Quality Assurance for MR-LINAC

Abstract

The present disclosure includes a process for automating quality assurance testing for a radiotherapy apparatus including an MR-Linac device that is agnostic with respect to the manufacturer and supplier of the apparatus. The present disclosure includes a process for performing a validation process on a linear accelerator of the MR-Linac apparatus. The present disclosure includes a process for analyzing the quality of images produced by an MR imaging device of the MR-Linac apparatus. The present disclosure includes a process for combining the performance of a validation process on a linear accelerator of the MR-Linac apparatus with the analysis of the quality of images produced by the MR imaging device of the MR-Linac apparatus.

Description

[0001] Priority claim

[0002] This application claims priority to U.S. Provisional Application No. 62 / 744,875, filed on October 12, 2018, the entire contents of which are incorporated herein by reference. Technical Field

[0003] Embodiments of the present disclosure generally relate to quality assurance techniques in radiotherapy apparatuses. In particular, the present disclosure relates to performing automatic quality assurance for an integrated linear accelerator and MRI imaging device (MR-Linac or MRI-Linac). Background Art

[0004] In radiosurgery or radiotherapy (collectively referred to as "radiotherapy"), a very intense and precisely collimated dose of radiation is delivered to a target area in a patient to treat or destroy a lesion. Typically, the target area consists of a large amount of tumor tissue. Radiotherapy requires extremely accurate spatial positioning of the targeted lesion. Computed tomography ("CT"), magnetic resonance imaging ("MRI") scanning and other imaging methods enable physicians to accurately locate the lesion relative to bone landmarks or implanted fiducial markers, and to control the position of the radiation source so that its beam can be accurately directed to the target tissue while avoiding adjacent critical body structures. Therefore, radiotherapy requires high-precision diagnosis and high-precision radiation source control. The consequences of deviations outside the prescribed tolerances for diagnosis and radiation source control may cause potential harm to the patient.

[0005] In view of the need for accurate implementation of radiotherapy, regulations require that quality assurance tests be performed to ensure that the radiotherapy device is operating properly. Such quality assurance tests are routinely performed on the radiotherapy device before the prescribed radiation dose is delivered to the patient, and are manually analyzed by the operator to ensure the proper alignment and configuration of the system. However, the manual analysis of quality assurance tests is extremely time-consuming, which results in a lot of downtime for the device, and the patient cannot be treated during the downtime. In addition, the manual analysis of the tests by various operators is often inconsistent. In addition, although there are some automatic quality assurance processes for analyzing radiotherapy devices, such techniques are generally machine-specific and require the use of supplier-specific hardware. This also leads to inconsistent determination of quality assurance between different radiotherapy device manufacturers. Summary of the invention

[0006] The present disclosure includes a process for automating quality assurance testing of a radiotherapy apparatus including an MR-Linac device that is agnostic to the manufacturer and supplier of the apparatus. The present disclosure includes a process for performing a validation process on a linear accelerator of the MR-Linac apparatus. The present disclosure includes a process for analyzing the quality of images produced by an MR imaging device of the MR-Linac apparatus. The present disclosure includes a process for combining the performance of a validation process on a linear accelerator of the MR-Linac apparatus with the analysis of the quality of images produced by the MR imaging device of the MR-Linac apparatus.

[0007] In some embodiments, a system, method, and computer-readable medium for verifying the operation of a radiotherapy apparatus including an integrated linear accelerator and an MRI imaging device (MR-Linac) are provided. The system, method, and computer-readable medium perform operations including: selecting a verification process associated with a linear accelerator of the MR-Linac; retrieving configuration information of a multi-leaf collimator (MLC) of the MR-Linac based on the selected verification process; obtaining a plurality of images from an imaging detector, each image being associated with a given radiation beam of a plurality of radiation beams that have passed through the MLC that has been adjusted based on the configuration information; processing the plurality of images to determine characteristics of the MR-Linac; and determining whether the MR-Linac is operating within a valid range based on the processing of the plurality of images.

[0008] In some implementations, systems, methods, and computer-readable media perform operations including comparing the determined characteristics to allowable thresholds associated with a selected authentication process, and in response to determining that MR-Linac is operating outside of a valid range, adjusting one or more characteristics of MR-Linac.

[0009] In some implementations, one or more characteristics are determined based on the selected authentication process.

[0010] In some implementations, systems, methods, and computer-readable media perform operations including automatically adjusting an MLC in response to retrieval of configuration information and selection of a verification process.

[0011] In some implementations, the plurality of verification processes include a leaf position verification process of the MLC, a radiation source and imaging detector movement verification process, a linear accelerator and MRI imaging device alignment verification process, and a linear accelerator flatness symmetry verification process.

[0012] In some implementations, the selected validation process includes a blade position validation process, wherein the plurality of images includes at least seven images obtained at the same gantry angle of the MR-Linac, wherein the characteristics include blade and jaw positions relative to nominal blade and jaw positions, and wherein the allowed threshold includes one millimeter.

[0013] In some implementations, a first image, a second image, and a third image among seven images are processed to determine a radiation isocenter, configuration information indicates a first position of a first group of blades and a second group of blades for the MLC, a second position of a third group of blades for the MLC between the first group of blades and the second group of blades, and the operation also includes: segmenting the first image, the second image, and the third image; calculating a center of mass for each of the segmented first image, the segmented second image, and the segmented third image; fitting a circular function to the calculated center of mass; determining a radiation isocenter associated with each of the first image, the second image, and the third image based on the center of the circular function; and detecting motion in the beam based on the difference in the determined radiation isocenter associated with each of the first image, the second image, and the third image.

[0014] In some implementations, a fourth image of the seven images is processed to identify an open field, a configuration is provided indicating an open position of blades for the MLC, and the operation further includes performing the following operations for each edge in the open field of the fourth image: generating an initial contour based on pixel intensity values ​​of the open field edge; modifying points of the initial contour to minimize the internal energy of the contour; moving the points toward the strongest gradient and maintaining regular spacing between the points; and calculating a second order derivative of the contour to identify a corner of the open field edge.

[0015] In some implementations, a fifth image of the seven images is processed to detect extended blades of the MLC, a first position indicating a first group of blades and a second group of blades for the MLC is configured, and a second position indicating a third group of blades between the first group and the second group is alternated between a second position and a third position, and the operation further includes: identifying a center of each of the third group of blades based on an edge of a field in the fifth image; and identifying a blade edge of the third group of blades.

[0016] In some implementations, a sixth image and a seventh image of seven images are processed to detect blade and jaw positions, and for the sixth image, a configuration indicates a first position of a first blade stack for MLC and a second position of a second blade stack for MLC, and for the seventh image, the configuration indicates the above-mentioned first position of the second blade stack for MLC and the above-mentioned second position of the first blade stack for MLC, and the operation also includes: generating an intensity distribution of the sixth image and the seventh image; identifying the blade position based on the distance between the horizontal center line of the field in the sixth image and the seventh image and the identified blade edge; and identifying the jaw position based on the field edge distribution of the sixth image and the seventh image.

[0017] In some implementations, the selected validation process includes a linear accelerator flatness symmetry validation process, wherein the plurality of images includes four images acquired at four different gantry angles of an MR-Linac, wherein the characteristics include symmetry and flatness of the images, and wherein the allowable threshold includes a specified percentage value.

[0018] In some implementations, the operations include performing the following operations for each image: determining the location of a field edge of a field in the image; calculating an average of the field edge locations to determine the center of the field; generating an average distribution from nine vertical distributions, wherein a middle distribution of the nine vertical distributions passes through the center of the field; identifying a location in the nine vertical distributions corresponding to 80% intensity; determining a difference between a first vertical distribution and a second vertical distribution that are equidistant from the middle distribution in the nine vertical distributions to determine symmetry; and determining a variation between values ​​of the nine vertical distributions to determine flatness.

[0019] In some embodiments, a system, method, and computer-readable medium are provided for determining image quality of a radiotherapy apparatus including an integrated linear accelerator and an MRI imaging device (MR-Linac). The system, method, and computer-readable medium perform operations, including: selecting a quality assessment process associated with an MRI imaging device of MR-Linac; obtaining a plurality of images of a phantom including at least one module from the MRI imaging device; processing features in the plurality of images representing components in the at least one module based on the selected quality assessment process; calculating image quality attributes associated with the MRI imaging device based on the processed features; determining whether the image quality attributes meet an allowable standard; and determining whether the MR-Linac is operating effectively in response to determining that the image quality attributes meet the allowable standard.

[0020] In some implementations, the operations include, in response to determining that the image quality attribute does not meet an allowable criterion, adjusting one or more characteristics of the MR-Linac.

[0021] In some implementations, the quality assessment process is a first quality assessment process for calculating the position of the phantom relative to the axial field of view and calculating the size of the phantom.

[0022] In some implementations, the operations include: selecting a first image from a plurality of images; binarizing the first image so that a phantom displayed in the image has uniform pixel values; identifying the column numbers of the first pixel and the last pixel of the phantom in each row of pixels; generating a region of interest (ROI) around the phantom based on the identified column numbers; identifying the center of the ROI based on the identified column numbers; identifying the center of the first image; calculating the difference between the coordinates of the phantom and the center of the first image to determine misalignment of the phantom; and calculating the height and width of the phantom based on the ROI.

[0023] In some implementations, the plurality of quality assessment processes include at least one of the quality assessment processes for checking rotation of the phantom relative to image space, calculating slice thickness accuracy, calculating high contrast spatial resolution, checking slice position accuracy, calculating percent image uniformity, calculating percent signal ghosting, calculating low contrast detectability, or calculating geometric accuracy.

[0024] In some embodiments, a system, method, and computer-readable medium for performing quality assessment on a radiotherapy apparatus including an integrated linear accelerator and an MRI imaging device (MR-Linac) are provided. The system, method, and computer-readable medium perform operations including: selecting a validation process associated with the linear accelerator of the MR-Linac and a quality assessment process associated with the MRI imaging device of the MR-Linac; obtaining a first plurality of images from an imaging detector, each image associated with a given radiation beam of a plurality of radiation beams that have passed through an MLC that has been adjusted based on the selected validation process; obtaining a second plurality of images of a phantom including at least one module from the MRI imaging device; processing the first plurality of images and the second plurality of images based on the selected validation process and the quality assessment process; and determining whether the MR-Linac is operating within a valid range based on the properties of the first plurality of images and the second plurality of images determined based on the processing of the first plurality of images and the second plurality of images.

[0025] In some implementations, the operations include, in response to determining that the MR-Linac is operating outside of a valid range, adjusting one or more characteristics of the MR-Linac.

[0026] In some implementations, one or more characteristics are determined based on a selected validation process or quality assessment process.

[0027] The above summary is intended to provide an overview of the subject matter of this patent application. It is not intended to provide an exclusive description or exhaustive description of the present disclosure. The specific implementation is included to provide additional information about this patent application. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In the drawings, which are not necessarily drawn to scale, similar reference numerals describe substantially similar components throughout the several views. Similar reference numerals with different letter suffixes represent different instances of substantially similar components. The accompanying drawings generally illustrate various embodiments discussed in this document by way of example and not by way of limitation.

[0029] Figure 1 An exemplary radiation therapy system suitable for performing a treatment plan generation process according to some examples is shown.

[0030] Figure 2 An exemplary image-guided radiation therapy apparatus according to some examples of the present disclosure is shown.

[0031] Figure 3 is a perspective view of a multi-leaf collimator according to some examples of the present disclosure.

[0032] Figure 4A and Figure 4B An illustrative process for testing the MLC and jaw configuration of an MRI-Linac device according to some examples of the present disclosure is provided.

[0033] Figure 5 The input and output of the MLC and jaw configuration tests of an MRI-Linac device according to some examples of the present disclosure are shown.

[0034] Figures 6 to 9 An illustrative process for testing an MRI-Linac device according to some examples of the present disclosure is provided.

[0035] Fig. 10A and Fig. 10B The input and output of an image quality test of an MRI-Linac device according to some examples of the present disclosure are shown.

[0036] Fig.11 An exemplary web-based interface for testing an MRI-Linac device according to some examples of the present disclosure is shown. DETAILED DESCRIPTION

[0037] The present disclosure includes techniques for automating quality assurance testing of radiotherapy apparatus including MR-Linac devices. The present disclosure includes processing for performing a validation process on a linear accelerator of an MR-Linac apparatus. The present disclosure includes processing for analyzing the quality of images produced by an MR imaging device of an MR-Linac apparatus. These technical advantages include increased accuracy and consistency between quality assurance tests performed across different apparatuses at different times. These technical advantages reduce downtime of radiotherapy apparatuses, thereby allowing more patients to be treated. Accordingly, in addition to these technical benefits, the present technology can also produce many significant medical treatment benefits (including improving the accuracy of radiotherapy treatments, reducing exposure to unintended radiation, etc.).

[0038] In some embodiments, the disclosed technology includes a web-based tool for analyzing 3D images from an MRI imaging device. These techniques are applicable to any MRI imaging device and are not related to a specific manufacturer. The disclosed technology processes images to find the center of mass of the phantom and the height and width of the phantom in pixels in a collection of image files (e.g., DICOM image files) from the MRI imaging device. Determine the rotation, slice thickness, high contrast spatial resolution, slice position accuracy, image uniformity percentage, and signal artifact percentage. Using this information, determine the accuracy of the MRI imaging device. In some implementations, the height and width of the phantom in pixels are converted to millimeters. In some embodiments, the rotation of the phantom is determined by analyzing the top and bottom of both sides of the black bar in the center of an image. The slice thickness is determined by analyzing the brightest portion of several lines in the central portion of the first image.

[0039] In some embodiments, the disclosed technology includes a supplier-agnostic web-based tool that provides quality control for all radiotherapy devices and imaging devices. In particular, the disclosed technology includes a web-based tool that can analyze the quality and proper performance of a system having a linear accelerator (Linac) associated with a first manufacturer and a digital imaging panel (detector) associated with a different second manufacturer. For example, a multi-leaf collimator (MLC) leaf position test is provided that enhances extended leaf detection and subtraction profile to support the digital imaging panel. In some embodiments, the image produced by the digital imaging panel has a different grayscale value than the imaging panel provided by the first manufacturer and used by the Linac associated with the first manufacturer. Various image sets are obtained using the digital imaging panel to determine the exact position of the leaf. An extended leaf detection test is used, which detects the leaf extended through the field to calculate the rotation between the imaging panel and the leaf travel. A subtraction profile test is used to produce a composite image by subtracting two 20×20 cm jaw position detection fields. In particular, the MLC includes one or more plates above or below the leaves of the MLC, which also control the direction and amount of radiation in a given beam. These jaw positions can be determined in the detected image field. Quality assurance is then determined for such radiotherapy devices that include components from different manufacturers.

