiation dose, therefore in Figure 4B High intensity is generated in the projection of the cost matrix. In contrast, it is assumed that OAR structure 404 has clinical goals that conflict less with the estimated radiation dose, therefore in Figure 4B The projection of the cost matrix produces a relatively low intensity.

[0064] Figure 5A and 5B Images illustrating a processing path according to the inventive concept of this disclosure are shown in two-dimensional (2-D) and three-dimensional (3-D) representations. Figure 5AAnother example is shown: a matrix of convergent projection values ​​and an optimal path 504 that minimizes the sum of the corresponding entries in the convergent projection matrix. Flat gray areas are blocked by those directions that cause conflict (gantry angle, treatment bed angle) or by beams entering the patient volume through the shear plane of the CT image. Figure 5B The optimal path 502 in 3-D space is shown. Figure 5B The diagram also illustrates the radiation beam at a point along the optimal path.

[0065] The computing system 150 or processor(s) 154 may employ method 200 to optimize a radiation plan for intensity-modulated radiotherapy (IMRT) or volume-modulated arc therapy (VMAT). For example, the objective function may be defined as optimizing an IMRT-based radiation plan or a VMAT-based radiation plan. In VMAT, a multi-leaf collimator (MLC) mounted at the head of the gantry is used to shape the radiation beam. The MLC includes a set of metal blades that move in and out and block portions of the radiation to modulate the beam and make the radiation more conform to the PTV shape. In VMAT, the gantry can continuously deliver radiation while moving around the patient, and the MLC can block radiation at certain points along the path. Thus, optimization of the VMAT processing path may involve determining the path segments where the MLC blocks radiation. In IMRT, the gantry stops at several angles (e.g., about 5 to 10 angles) and delivers radiation by modulating the beam. Thus, path optimization may include determining pairs of (gantry angle, treatment bed angle) where the gantry stops delivering radiation to the patient.

[0066] It should be noted that the examples discussed in this specification are for illustrative purposes only and should not be construed as limiting. For example, an estimate of the radiation dose distribution can be defined using functions other than the function D(x) described in equation (1). Furthermore, the computing system 150 can initiate the path search algorithm in various different ways.

[0067] The various methods described in this disclosure can be implemented by computer code instructions stored on a computer-readable medium. When executed by one or more processors of a computing device, the computer code instructions cause the computing device to perform the method.

[0068] Although the invention has been specifically shown and described with reference to particular embodiments, those skilled in the art will understand that various changes in form and detail may be made to the invention without departing from its spirit and scope.

[0069] While this disclosure contains numerous details of specific embodiments, these should not be construed as limiting the scope of any invention or the scope that may be claimed, but rather as descriptions of features characteristic of particular embodiments of a particular invention. Certain features described in this specification in the context of independent embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented individually in multiple embodiments or in any suitable sub-combination. Furthermore, although features may be described above as functioning in some combinations, and even initially claimed in this way, one or more features from a claimed combination may be removed from that combination in some cases, and the claimed combination may be for sub-combinations or variations thereof.

[0070] Similarly, although the operations are depicted in a specific order in the figures, this should not be construed as requiring such operations to be performed in the specific or sequential order shown, or requiring all illustrated operations to achieve the desired result. In some cases, multitasking and parallel processing may be advantageous. Furthermore, the separation of various system components in the above embodiments should not be construed as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated into a single software product or packaged into multiple software products.

[0071] A reference to “or” can be interpreted as inclusive, so any term described using “or” can refer to a single, more than one, or any of the terms described.

[0072] Therefore, specific embodiments of the subject matter have been described. Other embodiments are within the scope of the following claims. In some cases, the actions described in the claims can be performed in a different order and the desired result can still be obtained. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing may be advantageous.

Claims

1. A method for radiation treatment planning, comprising: An estimate of the radiation dose distribution within the patient's anatomical regions is determined by one or more processors; The cost matrix is ​​determined by the one or more processors using the estimate of the radiation dose distribution, the cost matrix representing an objective function defined based on the estimate of the radiation dose distribution and patient-specific data; The cost matrix is ​​projected by the one or more processors onto each of a plurality of fluence planes, each of the plurality of fluence planes being associated with a corresponding gantry-treatment bed orientation among a plurality of gantry-treatment bed orientations of a medical linear accelerator device; as well as Using the projection of the cost matrix onto each of the plurality of injection planes, a sequence of bench-to-bed orientations representing the treatment path is determined among the plurality of bench-to-bed orientations.

2. The method of claim 1, wherein determining the estimate of a radiation dose distribution comprises: The estimate of radiation dose distribution is determined based on the distance from the planned target volume (PTV) to the anatomical region.

3. The method according to claim 1 or 2, wherein the objective function reflects one or more radiation constraints for the patient.

4. The method according to claim 1 or 2, wherein the objective function is defined as optimizing a radiation plan based on intensity modulated radiation therapy (IMRT).