[0040] In some embodiments, the disclosed technology includes providing a supplier-agnostic web-based tool for quality control of MR-Linac devices. An automated quality control test specific to MR-Linac (MRL) devices is provided. Three image-based tests are provided, including an MRLMLC leaf position test, a radiation source and imaging panel (detector) movement test, and an output flatness symmetry versus gantry angle test. In this device, the photon beam characteristics and system geometry are different from the traditional Linac system considered in the disclosed quality control test for MR-Linac devices. In the MRL MLC leaf position test, the MLC leaf position accuracy of 28 leaves is determined. The test takes into account the MR-Linac field size, field characteristics, and machine size. In the radiation source and imaging panel (detector) movement test, the position of the radiation source is detected according to the change in the gantry angle to detect the 3D position of the megavolt (MV) phantom, thereby determining the maximum displacement of the radiation axis relative to the center of the average radiation field. Field edge detection uses a gradient technique. In the output flatness symmetry vs. gantry angle test, analysis of four open field images was performed to determine flatness and symmetry in the radial and transverse directions. The output calculated from the center portion of each field was also considered.

[0041] Figure 1 An exemplary radiation therapy system 100 is shown that is suitable for performing radiation therapy planning processing operations using one or more of the methods discussed herein. These radiation therapy planning processing operations are performed so that the radiation therapy system can provide radiation therapy to a patient based on captured medical imaging data and specific aspects of treatment dose calculations. The radiation therapy system includes a radiation therapy processing computing system 110 that hosts image processing logic 120. The radiation therapy processing computing system 110 can be connected to a network (not shown), and such a network can be connected to the Internet. For example, the network can connect the radiation therapy processing computing system 110 with one or more medical information sources (e.g., a radiation information system (RIS), a medical record system (e.g., an electronic medical record (EMR) / electronic health record (EHR) system), an oncology information system (OIS)), one or more image data sources 150, an image acquisition device 170 (e.g., an imaging modality), a treatment device 180 (e.g., a radiation therapy device), and a treatment data source 160.

[0042] The radiation therapy treatment computing system 110 may include a processing circuit system 112, a memory device 114, a storage device 116, and other hardware and software operable features such as a user interface 142, a communication interface (not shown), etc. The storage device 116 may store transient or non-transient computer executable instructions, such as an operating system, a radiation therapy treatment plan, a quality assurance process, an image quality assessment process, a verification process, a software program (e.g., image processing software, image or anatomical visualization software, AI implementations and algorithms such as provided by DL models, ML models, and neural networks, etc.), and any other computer executable instructions to be executed by the processing circuit system 112. The processing circuit system 112 may access the instructions stored in the storage device 116 to perform and execute any one or more of the quality assurance and verification processes discussed herein. The processing circuit system 112 may provide a web tool for an operator to instruct the processing circuit system 112 to perform the quality assurance and verification process, and the processing circuit system 112 may output and present the results of the quality assurance and verification process to the operator in graphical or tabular form via the web tool.

[0043] For example, users can Fig.11 The web tool shown is used to instruct the processing circuit system 112 to perform any one of the quality assurance and verification process set. After the processing circuit system 112 performs the process selected by the user, Fig.11 In some embodiments, Fig.11 The web tool shown allows a user to input one or more previously captured images of an MRI-linac device, and these images are then processed by the processing circuit system 112 to perform selected quality assurance and verification processes. In some embodiments, the processing circuit system 112 processes the images via Fig.11 The web tool shown instructs the user on the proper placement and positioning of a phantom or other components associated with testing the MRI-Linac device. After the user configures the MRI-Linac device as directed by the web tool, one or more test images may be captured and processed by the processing circuit system 112 to perform quality assurance and validation procedures. The quality assurance and validation procedures that a user may choose to perform are shown and described in conjunction with Figures 4 to 10. In some embodiments, the web tool may notify the user whether any one of a set of quality assurance and validation procedures has passed or failed, and may provide an operating range or graph within the allowable operating range for a given MRI-Linac device. In an embodiment, the user may use Fig.11 A web tool is provided to manipulate the allowed thresholds to configure whether a given MRI-Linac device passes or fails a given quality assurance and validation process.

[0044] In an example, processing circuit system 112 may include a processing device, for example, one or more general-purpose processing devices such as a microprocessor, a central processing unit (CPU), a graphics processing unit (GPU), an accelerated processing unit (APU), etc. More specifically, processing circuit system 112 may be a complex instruction set computing (CISC) microprocessor, a reduced instruction set computing (RISC) microprocessor, a very long instruction word (VLIW) microprocessor, a processor that implements other instruction sets, or a processor that implements a combination of instruction sets. Processing circuit system 112 may also be implemented by one or more special-purpose processing devices such as an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital signal processor (DSP), a system on a chip (SoC), etc. As will be appreciated by those skilled in the art, in some examples, processing circuit system 112 may be a special-purpose processor rather than a general-purpose processor. Processing circuit system 112 may include one or more known processing devices, such as those from Intel. TM Pentium manufactured TM 、Core TM , Xeon TM or series of microprocessors from AMD TM Made by Turion TM , Athlon TM 、Sempron TM 、Opteron TM FX TM 、Phenom TM series of microprocessors or any of a variety of processors manufactured by Sun Microsystems. Processing circuit system 112 may also include a graphics processing unit, such as a graphics processing unit from Nvidia TM Manufactured series, by Intel TM GMA, Iris TM Series or by AMD TM Made by Radeon TM The processing circuit system 112 may also include a GPU such as that provided by Intel TM Xeon Phi manufactured TMseries of accelerated processing units. The disclosed embodiments are not limited to any type of processor that is otherwise configured to meet the computational requirements of identifying, analyzing, maintaining, generating and / or providing large amounts of data or manipulating such data to perform the methods disclosed herein. In addition, the term "processor" may include more than one physical (circuitry-based) or software-based processor, such as a multi-core design or multiple processors each having a multi-core design. The processing circuit system 112 can execute a sequence of transient or non-transient computer program instructions stored in the memory device 114 and accessed from the storage device 116 to perform various operations, processes, and methods that will be described in more detail below. It should be understood that any component in the radiation therapy system 100 can be implemented separately and operated as a stand-alone device, and can be coupled to any other component in the radiation therapy system 100 to perform the techniques described in the present disclosure.

[0045] Memory device 114 may include read-only memory (ROM), phase-change random access memory (PRAM), static random access memory (SRAM), flash memory, random access memory (RAM), dynamic random access memory (DRAM) such as synchronous DRAM (SDRAM), electrically erasable programmable read-only memory (EEPROM), static memory (e.g., flash memory, flash disk, static random access memory), and other types of random access memory, cache memory, registers, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage device, cassette tape, other magnetic storage device, or any other non-transitory medium that can be used to store information including images, data, or transient or non-transitory computer executable instructions (e.g., stored in any format) that can be accessed by processing circuit system 112 or any other type of computer device. For example, computer program instructions can be accessed by processing circuit system 112, can be read from ROM or any other suitable memory location, and can be loaded into RAM for execution by processing circuit system 112.

[0046] The storage device 116 may constitute a drive unit including a transitory or non-transitory machine-readable medium on which is stored one or more transitory or non-transitory instruction sets and data structures (e.g., software) implementing or utilized by any one or more of the methods or functions described herein (including the user interface 142 in various examples). During execution of the instructions by the radiation therapy treatment computing system 110, the instructions may also reside, in whole or in part, within the memory device 114 and / or the processing circuit system 112, wherein the memory device 114 and the processing circuit system 112 also constitute transitory or non-transitory machine-readable media.

[0047] The memory device 114 and the storage device 116 may constitute a non-transitory computer-readable medium. For example, the memory device 114 and the storage device 116 may store or load transient or non-transitory instructions for one or more software applications on a computer-readable medium. Software applications stored or loaded using the memory device 114 and the storage device 116 may include, for example, operating systems for general-purpose computer systems and software-controlled devices. The radiotherapy processing computing system 110 may also operate various software programs including software codes for implementing the image processing logic 120 and the user interface 142. In addition, the memory device 114 and the storage device 116 may store or load the entire software application, a portion of the software application, or code or data associated with the software application that can be executed by the processing circuit system 112. In another example, the memory device 114 and the storage device 116 may store, load, and manipulate one or more radiotherapy treatment plans, imaging data, segmentation data, treatment visualization, histograms or measurement results, AI model data (e.g., weights and parameters), labeling and mapping data, etc. It is contemplated that the software program may be stored not only on storage device 116 and memory device 114, but also on removable computer media such as a hard drive, a computer disk, a CD-ROM, a DVD, a Blu-ray DVD, a USB flash drive, an SD card, a memory stick, or any other suitable media; such software programs may also be transmitted or received over a network.

[0048] Although not shown, the radiation therapy treatment computing system 110 may include a communication interface, a network interface card, and a communication circuit system. Exemplary communication interfaces may include, for example, a network adapter, a cable connector, a serial connector, a USB connector, a parallel connector, a high-speed data transmission adapter (e.g., such as optical fiber, USB 3.0, thunderbolt, etc.), a wireless network adapter (e.g., such as an IEEE 802.11 / Wi-Fi adapter), a telecommunications adapter (e.g., communicating with 3G, 4G / LTE, and 5G networks, etc.), etc. Such a communication interface may include one or more digital and / or analog communication devices that allow the machine to communicate with other machines and devices, such as remotely located components, via a network. The network may provide the functionality of a local area network (LAN), a wireless network, a cloud computing environment (e.g., software as a service, platform as a service, infrastructure as a service, etc.), a client-server, a wide area network (WAN), etc. For example, the network may be a LAN or WAN that may include other systems (including additional image processing computing systems or image-based components associated with medical imaging or radiation therapy operations).

[0049] In an example, the radiation therapy treatment computing system 110 may obtain image data 152 from an image data source 150 (eg, an MRI image) to be hosted on the storage device 116 and the memory device 114. Figure 2 An exemplary image data source 150 is described in detail. In an example, a software program running on the radiation therapy treatment computing system 110 can process image data 152 obtained by the image data source 150 to perform an image quality assessment to provide a quality assurance metric or verification for the radiation therapy device. In another example, a software program running on the radiation therapy treatment computing system 110 can process image data 162 obtained by a treatment device 180 (e.g., a beam detector) to perform a verification process to provide a quality assurance metric or verification for the radiation therapy device. In an embodiment, characteristics of the radiation therapy device can be adjusted based on an automatic quality assessment of the image data 162 obtained by the image data source 150 and / or the image data 162 obtained by the treatment device 180.

[0050] Processing circuit system 112 may be communicatively coupled to memory device 114 and storage device 116, and processing circuit system 112 may be configured to execute computer executable instructions stored thereon from memory device 114 or storage device 116. Processing circuit system 112 may execute instructions such that a medical image from image data 152 is received or obtained in memory device 114 and processed using image processing logic 120.

[0051] In an example, the image data 152 may include one or more MRI images (e.g., 2D MRI, 3D MRI, 2D flow MRI, 4D MRI, 4D volume MRI, 4D imaging MRI, etc.), functional MRI images (e.g., fMRI, DCE-MRI, diffusion MRI), computed tomography (CT) images (e.g., 2D CT, 2D cone beam CT, 3D CT, 3D CBCT, 4D CT, 4D CBCT), ultrasound images (e.g., 2D ultrasound, 3D ultrasound, 4D ultrasound), positron emission tomography (PET) images, X-ray images, fluoroscopic images, radiotherapy portal images, single photon emission computed tomography (SPECT) images, computer-generated synthetic images (e.g., pseudo CT images), etc. In addition, the image data 152 may also include or be associated with medical image processing data, such as training images and ground truth images, contour images, and dose images.

[0052] In an example, the image data 152 may be received from an image acquisition device 170 and stored in one or more image data sources 150 (e.g., a picture archiving and communication system (PACS), a vendor neutral archive (VNA), a medical record or information system, a data warehouse, etc.). Thus, the image acquisition device 170 may include an MRI imaging device, a CT imaging device, a PET imaging device, an ultrasound imaging device, a fluoroscope device, a SPECT imaging device, an integrated linear accelerator and MRI imaging device, a CBCT imaging device, or other medical imaging device for obtaining medical images of a patient. The image data 152 may be received and stored in any data type or any format type (e.g., in a Digital Imaging and Communications in Medicine (DICOM) format) that the image acquisition device 170 and the radiation therapy processing computing system 110 may use to perform operations consistent with the disclosed embodiments.

[0053] In an example, the image acquisition device 170 can be integrated with the treatment device 180 into a single device (e.g., an MRI device combined with a linear accelerator, also referred to as an "MRI-Linac" or "MR-Linac"). Such an MRI-Linac can be used, for example, to determine the location of a target organ or target tumor in a patient, thereby accurately guiding the radiation therapy to a predetermined target according to a radiation therapy treatment plan. For example, a radiation therapy treatment plan can provide information about a specific radiation dose to be applied to each patient. The radiation therapy treatment plan can also include other radiation therapy information, such as beam angles, dose-histogram-volume information, the number of radiation beams to be used during treatment, the dose of each beam, and the like.

[0054] The radiotherapy treatment computing system 110 can communicate with an external database over a network to send / receive a plurality of various types of data related to image processing and radiotherapy operations. For example, the external database may include machine data (including equipment constraints) that provides information associated with the treatment device 180, the image acquisition device 170, or other machines related to radiotherapy or medical procedures. The machine data information may include radiation beam size, arc placement, beam switch duration, machine parameters, segments, multi-leaf collimator (MLC) configuration, gantry speed, MRI pulse sequence, etc. The external database may be a storage device and may be equipped with an appropriate database management software program. In addition, such a database or data source may include multiple devices or systems located in a central manner or in a distributed manner.

[0055] The radiation therapy treatment computing system 110 may collect and obtain data via a network and communicate with other systems using one or more communication interfaces that are communicatively coupled to the processing circuitry 112 and the memory device 114. For example, the communication interface may provide a communication connection between the radiation therapy treatment computing system 110 and components of the radiation therapy system (e.g., allowing data to be exchanged with external devices). For example, in some examples, the communication interface may have appropriate interface circuitry with respect to an output device 146 or an input device 148 to connect to a user interface 142, which may be a hardware keyboard, keypad, or touch screen through which a user may input information into the radiation therapy system.

[0056] As an example, the output device 146 may include a display device that outputs: the user interface 142 ( Fig.11 and one or more aspects, visualizations, or representations of medical images, treatment plans, quality assurance test results, and the status of training, generation, validation, or implementation of such plans. Output device 146 may include one or more display screens that display medical images, interface information, treatment plan parameters (e.g., contours, doses, beam angles, markers, graphs, etc.), treatment plans, targets, positioning of targets, and / or tracking of targets, or any information relevant to the user. Input device 148 connected to user interface 142 may be a keyboard, keypad, touch screen, or any type of device through which a user can input information to the radiation therapy system. Alternatively, output device 146, input device 148, and the functionality of user interface 142 may be integrated into a user interface such as a smart phone or tablet computer (e.g., Apple Lenovo Samsung etc.) in a single device.

[0057] In addition, any and all components of the radiation therapy system can be implemented as virtual machines (e.g., via a virtualization platform such as VMWare, Hyper-V, etc.) or stand-alone devices. For example, a virtual machine can be software that functions as hardware. Thus, a virtual machine can include at least one or more virtual processors, one or more virtual memories, and one or more virtual communication interfaces that function collectively as hardware. For example, the radiation therapy processing computing system 110, the image data source 150, or similar components can be implemented as virtual machines or within a cloud-based virtualization environment.

[0058] The image processing logic 120 or other software program may cause the computing system to communicate with the image data source 150 to read images into the memory device 114 and the storage device 116 or to store images or associated data from the memory device 114 or the storage device 116 to and from the image data source 150. For example, the image data source 150 may be configured to store and provide a plurality of images (e.g., 3D MRI, 4D MRI, 2D MRI slice images, CT images, 2D fluoroscopic images, X-ray images, raw data from an MR scan or a CT scan, Digital Imaging and Communications in Medicine (DICOM) metadata, etc.) from an image set in image data 152 hosted by the image data source 150 and obtained from one or more patients or phantoms via an image acquisition device 170 in performing quality assurance testing. The image data source 150 or other database may also store data to be used by the image processing logic 120 when executing a software program that performs image processing operations to analyze image quality and perform validation processes to provide quality assurance metrics or results. Thus, the radiation therapy treatment computing system 110 may obtain and / or receive image data 152 (e.g., 2D MRI slice images, CT images, 2D fluoroscopic images, X-ray images, 3D MRI images, 4D MRI images, etc.) from an image data source 150, an image acquisition device 170, a treatment device 180 (e.g., MRI-Linac), or other information system related to performing radiation therapy or diagnostic operations.

[0059] The image acquisition device 170 can be configured to acquire one or more images of the patient's anatomical structure for an area of ​​interest (e.g., a target organ, a target tumor, or both). Each image, which is typically a 2D image or slice, can include one or more parameters (e.g., 2D slice thickness, orientation, and position, etc.). In an example, the image acquisition device 170 can acquire 2D slices of any orientation. For example, the orientation of the 2D slice can include a sagittal orientation, a coronal orientation, or an axial orientation. The processing circuit system 112 can adjust one or more parameters, such as the thickness and / or orientation of the 2D slice, to include the target organ and / or the target tumor. In an example, the 2D slice can be determined based on information such as a 3D CBCT or CT or MRI volume. When a patient is receiving radiation therapy treatment, such as when using a treatment device 180, such a 2D slice can be acquired by the image acquisition device 170 "near real time" (wherein, "near real time" means acquiring data at least in a few milliseconds or less).

[0060] Figure 2An exemplary image-guided radiation therapy device 202 is shown, which includes: a radiation source such as an x-ray source or a linear accelerator, a couch 216, an imaging detector 214, and a radiation therapy output 204. The radiation therapy device 202 can be configured to emit a radiation beam 208 to provide treatment to a patient. The radiation therapy output 204 can include one or more attenuators or collimators, such as a multi-leaf collimator (MLC).

[0061] Figure 3 An MLC according to the present disclosure is shown. Two blade stacks facing each other are provided, one on either side of the beam, so as to define the beam from opposite sides. Thus, the first stack 10 includes a frame 12 supporting an array of blades 14, and the second stack 20 includes a frame 22 supporting an array of blades 24. Each blade is oriented in a generally vertical manner relative to the beam, with most blades having a small deflection relative to perfect verticality, as will be briefly described. The blades 14, 24 are retained in the frames 12, 22 by ridges extending along the length of the upper and lower edges of the blades, which ridges engage in corresponding channels in the frames so that the blades can slide horizontally backwards (i.e., out of the beam) and forwards (i.e., into the beam). Each blade is driven in a known manner by a suitable motor or other drive means (not shown).

[0062] Return to reference Figure 2, as an example, a patient can be positioned in region 212, supported by treatment couch 216, to receive a radiation therapy dose according to a radiation therapy treatment plan. The radiation therapy output 204 can be mounted or attached to a gantry 206 or other mechanical support. When the couch 216 is inserted into the treatment area, one or more chassis motors (not shown) can rotate the gantry 206 and the radiation therapy output 204 about the couch 216. In an example, the gantry 206 can continuously rotate about the couch 216 when the couch 216 is inserted into the treatment area. In another example, the gantry 206 can rotate to a predetermined position when the couch 216 is inserted into the treatment area. For example, the gantry 206 can be configured to rotate the treatment output 204 about an axis ("A"). Both the couch 216 and the radiation therapy output 204 can be independently moved to other positions around the patient, for example, can be moved in a transverse direction ("T"), can be moved in a lateral direction ("L"), or can be rotated about one or more other axes, such as about a transverse axis (shown as "R"). A controller communicatively connected to one or more actuators (not shown) can control the movement or rotation of the couch 216 to appropriately position the patient in or out of the radiation beam 208 according to the radiation therapy treatment plan. Both the couch 216 and the gantry 204 can be moved independently of each other in multiple degrees of freedom, which allows the patient to be positioned so that the radiation beam 208 can be precisely targeted to the tumor.

[0063] Figure 2 208 (including axis A, axis T, and axis L) can have an origin located at an isocenter 210. The isocenter can be defined as the location where the central axis of the radiotherapy beam 208 intersects the origin of the coordinate axes, such as a location to deliver a prescribed radiation dose to a patient or a location within the patient. Alternatively, the isocenter 210 can be defined as the location where the central axis of the radiotherapy beam 208 intersects the patient for various rotational positions of the radiotherapy output 204 about axis A as positioned by the gantry 206.

[0064] The gantry 206 may also have an attached imaging detector 214. The imaging detector 214 is preferably positioned opposite the radiation source (radiotherapy output 204), and in an example, the imaging detector 214 may be located within the field of the radiotherapy beam 208. The imaging detector 214 may implement the image processing logic 120 ( Figure 1) to generate images in real time for performing validation process testing. The imaging detector 214 can be mounted on the gantry 206, preferably opposite the radiotherapy output 204, for example to maintain alignment with the radiotherapy beam 208. As the gantry 206 rotates, the imaging detector 214 rotates about the rotation axis. In an example, the imaging detector 214 can be a flat panel detector (e.g., a direct detector or a scintillator detector). In this manner, the imaging detector 214 can be used to monitor the radiotherapy beam 208, or the imaging detector 214 can be used to image the patient's anatomical structure, such as field imaging. The control circuit system of the radiotherapy device 202 can be integrated into the radiotherapy system 100 or remote from the radiotherapy system 100.

[0065] In the illustrative example, one or more of the couch 216, treatment output 204, or gantry 206 may be automatically positioned, and the treatment output 204 may create a radiation therapy beam 208 according to a specified dose for a particular treatment delivery instance. A sequence of treatment deliveries may be specified according to a radiation therapy treatment plan, for example, using one or more different orientations or positions of the gantry 206, couch 216, or treatment output 204. The treatment deliveries may occur sequentially, but may be staggered at desired treatment sites on or within the patient, for example, at the isocenter 210. Thus, a prescribed cumulative dose of radiation therapy may be delivered to the treatment site while reducing or avoiding damage to tissue near the treatment site.

[0066] therefore, Figure 2 Specifically shown is an example of a radiotherapy device 202 that is operable to provide radiotherapy treatment to a patient, the radiotherapy device 202 having a configuration in which a radiotherapy output can rotate about a central axis (e.g., axis "A"). Other radiotherapy output configurations may be used. For example, the radiotherapy output may be mounted to a robotic arm or manipulator having multiple degrees of freedom. In yet another example, the treatment output may be fixed, such as in an area laterally separated from the patient, and a platform supporting the patient may be used to align the radiotherapy isocenter with a designated target site in the patient. In another example, the radiotherapy device may be a combination of a linear accelerator and an image acquisition device. As will be appreciated by those of ordinary skill in the art, in some examples, the image acquisition device may be an MRI, X-ray, CT, CBCT, spiral CT, PET, SPECT, optical tomography, fluorescence imaging, ultrasound imaging, or a radiotherapy portal imaging device, among others.

[0067] In certain embodiments, a phantom is used to evaluate the image quality of an MRI imaging device of an MR-Linac device. That is, a registration process is used to identify any rotation in the coordinate space to verify the successful application of a predetermined region of interest process. In order to evaluate the image quality, an identifiable target in a commercially available phantom can be determined. This MRI phantom has a plurality of circular inserts radially spaced around the axis of a cylinder. Each insert has a plurality of spaced blades or target pins to simulate different spatial frequencies. In the image thus acquired, the angular rotation of the cylinder will be unknown. The object used for registration is embedded in the plane of the body of interest, which has a significantly different structure from the test object. For example, if the test pattern utilizes circular elements, the registration target is a rectangular (or at least surrounded by straight lines) object. It should also be a size or density difference so that when an appropriate pixel value or morphological (shape sensitive) filter is applied to the resulting image, a clear distinction between the registration target and the test feature is obtained. Once the pixel coordinates of the registration target are determined, the orientation of the test feature can be derived using a priori knowledge of the spatial relationship between all features in the phantom.

[0068] In some embodiments, a user can select a validation process to test the MLC leaf and jaw positions. The validation process uses seven electronic portal imaging device (EPID) images of the MRI-Linac device for various MLC and jaw defined field modes. In a given configuration, each of the seven EPID images is captured after the corresponding radiation beam passes through the MLC leaf and jaw.

[0069] The purpose of this verification process is to determine the MLC leaf position accuracy of the center 28 leaves at two nominal positions. In this case, the processing circuit system 112 detects the leaf and jaw positions and indicates whether there are any errors associated with the positions. In some cases, the processing circuit system 112 generates and reports the sum distribution of two picket images, with 0.5mm and 1mm leaf gaps as approximate criteria. In some implementations, the processing circuit system 112 outputs the average radiation isocenter (using the first three of the seven images) and the location of the PEID reference pixel. The processing circuit system 112 can also output the MLC leaf positions of the center 30 leaves (of which there are 60 leaves in total) in each stack at the two nominal positions. The processing circuit system 112 can also output test graphs, including MLC leaf position-leaf number graphs, parameter-parameter graphs, and / or average radiation isocenter x&y coordinate graphs that trend over time.

[0070] In an embodiment, to begin this verification process, the operator enters Fig.11The web tool shown enters parameters that define the MRI-Linac device being tested and optionally the test parameters. After the user enters instructions to begin the verification process, the processing circuit system 112 performs a combination of Figure 4A and Figure 4B A process is described to verify proper operation and alignment of the MLC blade and jaw positions.

[0071] Figure 4A and Figure 4B An illustrative process for testing the MLC and jaw configuration of an MRI-Linac device according to some examples of the present disclosure is provided. Specifically, Figure 4A and Figure 4B 400 and 401 according to an example embodiment. Processes 400 and 401 may be implemented in computer-readable instructions for execution by one or more processors, such that the operations of processes 400 and 401 may be performed in part or in whole by functional components of image processing logic 120; therefore, processes 400 and 401 are described below by way of example with reference to image processing logic 120. However, in other embodiments, at least some of the operations of processes 400 and 401 may be deployed on various other hardware configurations. Therefore, processes 400 and 401 are not intended to be limited to image processing logic 120, and may be implemented in whole or in part by any other component. Some or all of the operations of processes 400 and 401 may be performed in parallel, out of order, or omitted entirely.

[0072] In an embodiment, during verification of the MLC blade and jaw positions, the processing circuitry 112 uses the three images of the acquired 4×4 cm field to estimate the location of the mean radiation field center within the plane of the panel. The image grayscale histogram is divided into two regions using an iterative thresholding technique. The two regions are used to segment the open field in each image. The centers of the segmented regions are determined based on a centroid calculation. A circular function is fitted to the calculated centroid, and the center of the circular fit is taken as the mean field center.

[0073] Specifically, at operation 410, the image processing logic 120 (e.g., using the processing circuit system 112) arranges the MLC and jaw arrangement of the MRI-Linac device in a first configuration. For example, to estimate the position of the center of the average radiation field, the image processing logic arranges the MLC and jaws of the MRI-Linac device so that a small open field is defined only by the MLC. The X jaws stop at 10 cm on each side. The sides of the blades produce field edges in the x-direction. This can eliminate the effect of the jaw calibration on changes in the position of the radiation beam axis. If there is movement of the radiation beam axis as a function of time in the x-direction, this can indicate a change in the MLC stack position relative to the beam source.

[0074] At operation 411, the image processing logic 120 passes the radiation beam three times through the MLC and jaws of the MRI-Linac device in a first configuration. As shown in Tables 1 and 2 below, a first configuration of jaw and MLC leaf positions is defined for each of the three radiation beams:

[0075]

[0076] Table 1: Jaw positions and gantry angles for the MLC blade and jaw position verification process MLC blade positions for images 1 to 3

[0077]

[0078] Table 2: MLC leaf positions for the first three beams during the MLC leaf and jaw position verification process

[0079] At operation 412 , the image processing logic 120 captures three images representing beam shapes detected by the radiation detector of the MRI-Linac device when the MRI-Linac is in the first configuration 510 . Figure 5 The resulting one of three images of a first configuration 510 of jaw and MLC leaf positions of an MRI-Linac device for three beams during the MLC leaf and jaw position verification process is shown. The first configuration 510 enables the image processing logic 120 to estimate the location of the mean radiation field center by arranging the MLC and jaws of the MRI-Linac device so that only a small open field (represented as a black rectangular square in the middle of the first configuration 510) is defined by the MLC.

[0080] At operation 413, the image processing logic 120 calculates the centroid of each of the three images to calculate the radiation field center and detect motion of the beam spot. Specifically, for each of the three images generated when the MLC and blade positions are in the first configuration, the image processing logic 120 segments the open field portion of the image using an iterative thresholding technique. The image processing logic 120 calculates the centroid to determine the radiation field center (average isocenter), and repeats the analysis for the three images to determine if there is any motion of the beam spot. If motion is detected, an indication is provided to the user in the web tool (e.g., by indicating that the verification process has not passed). In an implementation, by delimiting the field in the x-direction using the MLC blades, the position of the radiation axis in the x-direction can be precisely coincident with the joint between blade 40 and blade 41.

[0081] At operation 414, the image processing logic 120 arranges the MLCs and jaws of the MRI-Linac device in a second configuration. For example, the processing circuit system 112 uses an 8×20 cm2 The open field performs the first detection of the MLC and jaw position.

[0082] At operation 415, the image processing logic 120 passes the radiation beam once through the MLC and jaws of the MRI-Linac device in the second configuration. The second configuration of the jaws and MLC leaf positions for the radiation beam is defined as shown in Table 3 below:

[0083]

[0084] Table 3: MLC leaf positions for the fourth beam during the MLC leaf and jaw position verification process

[0085] At operation 416 , the image processing logic 120 captures a fourth image representing the beam shape detected by the radiation detector of the MRI-Linac device. Figure 5 The resulting image of a second configuration 520 of the jaw and MLC leaf positions of the MRI-Linac device for the fourth beam during the MLC leaf and jaw position verification process is shown. The second configuration 520 enables the image processing logic 120 to estimate the four field edge positions and estimate the rotation between the panel and the collimator. The rotation between the MLC and the panel can be determined and compared to a geometry file that specifies the expected rotation to determine if the rotation of the system has changed over time.