5. The method according to claim 1 or 2, wherein the objective function is defined as optimizing a radiation program based on volume modulated arc therapy (VMAT).

6. The method according to claim 1 or 2, further comprising: The plurality of frame-treatment bed orientations are determined by discretizing the space of possible frame-treatment bed orientations, where each point in the discretized space of possible frame-treatment bed orientations represents the corresponding frame-treatment bed orientation among the plurality of frame-treatment bed orientations.

7. The method of claim 1 or 2, wherein projecting the cost matrix onto each of the plurality of injection planes includes applying a weighted projection.

8. The method of claim 7, wherein applying the weighted projection comprises: The projected values ​​of the cost matrix are weighted according to the depth of the planned target volume (PTV) within the anatomical region relative to the direction of the radiation beam.

9. The method of claim 1 or 2, wherein determining the sequence of frame-treatment bed orientation comprises: For each gantry-treatment bed orientation, the corresponding matrix sum is calculated, whereby the matrix sum represents the sum of the projections of the cost matrix onto the target mask of the injection plane associated with the gantry-treatment bed orientation; as well as The sequence of bench-bed orientations is determined using the sum of matrices calculated for the plurality of bench-bed orientations.

10. The method of claim 9, wherein determining the sequence of gantry- couch orientations comprises: The sum of the matrix values ​​along the processing path is minimized, and the processing path extends over a predefined range of the frame-treatment bed orientation.

11. A radiation treatment planning system, comprising: One or more processors; as well as Non-transitory memory for storing computer code instructions, which, when executed, cause the one or more processors to: Determine an estimate of the radiation dose distribution within the patient's anatomical regions; The cost matrix is ​​determined using the estimate of the radiation dose distribution, the cost matrix representing an objective function defined based on the estimate of the radiation dose distribution and patient-specific data; The cost matrix is ​​projected onto each of a plurality of fluence planes, each of the plurality of fluence planes being associated with a corresponding gantry-treatment bed orientation among a plurality of gantry-treatment bed orientations of a medical linear accelerator device; as well as Using the projection of the cost matrix onto each of the plurality of injection planes, a sequence of bench-to-bed orientations representing the treatment path is determined among the plurality of bench-to-bed orientations.

12. The radiation treatment planning system of claim 11, wherein determining the estimate of a radiation dose distribution comprises: The estimate of radiation dose distribution is determined based on the distance from the planned target volume (PTV) to the anatomical region.

13. The radiation treatment planning system of claim 11 or 12, wherein the objective function reflects one or more radiation constraints for the patient.

14. The radiation treatment planning system according to claim 11 or 12, wherein the objective function is defined as optimizing a radiation plan based on intensity modulated radiation therapy (IMRT) or as optimizing a radiation plan based on volume modulated arc therapy (VMAT).

15. The radiation treatment planning system according to claim 11 or 12, wherein the one or more processors are further configured to: The plurality of frame-treatment bed orientations are determined by discretizing the space of possible frame-treatment bed orientations, where each point in the discretized space of possible frame-treatment bed orientations represents the corresponding frame-treatment bed orientation among the plurality of frame-treatment bed orientations.

16. The radiation processing planning system of claim 11 or 12, wherein the one or more processors are configured to apply weighted projection when projecting the cost matrix onto each of a plurality of fluence planes.

17. The radiation processing planning system of claim 16, wherein when applying the weighted projection, the one or more processors are configured to weight the projected values ​​of the cost matrix according to the depth of the planned target volume (PTV) within the anatomical region in the direction of the radiation beam.

18. The radiation treatment planning system of claim 11 or 12, wherein, in determining the sequence of gantry-treatment bed orientation, the one or more processors are configured to: For each gantry-treatment bed orientation, calculate the corresponding matrix sum, whereby the matrix sum represents the sum of the projections of the cost matrix onto the target mask of the injection plane associated with that gantry-treatment bed orientation; and The sequence of bench-bed orientations is determined by using the sum of matrices calculated for the plurality of bench-bed orientations.

19. The radiation treatment planning system of claim 18, wherein, in determining the sequence of gantry-treatment bed orientations, the one or more processors are configured to minimize the sum of matrix values ​​along the processing path, the processing path extending over a predefined range of gantry-treatment bed orientations.

20. A computer-readable medium including computer code instructions stored thereon, the computer code instructions causing one or more processors to: Determine an estimate of the radiation dose distribution within the patient's anatomical regions; The cost matrix is ​​determined using the estimated radiation dose distribution, the cost matrix representing an objective function defined based on the estimated radiation dose distribution and patient-specific data; The cost matrix is ​​projected onto each of a plurality of fluence planes, each of the plurality of fluence planes being associated with a corresponding gantry-treatment bed orientation among a plurality of gantry-treatment bed orientations of a medical linear accelerator device; as well as The sequence of bench-to-bed orientations representing treatment paths among the plurality of bench-to-bed orientations is determined by projecting the cost matrix onto each of the plurality of injection planes.

Images