[0086] At operation 417, the image processing logic 120 segments the fourth image using an active contour process to estimate the rotation between the radiation detector and the MLC. The active contour process provides an estimate of the positions of the four field edges. Specifically, if the field is truncated or if the field edge is within 3 mm of the image edge, the image processing stops. The active contour process iteratively modifies the points of the initial contour to minimize the internal energy of the contour (represented by the continuity and curvature of the contour) and the image power (represented by the image gradient). During the iterative process, the process moves the contour points toward the strongest gradient while minimizing the curvature of the contour and maintaining regular intervals between the points. In the final solution, the points for which the second-order derivative of the contour exceeds a set threshold are identified as field angles. Once the field angles are located, the remaining contour points can be grouped by the field edge. The field edge is parameterized using a linear equation using points on a given field edge. The process is repeated for the four field boundaries.

[0087] In operation 421 (in Figure 4BContinuing in the figure), image processing logic 120 arranges the MLCs and jaws of the MRI-Linac device in a third configuration corresponding to a picket fence. For example, processing circuit system 112 uses an 8×20 cm extended leaf pattern with a leaf passing through the isocenter. 2 To calculate the rotation between the plate and the blade stroke.

[0088] At operation 422, the image processing logic 120 passes the radiation beam once through the MLC and jaws of the MRI-Linac device in the third configuration. The third configuration of the jaws and MLC leaf positions for the radiation beam is defined as shown in Tables 4 and 5 below:

[0089]

[0090] Table 4: Jaw position and gantry angle for the MLC blade and jaw position verification process

[0091] MCL leaflet positions for image 5

[0092]

[0093]

[0094] Table 5: MLC leaf positions for the fifth beam during the MLC leaf and jaw position verification process

[0095] At operation 423 , the image processing logic 120 captures a fifth image representing the beam shape detected by the radiation detector of the MRI-Linac device. Figure 5 The resulting image shows a third configuration 530 of jaw and MLC leaf positions of the MRI-Linac device for a fifth beam during the MLC leaf and jaw position verification process. The third configuration 530 enables the image processing logic 120 to calculate the rotation between the plate and the leaf stroke.

[0096] The extended blade center in the image is detected using Otsu's method. The location of the extended blade center is then used to determine all blade positions in the x-axis direction, for example using the width of the blade. The extended blade pattern also provides a rotation between the MLC and the plate, which can be compared to a previously stored predetermined expected rotation. Specifically, if there is an angle in the fence or blade pattern captured in the image, the extended blade pattern can indicate a rotation.

[0097] At operation 424 , the image processing logic 120 applies thresholding to detect the center of the extended leaves of the MLC in the third configuration.

[0098] At operation 425, the image processing logic 120 determines all blade positions of the MLC along the x-axis direction based on the position of the center of the extended blade according to the width of each blade. For example, the image processing logic 120 determines the center of each extended blade by starting from the edge of the field and expecting 28 visible blades in the field. That is, the image processing logic 120 can detect the edge of the black box generated from the third configuration 530 and identify the 28 blades between the edges up to the opposite edge. The image processing logic 120 plots a vertical distribution along the blade direction, and their maximum gradient indicates the blade edge. The average angle between the detected extended blades and the imager row is calculated, and a linear equation for the field centerline perpendicular to the blade path and passing through the radiation isocenter is obtained. Based on the angle, the image processing logic determines the rotation between the MLC and the plate, and determines whether such rotation is within the allowed threshold or matches the expected rotation.

[0099] At operation 426, the image processing logic 120 arranges the MLC and jaws of the MRI-Linac device in a fourth configuration and a fifth configuration with respective leaf stacks in an extended position. For example, the processing circuit system 112 uses a 4×20 cm with two adjacent struts. 2 The configuration determines whether the blade position is within or outside an allowable threshold or specification.

[0100] At operation 427, the image processing logic 120 passes the radiation beam once through the MLC and jaws of the MRI-Linac device in each of the fourth and fifth configurations. The fourth and fifth configurations of the jaws and MLC leaf positions for the radiation beam are defined as shown in Tables 6 and 7 below:

[0101]

[0102]

[0103] Table 6: Jaw positions and gantry angles for the MLC blade and jaw position verification process MLC blade position for Image 6

[0104]

[0105]

[0106] MLC blade positions for image 7

[0107]

[0108]

[0109] Table 7: MLC blade positions for beam 6 and beam 7 during MLC blade and jaw position verification

[0110] At operation 428 , the image processing logic 120 captures sixth and seventh images representing beam shapes detected by the radiation detector of the MRI-Linac device. Figure 5 Resulting images of a fourth configuration 541 and a fifth configuration 542 of the jaw and MLC leaf positions of the MRI-Linac device for the sixth beam and the seventh beam, respectively, are shown during the MLC leaf and jaw position verification process.

[0111] At operation 429, the image processing logic 120 detects the positions of the left and right jaws based on the sixth and seventh images. For example, the image processing logic 120 detects the blade positions of all visible blades in the field of view for two different blade positions seen in two different struts (shown in the resulting images of the fourth configuration 541 and the fifth configuration 542). The image processing logic 120 extracts the intensity distribution perpendicular to the horizontal field centerline below each blade and normalizes the distributions to their maximum open field values. The individual blade positions are calculated along the extracted distributions as the distance between the horizontal field centerline and the blade edge (as determined from operation 425). The blade edges are identified using a predefined penumbra detection level, which can be specified using a web tool or hard-coded into the system. The blade positions are converted to millimeters using a pixel-scaling factor. The image processing logic 120 calculates the intensity distribution of the horizontal field centerline perpendicular to the horizontal field centerline below each blade by analyzing the 20×8 cm 2 The image processing logic 120 analyzes the gradient of 11 distributions across the edge of the field per field edge. The location of the edge of the field is determined by averaging the results of the maximum gradient method for these 11 distributions.

[0112] In some embodiments, the image processing logic 120 presents the results of the previous operation in two different tables for the Y1 and Y2 MLC blade stacks. The measured position of each blade at each of the two strut positions may be indicated. In an embodiment, if all measured blade positions are within 1 mm of the nominal blade position, the verification process passes the test. In some embodiments, blade positions outside of this range may be indicated in red in the table. The image processing logic 120 uses the open field image to determine the position of the x jaws (left jaw and right jaw). The overall test results may be presented, and if the jaw position is within 1 mm of the nominal jaw position, the test passes. In some embodiments, jaw positions outside of this range may be indicated in red in the table.

[0113] At operation 430, the image processing logic 120 generates a composite image by summing the sixth image and the seventh image to determine whether the blade position is within the allowed threshold. For example, a distribution can be obtained across two struts (shown in the resulting images of the fourth configuration 541 and the fifth configuration 542) under each blade and averaged to provide a standardized basis. The height of the peaks and valleys at the field junction depends on the blade gap at the field junction. The tolerance line on the graph can be used as an approximate criterion for 0.5mm and 1mm blade gaps. The graph can provide visual verification of the tabulated results. In some cases, if the distribution peak exceeds the tolerance line, the corresponding blade position can be indicated in the tabulated results as being outside the 1mm tolerance.

[0114] Figures 6 to 9 An illustrative process for testing an MRI-Linac device according to some examples of the present disclosure is provided. Specifically, Figures 6 to 9 is a flowchart illustrating exemplary operations of the image processing logic 120 when performing processes 600 to 900 according to an exemplary embodiment. Processes 600 to 900 may be implemented in computer-readable instructions for execution by one or more processors, such that the operations of processes 600 to 900 may be performed in part or in whole by functional components of the image processing logic 120; therefore, processes 600 to 900 are described below by way of example with reference to the image processing logic 120. However, in other embodiments, at least some of the operations of processes 600 to 900 may be deployed on various other hardware configurations. Therefore, processes 600 to 900 are not intended to be limited to the image processing logic 120, and may be implemented in whole or in part by any other component. Some or all of the operations of processes 600 to 900 may be performed in parallel, out of order, or omitted entirely.

[0115] In some embodiments, a user can select a validation process to test the radiation source and EPID movement as a function of gantry angle using a MV geometry phantom. The validation process uses 8 images of the MV phantom captured at 8 different gantry angles using an 8×8 cm field. In some cases, the phantom includes a ball bearing (BB) located at the isocenter. These 8 images are acquired at gantry angles of 180 degrees, 135 degrees, 90 degrees, 45 degrees, 0 degrees, -45 degrees, -90 degrees, and -135 degrees. Similarly, any other suitable additional or different angles may be used.

[0116] The purpose of this verification process is to detect and report the change in the position of the radiation source and EPID reference pixels as a function of the gantry angle. In addition, the 3D position of the phantom is detected and reported, and the maximum displacement of the radiation axis relative to the center of the mean radiation field is reported to the user. In an embodiment, Fig.11 The web tool shown outputs the central ball bearing position in 3D space and the maximum displacement of the radiation axis relative to the average radiation isocenter. The test also outputs perpendicular to the rotation axis: the position of the x reference pixel, the displacement of the radiation axis relative to the radiation isocenter at each angle, and the constancy of the average radiation axis relative to the x reference pixel (sensitive to changes in the MLC leaf stack position). The test also outputs parallel to the rotation axis: the position of the y reference pixel and the constancy of the average radiation isocenter relative to the reference pixel (sensitive to changes in mechanical alignment and asymmetric changes in the MLC leaf stack alignment). In some cases, a graph illustrating the ball bearing center coordinates versus the gantry angle is generated and provided to the user. Prior to testing (e.g., when the MRI-Linac device is installed), a reference pixel is determined in the megavolt (MV) panel coordinate system corresponding to the average isocenter of the MRI-Linac device.

[0117] In an embodiment, to begin the verification process, the operator Fig.11 The web tool shown enters parameters that define the MRI-Linac device being tested and optionally the test parameters. After the user enters instructions to begin the verification process, the processing circuit system 112 performs a combination of Figure 6 Describes the process to verify radiation source and EPID movement.

[0118] Specifically, at operation 601 , the image processing logic 120 passes a radiation beam eight times through a phantom located within the isocenter of an MRI-Linac device at eight different gantry angles.

[0119] At operation 602, the image processing logic 120 captures eight images of the phantom representing the beam shape detected by the radiation detector of the MRI-Linac device for each of the eight radiation beams. In some cases, the images are loaded from a storage device as they may have been previously captured and provided to the image processing logic 120. The configuration of the jaw and MLC leaf positions and the gantry angle for each of the eight radiation beams are defined as shown in Table 8 below:

[0120]

[0121]

[0122] MLC blade positions for radiation beam source testing

[0123]

[0124] Table 8: Jaw and blade positions and gantry angles for the radiation source and EPID movement validation process

[0125] At operation 603, the image processing logic 120 detects field edges for each image using gradient processing. For example, by detecting the field edges in each direction of the open field, the field center is determined in each image.

[0126] At operation 604, the image processing logic 120 calculates the mean of the field edges to determine the location of the radiation beam axis and determines the mean field center. Using the mean field center for each image minimizes the effects of field asymmetry.

[0127] At operation 605, the image processing logic 120 detects the location of the ball bearing in the phantom in each of the eight images. Specifically, the center of the ball bearing is determined using a standardized cross-correlation routine based on previously captured reference images.

[0128] At operation 606 , the image processing logic 120 calculates the distance between the mean field center in the image and the location of the ball bearing for each gantry angle.

[0129] At operation 607, the image processing logic 120 applies the calculated distance and gantry angle to a function that fits the data points along a graph representing the projection of the distance between the isocenter of the radiation and the location of the ball bearings. The function represents the projection of the distance between the location of the ball bearings and the isocenter of the radiation for an MV source that describes a circular orbit about the gantry rotation axis.

[0130] At operation 608, the image processing logic 120 determines whether the MRI-Linac device is operating within the effective range based on the curve graph. Specifically, the image processing logic 120 determines whether the distance between the mean field center in the calculated image and the position of the ball bearing exceeds an allowable threshold. In such a case, the image processing logic 120 can indicate that the MRI-Linac device is not operating within the effective range. The image processing logic 120 can compare the generated curve graph with the expected curve graph to determine whether the difference between the curve graphs exceeds the allowable threshold, thereby determining whether the device is operating within the effective range. In some cases, the image processing logic 120 can compare the position of the mean field center with a reference pixel to determine whether the difference exceeds the allowable threshold. In such a case where the threshold is exceeded, the image processing logic 120 determines that the device is not operating within the effective range.

[0131] In some embodiments, the image processing logic 120 generates a plurality of graphs including the following as an execution Figure 6 The results of the operation: BB center coordinate-gantry angle curve; field center coordinate-gantry angle curve; radiation axis coordinate-gantry angle curve; average radiation axis coordinate-time; and a curve diagram of the maximum displacement of the axis relative to the average radiation isocenter changing with time.

[0132] In some embodiments, the user can select a validation process to test the output flatness symmetry as a function of gantry angle. The validation process uses 4 open field images captured at 8 different gantry angles using an 8×20 cm field. The test calculates and reports the output, flatness, and symmetry for the four images. The 4 images are acquired at gantry angles of 180 degrees, 90 degrees, 0 degrees, and -270 degrees. Similarly, any other suitable additional or different angles may be used.

[0133] The purpose of this verification process is to determine the flatness and symmetry in both the radial and transverse directions. Relative measurements of the output are calculated based on the central portion of each field. In an embodiment, Fig.11 The web tool shown outputs this information. The test also outputs the relative output in the central 2×2 cm area of ​​the panel. The web tool presents graphs including: Flatness (Radial & Transverse) over time; Symmetry (Radial & Transverse) over time, Output-Gantry Angle; and Average Output over time.

[0134] In an embodiment, to begin the verification process, the operator Fig.11 The web tool shown enters parameters that define the MRI-Linac device being tested and optionally the test parameters. After the user enters instructions to begin the verification process, the processing circuit system 112 performs a combination of Figure 7 The process described is to verify flatness symmetry. In an embodiment, a user may specify a region of interest (eg, a percentage of open field) for which testing is to be performed.

[0135] At operation 701, the image processing logic 120 passes the radiation beam four times through the MLC of the MRI-Linac device at four different gantry angles. As shown in Table 9 below, the configuration of the jaw and MLC leaf position is defined for each of the four radiation beams:

[0136]

[0137]

[0138] Table 9: Jaw position and gantry angle for the verification process for flatness symmetry vs. gantry angle

[0139] At operation 702 , the image processing logic 120 captures four images representing the beam shapes detected by the radiation detector of the MRI-Linac device for each of the four radiation beams.

[0140] At operation 703 , the image processing logic 120 uses gradient processing for each image to determine the location of the field edges.

[0141] At operation 704, the image processing logic 120 calculates the mean of the field edge positions to determine the coordinates of the field center. For example, the active contour algorithm iteratively modifies the points of the initial contour to minimize the internal energy of the contour (represented by the continuity and curvature of the contour) and the image force (represented by the image gradient). During the iterative process, the algorithm moves the contour points toward the strongest gradient while minimizing the curvature of the contour and maintaining regular spacing between the points. In the final solution, the points for which the second-order derivative of its contour exceeds a set threshold are identified as field corners. Once the field corners are located, the remaining contour points can be grouped by the field edge. The field edge is parameterized using the points on a given field edge using a linear equation. This process is repeated for four field boundaries. The field center is calculated as the midpoint of the four field boundaries found using the active contour algorithm. Additional information on the rotation of the field edges can also be provided.

[0142] At operation 705, the image processing logic 120 creates an average vertical distribution based on a central distribution of the x-coordinate along the center of the field, four vertical distributions from one side of the central distribution, and four vertical distributions from the other side of the central distribution. For example, the image processing logic 120 creates the average distribution from nine vertical distributions, including the central vertical distribution containing the x-coordinate and four vertical distributions on either side of the central distribution.

[0143] At operation 706, the image processing logic 120 calculates flatness / symmetry based on the average vertical distribution. Specifically, flatness / symmetry indicates how similar the radiation intensity is between and across different regions of interest. In an embodiment, 80% of the maximum open field intensity is determined centered on the center of the field. The image processing logic 120 calculates or determines the change in intensity values ​​relative to the center line and determines the level of equality between adjacent lines relative to the center line.

[0144] At operation 707 , the image processing logic 120 creates an average horizontal distribution based on a central distribution along the y-coordinate of the center of the field, four horizontal distributions from one side of the central distribution, and four horizontal distributions from the other side of the central distribution.

[0145] At operation 708, the image processing logic 120 calculates flatness / symmetry based on the average horizontal distribution. Specifically, flatness / symmetry indicates how similar the radiation intensity is between and across different areas of interest. In an embodiment, 80% of the maximum open field intensity is determined centered on the center of the field. The image processing logic 120 calculates or determines the variation of the intensity values ​​relative to the center line and determines the level of equality between adjacent lines relative to the center line. In an embodiment, if the intensity variation between adjacent distributions and / or across specified areas is greater than an allowed threshold, the image processing logic 120 generates an output indicating whether the test passed or failed.

[0146] After calculating the average horizontal and vertical distribution flatness / symmetry, a 2×2 cm square area is segmented with the field center coordinate as the center. The mean of the intensity is obtained from the inside of the square area. Specifically, the flatness and symmetry in the lateral (x) direction and the radial (y) direction are calculated based on the flattened filter (FF) beam and the flattened filter-free (FFF) beam. For the FF beam, the flatness is calculated as the percentage ratio of the maximum intensity point to the minimum intensity point in the region of interest. The symmetry is calculated as the percentage ratio of the mirror point relative to the center. In the 2×2 cm centered at the field center coordinate, the average intensity of the square area is obtained. 2 The output is calculated in segments of the square area. The mean of the intensity readings inside the square area is taken and the mean is normalized to give the output. Depending on how the device normalization factor is defined, the user can enter a conversion factor as a parameter to normalize the output. In an embodiment, an indication is provided to the user whether the device passed or failed, or a warning can be generated.

[0147] In some embodiments, the user can select a verification process to test the alignment of the MR to the MV. This allows, for example, regular verification of the MR to MV transformation. The results of this test are displayed in a Fig.11The web tool shown is provided to the user in the web tool and includes the transformation and the root mean square of the match between the MR and MV systems after the transformation is applied. In an embodiment, this test is performed using a phantom.

[0148] In an embodiment, to begin the verification process, the operator Fig.11 The web tool shown enters parameters that define the MRI-Linac device being tested and optionally the test parameters. After the user enters instructions to begin the verification process, the processing circuit system 112 performs a combination of Figure 8 A process is described to verify the alignment of MR to MV.

[0149] Specifically, a phantom is placed in the MRI-Linac device at operation 801. For example, an operator may select a phantom and appropriately position the phantom in the MRI-Linac device.

[0150] At operation 802 , the image processing logic 120 measures the transform and the RMS.

[0151] At operation 803 , the image processing logic 120 displays the rotation, translation, and RMS with allowed tolerances.

[0152] In some embodiments, the user can select a quality assessment process to test the MRI image quality of the MRI-Linac device's MRI. The results of this test are displayed in a format such as Fig.11 The web tool shown is provided to the user in the web tool and includes verification of the following: scanning parameters, position and size information of the phantom, rotation of the phantom relative to the image space, slice thickness accuracy, high contrast spatial resolution information, slice position accuracy, image uniformity percentage (PIU), signal artifact percentage (PSG), low contrast detectability (LCD) and geometric accuracy information of the MRI device of the MRI-Linac device.

[0153] In an embodiment, to initiate one or more quality assessment processes, an operator may initiate a Fig.11 The web tool shown inputs parameters defining the MRI-Linac device being tested and optionally inputs test parameters. After the user enters instructions to begin the quality assessment process, the processing circuit system 112 executes the combined Fig. 9 A process is described to verify whether MRI-Linac satisfies one or more quality metrics.

[0154] At operation 901, the image processing logic 120 obtains a plurality of images representing different slices of a phantom that includes at least one module positioned within an MRI-Linac device. For example, the processing circuit system 112 may instruct the MRI-Linac device to capture one or more MRI images of a phantom positioned within the MRI-Linac device. The phantom may include a plurality of modules (physical elements), each module being designed to provide a different quality assessment measurement. The modules may be detected in different slices of the captured one or more MRI images. In some embodiments, a total of 11 slices may be obtained from one or more images of the phantom.

[0155] In some embodiments, a test is performed to find the position of the phantom relative to the axial field of view and also verify the size of the phantom. In some implementations, the output of this test is used to perform one or more additional quality assessment processes. Specifically, the test is performed to find the distance between the center of the phantom slice and the center of the image in the x and y directions, and to find the height and width of the phantom image in millimeters. Operations 902 to 904 are performed to find the position of the phantom relative to the axial field of view and verify the size of the phantom. The output of operations 902 to 904 are the coordinates of the center pixel of the first and last slices of the phantom, the coordinates of the center pixel in the image depicting the phantom, a quantification of the misalignment of the phantom in each image in terms of x and y coordinates, and a quantification of the height and width of the phantom in the first slice (and optionally, additional slices).

[0156] In particular, at operation 902 , the image processing logic 120 selects a first slice corresponding to a side of a phantom from a plurality of images. Fig. 10A The first slice 1010 selected at operation 902 is shown. The processing circuit system 112 uses the Otsu method to calculate a threshold for the selected first slice 1010. The processing circuit system 112 converts the grayscale image of the first slice 1010 into a binary image using the calculated threshold so that only the phantom has pixel values ​​of unit value. Specifically, the processing circuit system 112 processes the grayscale image of the first slice 1010 in the following manner, that is, any pixel value that appears to exceed a certain threshold is set to a first value (e.g., 0), and any other pixels are set to a second value (e.g., 1). In this way, the processing circuit system 112 can perform measurements and calculations only on the portion of the image slice that depicts the phantom.

[0157] After converting the image of the first slice 1010 into a binary image, the processing circuit system 112 identifies the column numbers of the first pixel and the last pixel of the phantom. Specifically, the processing circuit system 112 identifies the first column of pixels in the first slice 1010 that include non-zero binary values. The processing circuit system 112 also identifies the last column of pixels in the first slice 1010 that include non-zero binary values. Then, the processing circuit system 112 identifies a set of rows that include non-zero binary values ​​in each column between the first column and the last column and including the first column and the last column. The output of this step produces the outline of the phantom. In some embodiments, the first 5 rows of pixels and the last 5 rows of pixels are skipped in this identification process because there may be some noise or white spots along the top and bottom edges of the image. The output of this process is in rows (Ph rows ) and columns (Ph cols ) coordinate set, row (Ph rows ) and columns (Ph cols ) has a binary non-zero value representing a phantom.

[0158] After identifying the contour of the phantom in the binary image, the processing circuit system 112 retrieves the grayscale image of the first slice 1010 and maps the contour of the phantom from the binary image to the grayscale image. The processing circuit system 112 maps the contour of the phantom from the binary image to the grayscale image based on the rows (Ph rows ) and columns (Ph cols ) to create a region of interest (ROI) around the phantom in the grayscale image.

[0159] At operation 903, the image processing logic 120 identifies the centroid of the first slice depicted in the first slice. For example, the processing circuit system 112 converts the y coordinate (Y ph ) is calculated as the mean of the row values ​​(row value locations) found for the ROI coordinates of the grayscale image of the first slice 1010 (e.g., the mean of the row values ​​for the portion of the image depicting only the phantom that includes the ROI of the phantom). The processing circuitry 112 converts the x-coordinate (X) of the center of the phantom to ph ) is calculated as the mean of the column values ​​(column value positions) found for the ROI coordinates of the grayscale image of the first slice 1010 (e.g., the mean of the column values ​​of the portion of the image depicting only the phantom including the ROI of the phantom). In some embodiments, the image processing logic 120 repeats operations 901 to 903 (e.g., calculating the x and y coordinates of the center of the phantom) for multiple image slices (e.g., image slice 5 and image slices 7 to 11) in addition to the first image slice.

[0160] At operation 904, image processing logic 120 determines an offset of the phantom based on the difference between the identified center of mass and the true center of mass. For example, processing circuit system 112 calculates the midpoint of the image space (e.g., the midpoint of the entire image including first slice 1010). In an embodiment, to calculate the midpoint of the entire image space, processing circuit system 112 divides the column and row header values ​​by 2, respectively, thereby outputting x-coordinates and y-coordinates (x, y, and y, respectively). sp ,Y sp ). Specifically, processing circuit system 112 finds the maximum column number and divides it by 2 and finds the maximum row value and divides the row value by 2 to output the x-coordinate and y-coordinate. Next, processing circuit system 112 calculates the absolute difference between the center of the phantom and the center of the image space in the x-direction and the y-direction in pixels to determine the offset of the midpoint of the phantom relative to the midpoint of the image space. The result of this calculation is provided as X ph diff pix =|X sp -X ph | and Y ph diff pix =|Y sp -Y ph |.

[0161] In some implementations, processing circuitry 112 converts these differences to millimeters using a pixel pitch (P0) value that is pre-calculated and provided using image header information. In some embodiments, P0(1) represents the height of a pixel and P0(2) represents the width of a pixel. In such a case, the offset value in millimeters is calculated as X ph diff mm =X ph diff pix *P0(2) and Y ph diff mm =Y ph diff pix *P0(1).

[0162] At operation 905, the image processing logic 120 calculates the height of the phantom in pixels. For example, the processing circuit system 112 calculates the height of the phantom in pixels as Ph rows Similarly, the processing circuit system 112 calculates the width of the phantom in pixels as Ph colsThe processing circuit system 112 then converts the width and height of the phantom in pixels to millimeters according to the following expressions: (width of the phantom in pixels)*P0(2); (height of the phantom in pixels)*P0(1).

[0163] Image processing logic 120 (e.g., from Fig.11 The image processing logic 120 retrieves the real physical values ​​of the phantom being used (using the information provided by the user of the web tool shown). Specifically, the image processing logic 120 obtains the actual physical height and width of the phantom placed within the MRI-Linac device (e.g., 190 mm). The image processing logic 120 compares the calculated height and width (dimensions) of the phantom obtained from the image with the actual height and width of the phantom entered by the user (or read from the RFID code in the phantom). The image processing logic 120 determines whether the difference between the calculated dimensions of the phantom and the actual dimensions of the phantom exceeds an allowed threshold. In some cases, the user may enter a value in the RFID tag. Fig.11 In response to determining that the difference exceeds the allowable threshold, the image processing logic 120 Fig.11 The operator or user is alerted in the web tool shown; otherwise, the image processing logic 120 notifies the operator that MRI-Linac has passed the image quality assessment process.

[0164] In some embodiments, the image processing logic 120 repeats the steps of calculating the size of the phantom and comparing the calculated size to the actual size of the phantom for one or more additional slices (e.g., slice 5 for one or more images of the captured phantom). In some cases, the user selects which slices to perform the comparison between the calculated size and the actual size on. In some cases, the user can prioritize the slices so that if it is determined that the calculated size of a slice exceeds the actual size by more than a threshold, if the slice is associated with a high priority, the user is warned that the difference exceeds the threshold. If the slice for which the calculated size exceeds the actual size by more than a threshold is associated with a low priority, the user is not warned that the difference exceeds the threshold. In some cases, the size of the phantom is calculated or obtained from more than one image slice and is an average. The calculated average of the size can be compared to the actual size of the phantom. If the difference between the average calculated and actual size of the phantom exceeds an allowed threshold, the user is warned that the difference exceeds the threshold.

[0165] At operation 906, the image processing logic 120 uses the first slice and the first module depicted in the first slice to determine the rotation of the phantom relative to the image space. Specifically, the image processing logic 120 determines the degree of the phantom to help align the ROI (the region of the image slice depicting the phantom) for subsequent testing. The image processing logic 120 finds the rotation of the phantom based on the rotation of the edges of the first physical module 1016 (e.g., the large center rectangle) present in the phantom image slice 1010. For example, Fig. 10A As shown, when the first physical module 1016 is a rectangle, the image processing logic 120 creates a rectangular ROI 1018 of 120 mm width and 20 mm height, and places the rectangular ROI 1018 within the first slice 1010 at the following position: 3 mm or higher in the y direction relative to the phantom center of the first slice 1010. That is, the rectangular ROI 1018 is positioned relative to the phantom center calculated at operation 903. This results in the ROI 1018 being created around the center of the top long edge and the bottom long edge of the large center rectangle. The created rectangular ROI 1018 is placed parallel to the edge of the image space of the first slice 1010. The first physical module 1016 can be any other shape, and in such a case, the image processing logic 120 creates a shape similar to the shape of the first physical module 1016 as the ROI positioned above the first physical module 1016 depicted in the first slice 1010.

[0166] Next, the image processing logic 120 creates a matrix of values ​​of the segmented phantom image within the ROI 1018, which includes the created rectangle. That is, the image processing logic 120 retrieves the pixel values ​​of the pixels in the first slice 1010 that fall within the rectangular ROI 1018, which is placed on top of the first slice 1010 and at the specified position. The image processing logic 120 identifies the matrix row numbers (L 1 ) in the first column of the matrix and the last column of the matrix where the values ​​change from 1 to 0. 1 , L 2 ). Specifically, the first pixel in the columns of the large rectangle at the left and right edges of the ROI 1018 that includes the first physical module 1016. The image processing logic 120 identifies the last pixel in the first and last columns of the matrix (M 1 、M 2 ). The result is output P1(X 1 ,L 1 )、P2(X n ,L 2 )、P3(X 1 ,M 1 ) and P4(X n ,M 2), where n is the number of columns in the ROI matrix. P1, P2, P3, and P4 represent the points where the first physical module 1016 intersects with a rectangular ROI 1018 placed on top of the first physical module 1016 in the first slice 1010.

[0167] Image processing logic 120 determines the slope of the top and bottom lines of the rectangle relative to the axial field of view (m top and m bottom ) to calculate the rotation of the phantom. The slope represents the rotation of the rectangle relative to the axial field of view (alpha), where alpha is calculated in degrees. Specifically, the image processing logic 120 calculates the slope of the top and bottom edges of the first physical module 1016 (e.g., the long edges of the rectangle) relative to the corresponding top and bottom edges of the rectangular ROI, which is placed on top of the first physical module 1016 in the first slice 1010. The slope and alpha are calculated according to the following expressions: m top =(L 2 -L 1 ) / (n-1), m bottom =(M 2 -M 1 ) / (n-1), m average =(m top +m bottom ) / 2, α = tan -1 (mean).

[0168] At operation 907, the image processing logic 120 determines the slice thickness accuracy using the first slice based on the offset of the second module and the phantom depicted in the first slice. For example, the image processing logic 120 determines the slice thickness of the first slice 1010 according to a wire ramp included in the center of the phantom. In an embodiment, when the slice thickness is approximately 0.7 mm, it is determined that the slice thickness passes the quality assessment process.

[0169] In some embodiments, to calculate the slice thickness, the image processing logic 120 creates two ROIs to be placed around the upper line slope and the lower line slope, respectively, based on predetermined values. The image processing logic retrieves the previously calculated phantom center position and the rotation of the phantom to determine the slice thickness using the upper line slope and the lower line slope. In some implementations, the two ROIs are set and placed using set values ​​so that the vertices of the image file relative to the first slice 1010 are: upper left: x=-48.83, y=-10.0; upper right: x=48.83, y=-10.0; lower right: x=-48.83, y=10.0; and lower right: x=48.83, y=10.0, or as specified below:

[0170] tt1x=-48.83 / P0(2); ttly=-4.883 / P0(1)

[0171] tt2x=48.83 / P0(2); tt2y=-4.883 / P0(1)

[0172] tt3x=48.83 / P0(2); tt3y=-0.9766 / P0(1)

[0173] tt4x=-48.83 / P0(2); tt4y=-0.9766 / P0(1)

[0174] tt5x=-48.83 / P0(2); tt5y=2.9298 / P0(1)

[0175] tt6x=48.83 / P0(2); tt6y=2.9298 / P0(1)

[0176] tt7x=48.83 / P0(2); tt7y=-0.9766 / P0(1)

[0177] tt8x=-48.83 / P0(2); tt8y=-0.9766 / P0(1)

[0178] The image processing logic 120 corrects the setting value by converting the setting value from a millimeter value to a pixel value and applying the previously calculated rotation based on the centroid position of the phantom calculated in operation 903. The corrected setting value is referred to as ROI full Next, the image processing logic 120 retrieves the full The image processing logic 120 then trims the ROI by trimming all horizontal lines found in the image processing logic 120 and identifying the first and last rows that consist entirely of pixel values ​​less than the calculated slice threshold. In an embodiment, the calculated threshold is the same as the threshold used in operation 902. If no qualifying rows are found (e.g., if the image processing logic 120 does not identify a row that consists entirely of pixel values ​​less than the calculated threshold), the test fails. Otherwise, the image processing logic 120 trims the ROI full The value is set to be between two rows of pixels (ROI trimmed )between.

[0179] Next, the image processing logic 120 converts the ROI trimmed The row is divided vertically into two ROIs of equal height (ROI upper and ROI lower ). Image processing logic 120 calculates ROI upper and ROI lower The image processing logic 120 selects the ROI with the highest average pixel value. upperRow max-upper ). The image processing logic 120 identifies Row max-upper The highest pixel value in , and divide it by 2P half-max-upper Next, the image processing logic 120 identifies Row max-upper Above P half-max-upper The first and last pixel values ​​of the profile are given in pixels. The difference between the first and last positions provides the upper profile width in pixels. This value is converted to millimeters by multiplying by P0(2) and is used as ProfileWidth upper Output and storage. Image processing logic 120 repeats based on ROI lower ProfileWidth lower Specifically, the image processing logic 120 selects the ROI with the highest average pixel value. lower Row max-lower ). The image processing logic 120 identifies Row max-lower The highest pixel value in , and divide it by 2P half-max-lower Next, the image processing logic 120 identifies Row max-lower Above P half-max-lower The first and last pixel values ​​of the profile are given in pixels. The difference between the first and last positions provides the lower profile width in pixels. This value is converted to millimeters by multiplying by P0(2) and is used as ProfileWidth lower In some embodiments, the image processing logic 120 can be used to output and store Fig.11 The web tool shown in FIG. 1 outputs an image for presentation to the user showing the ROI. upper and ROI lower The image processing logic 120 may also present two graphs (one for the upper line and another for the lower line) showing pixel positions - Row max-upper and Row max-lower The value on P half-max-upper and P half-max-lower .

[0180] The image processing logic 120 calculates the measured slice thickness as: MeasuredSliceThickness = 0.2*((ProfileWidth upper *ProfileWidth lower ) / (ProfileWidth upper *ProfileWidth lower ). Image processing logic 120 (e.g., based on Fig.11The image processing logic 120 determines the slice thickness TagSliceThickness of the phantom being used by receiving user input from the web tool or user interface shown in , or by retrieving information from the header of the file storing one or more images of the phantom. The image processing logic 120 calculates the slice thickness accuracy as SliceThicknessAccuracy = the absolute value of (MeasuredSliceThickness-TagSliceThickness). The image processing logic 120 determines whether the calculated slice thickness accuracy is less than an allowable threshold (e.g., 0.7 mm). The allowable threshold can be set by the user via Fig.11 In response to determining that the slice thickness accuracy is less than the allowed threshold, the image processing logic 120 instructs the MRI-Linac device to pass the image quality assessment process. In response to determining that the slice thickness accuracy is greater than the allowed threshold, the image processing logic 120 (e.g., via Fig.11 The web tool shown in ) warns the user that the MRI-Linac device has not passed the image quality assessment process.

[0181] At operation 908, the image processing logic 120 determines the high contrast spatial resolution based on the offset of the third module and the phantom depicted in the first slice. For example, the image processing logic 120 determines the spatial resolution of the first slice 1010 from the physical resolution pattern ROI 1018 in the phantom. In an embodiment, the high contrast spatial resolution is determined to pass the quality assessment process when the resolution is 1.0 mm or less.

[0182] In an embodiment, to perform this quality assessment process, the image processing logic 120 divides a portion of the image corresponding to the physical resolution pattern ROI 1018 into 6 ROIs. The two ROIs on the left represent a resolution of 1.1 mm, the 2 ROIs in the center represent a resolution of 1.0 mm, and the 2 ROIs on the right represent a resolution of 0.9 mm. Each grouping of 2 ROIs with the same target resolution is read as a horizontal line (upper left) and a vertical line (lower right). The image processing logic 120 determines the lowest resolution at which 4 distinct holes can be found.

[0183] Specifically, to calculate high contrast spatial resolution, the image processing logic 120 creates six ROIs. One ROI is positioned around a series of holes (e.g., physical resolution pattern ROI 1018) in the upper left portion, and one ROI is positioned around a series of holes (e.g., physical resolution pattern ROI 1018) in the lower right portion. The image processing logic 120 creates an ROI for each of the three resolution intervals of 1.1 mm, 1.0 mm, and 0.9 mm regions based on previously set values, previously calculated phantom center positions, and rotations of the phantom. The vertices of the physical resolution pattern ROI 1018 are (in mm): upper left: x=-63.5, y=12.0; upper right: x=61.5, y=12.0; lower left: x=-63.5, y=62.0; and lower right: x=61.5, y=62.0. These vertices are converted from millimeters to pixels, and rotations are then applied to create ROIs. full The image processing logic 120 calculates the ROI using the Otsu method. full Then, the image processing logic 120 identifies the ROI full Point P upper-right , the point P upper-right Physically closest to the ROI full and has a value less than the calculated threshold. upper-right Identify three ROI pairs 1030 ( Fig. 10A ) position. Relative to P upper-right , the upper left portion of each ROI pair is (in mm): ROI pair-1 Upper left: x = -80.14, y = 8.4; ROI pair-2 Upper left: x=-55.64, y=8.4; and ROI pair-2 Upper left: x = -33.64, y = 9.4. These values ​​are converted to pixel coordinates by applying the previously calculated rotation.

[0184] Next, for each ROI pair , the image processing logic 120 removes any horizontal and vertical lines on the four edges of the ROI that are entirely composed of pixels having values ​​less than the calculated threshold. Calculate the relative pair The ROIs of the locations of the holes contain 6 separate well groups. 1 ROI 2 etc.) Fig. 10A 1030. The upper ROI (box 1032) is obtained from the matching ROI pairThe upper left portion of the ROI extends a given distance in height and width, while the lower ROI (box 1033) extends the same distance from the lower right portion. pair-1 9mm, for ROI pair-2 is 8.5 mm, and for ROI pair-3 These values ​​are converted to pixel values ​​by using the previously calculated values ​​for the conversion.

[0185] The image processing logic 120 reads the ROI of each row 1 and find the 4 rows with the highest average pixel values ​​(Line 1 、Line 2 Image processing logic 120 from Line 1 The image processing logic 120 calculates the threshold value for the remaining pixels using the Otsu method and sets the Line 1 The pixels in the line are grouped as "on" (if the pixel has a value greater than a threshold) and "off" (if the pixel is less than or equal to the threshold). The image processing logic 120 counts the number of transitions from on to off and determines the number of identifiable holes in the line. The image processing logic 120 counts the number of transitions from on to off and determines the number of identifiable holes in the line. 2-4 Repeat this process. If 1-4 If the number of holes in any of the lines is greater than four, the image processing logic 120 determines that the ROI 1 The image processing logic 120 processes the ROI 2-6 Repeat the process of finding and counting holes (for ROI 2 ROI 4 and ROI 6 , processing columns instead of rows).

[0186] Each ROI is associated with a size in mm: ROI 1 and ROI 2 Associated with 1.1mm, ROI 3 and ROI 4 is associated with 1.0 mm, and ROI 5 and ROI 6 The horizontal resolution of the phantom is such that more than 4 holes can be successfully generated in a given line ROI. 1,3,5 The vertical resolution of the phantom is the minimum resolution that can successfully create more than 4 holes in a given line ROI. 2,4,6In some embodiments, the image processing logic 120 determines whether the vertical resolution and the horizontal resolution are greater than an allowable threshold (e.g., 1.0 mm). The allowable threshold may be set by the user via Fig.11 In response to determining that the resolution is greater than the allowed threshold, the image processing logic 120 instructs the MRI-Linac device to pass the image quality assessment process. In response to determining that the resolution is less than the allowed threshold, the image processing logic 120 (e.g., via Fig.11 The web tool shown in ) warns the user that the MRI-Linac device has not passed the image quality assessment process.

[0187] Typically, the phantom includes a physical resolution pattern ROI 1018, which allows the image processing logic 120 to calculate the horizontal resolution and vertical resolution of the MRI-Linac device. The physical resolution pattern ROI 1018 includes multiple regions, each of which is configured to enable different levels of resolution to be measured along different dimensions (horizontally and vertically). The image processing logic 120 identifies each of these different regions and determines whether a known pattern can be identified in the image captured from the region. If a pattern or a portion of a pattern can be identified from the image captured from the region, the image processing logic 120 determines that the resolution level of MRI-Linac is at least as good as the corresponding resolution of the region. Otherwise, the image processing logic 120 determines that the resolution level of MRI-Linac cannot meet the resolution level associated with the region. For example, if each region includes a set of holes, the image processing logic 120 determines whether a threshold number of holes in the set of holes in the region can be seen in the image including the region. If a threshold number of holes are visible, the image processing logic 120 determines that the resolution level of MRI-Linac is at least as high as the resolution associated with the particular region. When a region is found in which the threshold number of holes is not found or cannot be identified, the image processing logic 120 determines that MRI-Linac cannot provide the resolution level associated with the region.

[0188] At operation 909, the image processing logic 120 determines the slice position accuracy based on another slice at the opposite end of the phantom and the fourth module depicted in the first slice. For example, the image processing logic 120 determines the slice position accuracy based on the image of the physical strip 1014 in the first slice 1010 and the image of the physical strip 1022 of the eleventh slice 1020 in the phantom. In an embodiment, when the strip length is determined to be 5.0 mm or less, it is determined that the slice position accuracy passes the quality assessment process.

[0189] In an embodiment, to perform this quality assessment process, the image processing logic 120 generates ROIs around the left portion (e.g., the left half of the bar) and the right portion (e.g., the right half of the bar) of the bar at the top of the phantom image in the first and eleventh slices 1010 and 1020, respectively. In an embodiment, this process is performed on the grayscale image of the slice. In some implementations, the ROIs in the left and right portions are positioned at the following coordinates, and corrections based on the rotation and the phantom center (previously calculated) are applied:

[0190] Left ROI:

[0191] PALt1x=-4.883 / P0(2); PALt1y=-71.289 / P0(1);

[0192] PALt2x=-1.953 / P0(2); PALt2y=-71.289 / P0(1);

[0193] PALt3x=-4.883 / P0(2); PALt3y=-51.758 / P0(1);

[0194] PALt4×=-1.953 / P0(2); PALt4y=-51.758 / P0(1);

[0195] Right ROI:

[0196] PARt1x=0.9766 / P0(2); PARt1y=-71.289 / P0(1);

[0197] PARt2x=3.906 / P0(2); PARt2y=-71.289 / P0(1);

[0198] PARt3x=0.9766 / P0(2); PARt3y=-51.758 / P0(1);

[0199] PARt4x=3.906 / P0(2); PARt4y=-51.758 / P0(1);

[0200] Apply corrections for rotation and phantom center coordinates

[0201] PALt11x=((PALt1x*(cosd(alpha)))·(PALt1y*(sind(alpha))))+xPhCntr;

[0202] PALt11y=((PALt1x*(sind(alpha)))+(PALt1y*(cosd(alpha))))+yPhCntr;

[0203] PALt12x=((PALt2x*(cosd(alpha)))·(PALt2y*(sind(alpha))))+xPhCntr;

[0204] PALt12y=((PALt2×*(sind(alpha)))+(PALt2y*(cosd(alpha))))+yPhCntr;

[0205] PALt13x=((PALt3x*(cosd(alpha)))-(PALt3y*(sind(alpha))))+xPhCntr;

[0206] PALt13y=((PALt3×*(sind(alpha)))+(PALt3y*(cosd(alpha))))+yPhCntr;

[0207] PALt14x=((PALt4x*(cosd(alpha)))-(PALt4y*(sind(alpha))))+xPhCntr;

[0208] PALt14y=((PALt4x*(sind(alpha)))+(PALt4y*(cosd(alpha))))+yPhCntr;

[0209] PARt11x=((PARt1×*(cosd(alpha)))-(PARt1y*(sind(alpha))))+xPhCntr;

[0210] PARt11y=((PARt1x*(sind(alpha)))+(PARt1y*(cosd(alpha))))+yPhCntr;

[0211] PARt12x=((PARt2x*(cosd(alpha)))-(PARt2y*(sind(alpha))))+xPhCntr;

[0212] PARt12y=((PARt2x*(sind(alpha)))+(PARt2y*(cosd(alpha))))+yPhCntr;

[0213] PARt13×=((PARt3x*(cosd(alpha)))-(PARt3y*(sind(alpha))))+xPhCntr;

[0214] PARt13y=((PARt3x*(sind(alpha)))+(PARt3y*(cosd(alpha))))+yPhCntr;

[0215] PARt14x=((PARt4x*(cosd(alpha)))-(PARt4y*(sind(alpha))))+xPhCntr;

[0216] PARt14y=((PARt4x*(sind(alpha)))+(PARt4y*(cosd(alpha))))+yPhCntr;

[0217] The image processing logic 120 extracts the distribution parallel to the edge of each ROI and applies Otsu's method to find a single threshold for these values. The image processing logic 120 segments and extracts regions from the image using a single threshold. Next, the image processing logic 120 searches for the boundaries of the strips (e.g., the row numbers where the values ​​change from black to white in the extracted regions) from the top to the bottom of the ROI. For example, the image processing logic 120 searches for region 1024 where the pixel values ​​change from black to white in the ROI around strip 1022. That is, Fig. 10A As shown, it is found that the left portion of the ROI changes from black to white in the area 1024, while the right portion of the ROI of the bar continues to remain black in the area 1024. The image processing logic 120 calculates the average column boundary value for each ROI (e.g., the image processing logic 120 calculates the height of the left portion of the ROI and the height of the right portion of the ROI). The image processing logic 120 calculates the difference between the column values ​​of the left ROI and the right ROI (e.g., the image processing logic 120 determines whether the left portion of the ROI is shorter than or smaller than the right portion of the ROI). If the image processing logic 120 determines that the left portion of the ROI is longer than the right portion, the image processing logic 120 assigns a negative sign to the length.

[0218] The image processing logic 120 converts the calculated difference in column values ​​for the first slice 1010 (e.g., the difference in height of two adjacent ROIs created above the bar 1014 and the bar 1022) to a millimeter value by multiplying the value by the previously calculated constant P0(1). The image processing logic 120 compares the calculated height difference in millimeters to an allowable threshold value (e.g., 5 mm). If the height difference is less than the allowable threshold value, the image processing logic 120 determines that the slice position accuracy of MRI-Linac passes the image quality assessment. The allowable threshold value may be set by the user via Fig.11The web tool shown in , or may be hard-coded into the image processing logic 120. The image processing logic 120 instructs the MRI-Linac device to go through the image quality assessment process. In response to determining that the height difference is greater than the allowed threshold, the image processing logic 120 (e.g., via Fig.11 The web tool shown in ) warns the user that the MRI-Linac device has not passed the image quality assessment process.

[0219] The image processing logic 120 repeats the comparison process of the left and right bar ROI heights for another slice (e.g., the eleventh slice). The image processing logic 120 converts the calculated difference in column values ​​of the eleventh slice 1020 (e.g., the difference in heights of two adjacent ROIs created above bar 1014 and bar 1022) into a millimeter value by multiplying the value by the previously calculated constant P0(1). The image processing logic 120 compares the calculated height difference in millimeters with an allowed threshold value (e.g., 4 mm). The allowed threshold value for the height difference of the eleventh slice 1020 may be less than or greater than the allowed threshold value for the first slice 1010. If the height difference is less than the allowed threshold value, the image processing logic 120 determines that the slice position accuracy of MRI-Linac passes the image quality assessment. The allowed threshold value may be set by the user via Fig.11 The web tool shown in , or may be hard-coded into the image processing logic 120. The image processing logic 120 instructs the MRI-Linac device to go through the image quality assessment process. In response to determining that the height difference is greater than the allowed threshold, the image processing logic 120 (e.g., via Fig.11 The web tool shown in ) warns the user that the MRI-Linac device has not passed the image quality assessment process.

[0220] At operation 910 , the image processing logic 120 determines an image uniformity percentage, a signal artifact percentage, a geometric accuracy, and low contrast detectability based on one or more slices between the first slice and the last slice.

[0221] To determine the image uniformity percentage (PIU) of the MRI-Linac device, the image processing logic performs an image quality assessment process on the slices 7 of one or more images of the phantom. In an embodiment, for an MRI system with a field strength of less than 3 Tesla, when the PIU is determined to be greater than or equal to 87.5%, the PIU is determined to pass the quality assessment process. In an embodiment, for an MRI system with a field strength of 3 Tesla, when the PIU is determined to be greater than or equal to 82.0%, the PIU is determined to pass the quality assessment process.

[0222] The image processing logic 120 retrieves the seventh slice 1040 of the phantom ( Fig. 10B) and identifies the position of the phantom relative to the axial field of view and the x and y coordinates of the center of the phantom. In some implementations, the image processing logic 120 performs operations similar to operations 902 to 904 to determine the coordinates of the center of the phantom that appears in the seventh slice 1040. Next, the image processing logic 120 generates a 200 cm 2 The circular ROI is depicted as (79.789 / P0(1) pixel radius). Fig. 10B The image processing logic 120 identifies the locations of the maximum and minimum pixels within the circular ROI 1042. The image processing logic generates a 1 cm 2 Two circular ROIs 1042 and 1044 (5.642 / P0(1) pixel radius) Fig. 10B ), and each ROI 1042 and 1044 is centered on the corresponding maximum pixel value and minimum pixel value within the larger circular ROI.

[0223] Image processing logic 1040 ensures that the smaller 1cm 2 The circular ROIs 1042 and 1044 are completely within the larger 200cm 2 To do this, the image processing logic 120 checks the distance from the maximum point and the minimum point to the center of the larger circular ROI and verifies that the distance is less than a threshold value (e.g., less than 79.789 / P0(1)-5.642 / P0(1)). If the distance is less than the threshold value, the small circular ROI is within the large circular ROI. If the distance is not less than the threshold value, the image processing logic 120 places the center of the small ROI at a point along the same line (at the same angle) from the center of the large ROI but at a distance of the threshold value from the center of the large circular ROI.

[0224] The image processing logic 120 calculates the average value of the pixels within each of the small ROIs 1042 and 1044 (meanSlice7MaxROI and meanSlice7MinROI). The image processing logic 120 calculates the PIU as follows: PIU=100*(1-((meansSlice7MaxROI-meanSlice7MinROI) / (meanSlice7MaxROI+meanSlice7MinROI))). Specifically, the image processing logic 120 calculates the PIU based on the average value of the pixels within each of the small ROIs 1042 and 1044. The image processing logic 120 retrieves the magnetic strength of the MRI-Linac device used to capture one or more images of the phantom. In an embodiment, when the quality assessment process is initiated, the value is input by a user (e.g., via Fig.11) and stored for retrieval by the image processing logic 120. In some embodiments, the magnetic intensity is encoded in the header of one or more image files of the image of the phantom and retrieved by the image processing logic 120.

[0225] The image processing logic 120 compares the calculated PIU to an allowable threshold (e.g., 87.5% for MRIs with magnetic field strengths less than 3 Tesla, or 82.0% for MRIs with magnetic field strengths of 3 Tesla). If the PIU is greater than the allowable threshold, the image processing logic 120 determines that the PIU of the MRI-Linac passes the image quality assessment. The allowable threshold can be set by the user via Fig.11 The web tool shown in , or may be hard-coded into the image processing logic 120. The image processing logic 120 instructs the MRI-Linac device to go through the image quality assessment process. In response to determining that the PIU is less than the allowed threshold, the image processing logic 120 (e.g., via Fig.11 The web tool shown in ) warns the user that the MRI-Linac device has not passed the image quality assessment process.

[0226] In some embodiments, the image processing logic 120 generates a graphic showing a large circular ROI and a small circular ROI for the maximum and minimum positions. Fig.11 The web tool shown presents the graphic to the user.

[0227] To determine the signal artifact percentage (PSG) of the MRI-Linac device, the image processing logic performs an image quality assessment process on slices 7 of one or more images of the phantom. In some cases, the image processing logic 120 uses the generated large circular ROI to calculate the PIU to determine the PSG. In an embodiment, when it is determined that the PSG is less than or equal to 0.025, it is determined that the PSG passes the quality assessment process.

[0228] To determine the PSG, the image processing logic 120 calculates the average value of the pixels inside the large circular ROI that is generated to calculate the PIU. The image processing logic 120 generates Fig. 10B The four elliptical ROIs shown in the image 1050 in FIG. The four elliptical ROIs are generated with a 10 cm 2 The ellipse may have an aspect ratio of about 4:1. In an embodiment, the ellipse is positioned along but not against the edge of the field of view and not against or against the edge of the phantom. The center of the ellipse is between the edges of the phantom. In some cases, the image processing logic 120 reduces the width of the ellipse to achieve this result and increases the length to maintain the specified area.

[0229] Next, the image processing logic 120 calculates the distance between the edge of the phantom and the edge of the field of view for each of the four edges. For each top edge, the image processing logic 120 checks the minimum row value based on the delineation of the segmented phantom image. For the bottom edge, the image processing logic 120 subtracts the maximum row value based on the delineation of the phantom image from the total number of rows obtained from the head of the image of the seventh slice. For the left edge, the image processing logic 120 checks the minimum column value based on the delineation of the phantom. For the right edge, the image processing logic 120 subtracts the maximum column value from the total number of columns in the image. The result of these checks outputs the number of pixels between the edge of the phantom and the edge of the field of view for each edge. The image processing logic 120 ensures that these values ​​are less than a threshold value (e.g., 17.842 / P0(2))+6 pixels (3 on each side) to ensure that the ROI does not touch the edge of the phantom or the edge of the field of view.

[0230] When the image processing logic 120 determines the space available for placement of the elliptical ROI, the radii of the major and minor axes of the ellipse are set. To set these radii, the image processing logic 120 determines whether the distance is less than a threshold value (e.g., 17.842 / P0(2)), and if the distance is less than the threshold value, the minor (minor) axis radius of the ellipse is set to w_pxl=8.921 / P0(2), and the major (major) axis radius is set to l_pxl=35.681 / P0(1). In the event that the value of the distance is greater than the threshold value+6 pixels, the minor axis radius is set to (distance+6 pixels) divided by 2, and the major axis radius is set to (318.31 / (P0(1)×P0(2))) divided by the minor axis radius to maintain the area of ​​the ROI.

[0231] Next, the image processing logic 120 calculates the average pixel value within each of the four elliptical ROIs (meanTopROI is the average of the top ellipse, meanBtmROI is the average of the bottom ellipse, meanLftROI is the average of the left ellipse, and meanRgtROI is the average of the right ellipse). The image processing logic 120 calculates the artifact ratio based on these averages according to the following formula: Artifact Ratio=absolute value of (((meanTopROI+meanBtmROI)-(meanLftROI+meanRgtROI)) / (2*meanLgROI)).

[0232] The image processing logic 120 compares the calculated artifact ratio with an allowable threshold (e.g., 0.025). If the artifact ratio is less than the allowable threshold, the image processing logic 120 determines that the artifact ratio of MRI-Linac passes the image quality assessment. The allowable threshold can be set by the user via Fig.11The web tool shown in , or may be hard-coded into the image processing logic 120. The image processing logic 120 instructs MRI-Linac to go through the image quality assessment process. In response to determining that the artifact ratio is greater than the allowed threshold, the image processing logic 120 (e.g., via Fig.11 The web tool shown in ) warns the user that the MRI-Linac device has not passed the image quality assessment process.

[0233] In some embodiments, the image processing logic 120 generates a graphic showing four elliptical ROIs and a large circular ROI at the top, bottom, left, and right sides of the phantom. Fig.11 The web tool shown presents the graphic to the user.

[0234] To determine the low contrast detectability (LCD) of the MRI-Linac device, the image processing logic performs an image quality assessment process on slices 9, 10, and 11 of one or more images of the phantom. In an embodiment, for an MRI system with a field strength of less than 3 Tesla, the LCD is determined to pass the quality assessment process when the image processing logic 120 calculates the total score of 9 spokes. In an embodiment, for an MRI system with a field strength of 3 Tesla, the LCD is determined to pass the quality assessment process when the image processing logic 120 calculates the total score of 37 spokes.

[0235] For slice 8, the image processing logic 120 obtains the thresholds calculated in operations 902 to 905. Specifically, the thresholds are calculated to find the position of the phantom relative to the axial field of view and the size of the phantom. Next, the image processing logic 120 searches both horizontally and vertically from the center of the image to find the x- and y-coordinates where the pixel value is less than the threshold. This produces P left , P right , P top , P bottom The image processing logic 120 performs the left and P right The x value and P top and P bottom The y values ​​are averaged to find the center point (P center ). center The image processing logic 120 calculates the mean and standard deviation of the pixel values ​​in a circular area with a diameter of 85 mm. The image processing logic 120 applies the window level (mean) and the window width (mean × standard deviation × 0.003) to generate an image of the slice for use in (e.g., via Fig.11 For example, the image processing logic 120 outputs Fig. 10B1060 shown in . The user reviews the image and counts the number of spokes observed and enters the number. Image processing logic 120 repeats the process for each of slices 9 to 11 or generates an image of the slice for display to the user. Image processing logic 120 stores the number of spokes entered by the user as visible on slices 8 to 11.

[0236] The image processing logic 120 compares the stored number of spokes to an allowable threshold (e.g., 9 spokes for MRIs with magnetic field strengths less than 3 Tesla, or 37 spokes for MRIs with magnetic field strengths of 3 Tesla). If the number of spokes is greater than the allowable threshold, the image processing logic 120 determines that the LCD of the MRI-Linac passes the image quality assessment. The allowable threshold may be set by the user via Fig.11 The web tool shown in , or may be hard-coded into the image processing logic 120. The image processing logic 120 instructs the MRI-Linac device to go through the image quality assessment process. In response to determining that the LCD is less than the allowed threshold, the image processing logic 120 (e.g., via Fig.11 The web tool shown in ) warns the user that the MRI-Linac device has not passed the image quality assessment process.

[0237] To determine the geometric accuracy (spatial linearity) of the MRI-Linac device, the image processing logic performs an image quality assessment process on slice 5 of one or more images of the phantom. In some cases, the image processing logic 120 obtains the previously calculated width and height of the first slice 1070 and the height and diagonal dimensions of the fifth slice 1072 to determine the geometric accuracy. In an embodiment, when the width and height of the first slice 1070 and the height and diagonal dimensions of the fifth slice 1072 are within 2.0 mm of the baseline (190.0 mm), it is determined that the geometric accuracy passes the quality assessment process.

[0238] For the first slice 1070, the image processing logic 120 scales the image with a scaling factor of 16 using Lanczos interpolation in an 8×8 pixel neighborhood. This enables sub-pixel accuracy at various sizes. The image processing logic 120 begins scanning horizontally from left to right and vertically from top to bottom to find the first pixel with a value less than the threshold for the first slice 1070. The length of the line from the leftmost pixel to the rightmost pixel and the length of the line from the top pixel to the bottom pixel are the scaled width and height of the phantom. The image processing logic 120 divides the value of the line length by 160.0 to reverse the scaling. For the fifth slice 1072, the image processing logic 120 repeats the process of scanning horizontally from left to right and vertically from top to bottom to find the first pixel with a value less than the threshold for the fifth slice 1072. The length of the line from the leftmost pixel to the rightmost pixel and the length of the line from the top pixel to the bottom pixel are the scaled width and height of the phantom in the fifth slice 1072. The image processing logic 120 divides the value of the length of the line calculated for the fifth slice 1072 by 160.0 to reverse the scaling.

[0239] Next, the image processing logic 120 repeats the process of calculating a line for each diagonal (45 degrees) through the COM of the slice. The image processing logic 120 converts all of the lines to millimeters using the previously calculated values. The image processing logic 120 then determines whether the calculated length meets an allowable threshold (which may be user provided). If the length is within 2.0 millimeters of the allowable threshold, the image processing logic 120 determines that the geometric accuracy of the MRI-Linac passes the image quality assessment. The image processing logic 120 instructs the MRI-Linac device to pass the image quality assessment process. In response to determining that the length is outside the allowable threshold, the image processing logic 120 (e.g., via Fig.11 The web tool shown in ) warns the user that the MRI-Linac device has not passed the image quality assessment process.

[0240] In some embodiments, if there is a sagittal view slice, the image processing logic 120 determines the geometric accuracy of the sagittal view slice. The length should be within 2.0 mm of the ideal length of 148.0 mm. The image processing logic 120 finds the threshold of the sagittal view slice and divides it by 2. The image processing logic 120 creates a binary image of pixels with values ​​greater than the threshold. The image processing logic 120 scans each row in the binary image and fills any internal blank space with a width of 20 pixels or less. Based on the modified binary image, the image processing logic 120 calculates the COM of the slice. The image processing logic 120 reads the pixel column from the top to the bottom of the binary image at COM x and 2 columns on either side (a total of 5 columns) to find the top and bottom "on" pixels. The difference between the top and the bottom is the length of the column in pixels. The image processing logic 120 adds the results and divides the sum by 5 to provide the length of the lost surface in pixels. The image processing logic 120 converts the value into millimeters using the previously determined factor or value. In an embodiment, the image processing logic 120 outputs an image representing the result.

[0241] Fig.11 An exemplary web-based interface 1100 for testing an MRI-Linac device according to some examples of the present disclosure is shown. In some embodiments, a user accesses the web-based interface 1100 via the Internet by accessing a website. The web-based interface 1100 includes an option 1110 to enter MRI-Linac device parameters, an option 1120 to upload radiation detector images and / or MRI images, an option 1130 to specify thresholds and allowable tolerances, an option 1140 to select a verification process, an option 1150 to indicate that the MRI-Linac device is selected, an option 1160 to select an imaging quality assessment process, and a pass / fail report and image processing result display 1170.

[0242] In response to receiving a user selection of the input MRI-Linac device parameters option 1110, a prompt is provided to the user that allows the user to enter any of the parameters discussed above for generating images of the MRI-Linac device for testing. For example, the user can specify parameters for the MLC and jaws, such as leaf size and width, and can specify the magnetic field strength of the MRI-Linac device. The user can also specify various parameters of the phantom used to perform imaging quality assessment and other validation processes (e.g., the number and type and location of physical modules in one or more phantoms).

[0243] In some embodiments, the user can choose to upload a radiation detector image and / or an MRI image by selecting option 1120. Specifically, any image discussed above for performing a validation process or quality assessment can be pre-captured, stored, and uploaded by the user. After the user uploads the image and / or after capturing the image, the user can specify an allowable threshold for the MRI-Linac device, which is used to determine whether the MRI-Linac device passes or fails a given validation process or quality assessment. Therefore, the user selects option 1130. In response, a list of validation processes and quality assessments and default test thresholds are presented. The user can modify any default values, and the image processing logic 120 uses the modified values ​​to determine whether a given test passes or fails.

[0244] The user can select a verification process to be performed by selecting option 1140. In response to receiving the user selection of option 1140, the web-based interface 1100 presents a list of available verification processes to the user. For example, the web-based interface 1100 presents an interactive list including a position verification process of the MLC, a radiation source and imaging detector movement verification process, a linear accelerator and MRI imaging device alignment verification process, and a linear accelerator flatness symmetry verification process. The user can select any one or all of the listed processes to be performed on the uploaded image or on the newly captured image of the MRI-Linac device.

[0245] The user can select an imaging quality assessment process (or imaging quality assessments) to perform by selecting option 1160. In response to receiving a user selection of option 1160, the web-based interface 1100 presents a list of available imaging quality assessments to the user. For example, the web-based interface 1100 presents an interactive list that includes quality assessment processes for checking rotation of the phantom relative to image space, calculating slice thickness accuracy, calculating high contrast spatial resolution, checking slice position accuracy, calculating image uniformity percentage, calculating signal artifact percentage, calculating low contrast detectability, or calculating geometric accuracy. The user can select any one or all of the listed processes to perform on an uploaded image or on a newly captured image of the MRI-Linac device.

[0246] In some embodiments, when a given image set has not been previously captured and is available for the user to upload, the user may select option 1150 to instruct the MRI-Linac device to perform a given operation to capture an image set suitable for the procedure(s) selected by options 1140 and 1160. In some embodiments, the web-based interface 1100 instructs the operator how to position the phantom and / or MLC to capture appropriate images for a given procedure. In some embodiments, the web-based interface 1100 communicates with the MRI-Linac device to automatically set the MLC of the MRI-Linac device to a configuration suitable for capturing images for a given procedure.

[0247] After capturing or uploading an image, the web-based interface 1100 presents the results of all the processes selected by the user using options 1140 and 1160 in a results display 1170. The results display 1170 may include any graphics generated by the image processing logic 120. The user may select any of the displayed results to obtain additional information about why a given process or test passed or failed.

[0248] Additional Notes

[0249] As previously discussed, each electronic computing system or device can implement one or more of the methods or functional operations discussed herein. In various embodiments, such an electronic computing system or device operates as a standalone device or can be connected (e.g., networked) to other machines. For example, such a computing system or device can operate with the capabilities of a server or client machine in a server-client network environment, or operate as a peer machine in a peer-to-peer (or distributed) network environment. The features of a computing system or device can be implemented by a personal computer (PC), a tablet PC, a personal digital assistant (PDA), a cellular phone, a web device, or any machine capable of executing instructions (sequentially or otherwise) specifying the action to be performed by the machine.

[0250] As also explained above, the functionality discussed above may be implemented via instructions, logic, or other information storage on a machine-readable medium. Although a machine-readable medium may have been described in various examples with reference to a single medium, the term "machine-readable medium" may include a single medium or multiple media (e.g., a centralized or distributed database and / or associated cache memory and servers) storing one or more transient or non-transient instructions or data structures. The term "machine-readable medium" should also be considered to include any tangible medium that is capable of storing, encoding, or carrying transient or non-transient instructions for execution by a machine and causing the machine to perform any one or more of the methods of the present disclosure, or that is capable of storing, encoding, or carrying data structures utilized by or associated with such instructions.

[0251] The above specific embodiments include reference to the accompanying drawings, which form a part of the specific embodiments. The accompanying drawings show the specific embodiments in which the present disclosure can be put into practice by way of explanation rather than by way of limitation. These embodiments are also referred to as "examples" in this article. Such examples may include elements other than those shown or described. However, the present disclosure is also intended to provide only examples of those elements shown or described. In addition, the present disclosure is also intended to use any combination or arrangement of those elements (or one or more aspects of those elements) shown or described with respect to a specific example (or one or more aspects of a specific example) or with respect to other examples (or one or more aspects of other examples) shown or described in this article.

[0252] All publications, patents, and patent documents referenced in this document are incorporated herein by reference in their entirety, as if individually incorporated by reference. If there are inconsistent usages between this document and those documents incorporated by reference, the usage in the incorporated reference(s) should be considered supplementary to that of this document; for conflicting inconsistencies, the usage in this document shall prevail.

[0253] In this document, when introducing elements of various aspects of the present disclosure or embodiments thereof, the terms "a", "an", "the", and "said" are used, as is common in patent documents, to include one or more than one or more of the elements, independent of any other instances or uses of "at least one" or "one or more". In this document, unless otherwise indicated, the term "or" is used to represent a non-exclusive or, such that "A or B" includes "A but not B", "B but not A", and "A and B".

[0254] In the following claims, the terms "including" and "in which" are used as the plain-English equivalents of the respective terms "comprising" and "wherein." Further, in the following claims, the terms "including," "comprising," and "having" are intended to be open-ended, to mean that additional elements may be present in addition to the listed elements, such that anything following such a term (e.g., including, comprising, having) in a claim is still considered to fall within the scope of the claim. Moreover, in the following claims, the terms "first," "second," and "third," etc. are used merely as labels, and are not intended to impose numerical requirements on their objects.

[0255] The present disclosure also relates to a computing system adapted, configured or operated to perform the operations herein. The system may be specially constructed for the desired purpose, or the system may include a general-purpose computer selectively started or reconfigured by a computer program (e.g., instruction, code, etc.) stored in a computer. Unless otherwise specified, the order of operation or execution of the operations in the embodiments of the present disclosure shown and described herein is not necessary. That is, unless otherwise specified, operations may be performed in any order, and compared with the operations disclosed herein, the embodiments of the present disclosure may include other operations or fewer operations. For example, it is expected that the following falls within the scope of various aspects of the present disclosure: before another operation, simultaneously with another operation, or after another operation, a specific operation is run or performed.

[0256] In view of the above, it will be seen that the several objects of the present disclosure are achieved and other advantageous results are obtained. Having described aspects of the present disclosure in detail, it will be apparent that modifications and variations are possible without departing from the scope of aspects of the present disclosure as defined in the appended claims. Since various changes may be made to the above-described constructions, products and methods without departing from the scope of various aspects of the present disclosure, it is intended that all that is contained in the above description and shown in the accompanying drawings should be interpreted in an illustrative and not restrictive sense.

[0257] The examples described herein can be implemented in a variety of embodiments. For example, one embodiment includes a computing device that includes processing hardware (e.g., a processor or other processing circuit system) and memory hardware (e.g., a storage device or volatile memory), the memory hardware including instructions implemented thereon, so that the instructions, when executed by the processing hardware, cause the computing device to implement, perform, or coordinate electronic operations for these technologies and system configurations. Another embodiment discussed herein includes a computer program product that can be implemented, for example, by a machine-readable medium or other storage device, and the computer program product provides transient or non-transient instructions for implementing, performing, or coordinating electronic operations for these technologies and system configurations. Another embodiment discussed herein includes a method that can operate on the processing hardware of a computing device to implement, perform, or coordinate electronic operations for these technologies and system configurations.

[0258] In other embodiments, the logic, commands, or transient or non-transient instructions implementing various aspects of the above-described electronic operations may be provided in a distributed or centralized computing system including any number of form factors for computing systems such as desktop computers or notebook personal computers, mobile devices such as tablet computers, netbooks, and smart phones, client terminals, and server-hosted machine instances. Another embodiment discussed herein includes incorporating the techniques discussed herein into other forms, including other forms of programming logic, hardware configurations, or dedicated components or modules, including devices having various components that perform the functions of such techniques. Various algorithms for implementing the functions of such techniques may include sequences of some or all of the above-described electronic operations or other aspects depicted in the accompanying drawings and in the above specific embodiments.

[0259] The above description is intended to be illustrative rather than restrictive. For example, the examples described above (or one or more aspects of the examples) may be used in combination with each other. In addition, without departing from the scope of the present disclosure, many modifications may be made to adapt specific situations or materials to the teachings of the present disclosure. Although the dimensions, material types and example parameters, functions and implementations described herein are intended to limit the parameters of the present disclosure, they are by no means restrictive embodiments, but exemplary embodiments. After browsing the above description, many other embodiments will be apparent to those skilled in the art. Therefore, the scope of the present disclosure should be determined with reference to the appended claims together with the full scope of equivalents enjoyed by these claims.

[0260] In addition, in the above detailed description, various features can be combined together to organize the present disclosure. This should not be interpreted to mean that: for any claim, the disclosed features that are not claimed for protection are all necessary. Instead, the subject matter of the invention may lie in less than all the features of a particular disclosed embodiment. Therefore, the attached claims are hereby incorporated into the detailed description, wherein each claim exists by itself as an independent embodiment. The scope of the present disclosure should be determined by reference to the attached claims together with the full scope of equivalents to which these claims are entitled.

Claims

1. A method for verifying the operation of a radiotherapy apparatus comprising an integrated linear accelerator and an MRI imaging device MR-Linac, the method comprising: selecting a validation process associated with the MR-Linac linear accelerator; Based on the selected verification process, retrieving configuration information for the multi-leaf collimator MLC of the MR-Linac; obtaining a plurality of images from an imaging detector, each image associated with a given radiation beam of a plurality of radiation beams that have passed through the MLC that has been adjusted based on the configuration information; processing the plurality of images to determine a characteristic of the MR-Linac associated with the selected verification procedure, wherein processing the plurality of images comprises: calculating a centroid of each of a first image, a second image, and a third image in the plurality of images to determine a radiometric isocenter associated with each of the first image, the second image, and the third image; and detecting motion in the plurality of radiation beams associated with the first image, the second image, and the third image based on the determined differences in radiation isocenters associated with each of the first image, the second image, and the third image; and A determination is made based on processing the plurality of images as to whether the MR-Linac is operating within a valid range.

2. The method according to claim 1, further comprising: The determined characteristics are compared to allowable thresholds associated with the selected authentication process, and in response to determining that the MR-Linac is operating outside of the valid range, one or more characteristics of the MR-Linac are adjusted.

3. The method according to claim 2, wherein: The one or more characteristics are determined based on the selected authentication process.

4. The method according to any one of claims 1 to 3, further comprising: In response to the retrieval of the configuration information and the selection of the verification process, the MLC is automatically adjusted.

5. The method according to any one of claims 1 to 3, wherein: The selected verification process is one of a plurality of verification processes, the plurality of verification processes comprising a blade position verification process of the MLC, a radiation source and imaging detector movement verification process, a linear accelerator and MRI imaging device alignment verification process, and a linear accelerator flatness symmetry verification process.

6. The method according to any one of claims 1 to 3, wherein: The selected validation process comprises a blade position validation process, wherein the plurality of images comprises at least seven images acquired at a same gantry angle of the MR-Linac, wherein the determined characteristics comprise blade and jaw positions relative to nominal blade and jaw positions, and wherein an allowable threshold associated with the selected validation process comprises one millimeter.

7. The method according to claim 5, wherein: The configuration information indicates first positions of a first set of blades and a second set of blades for the MLC, and second positions of a third set of blades for the MLC between the first set of blades and the second set of blades, and processing the plurality of images further comprises: segmenting the first image, the second image, and the third image; calculating the centroid for each of the segmented first image, the segmented second image, and the segmented third image; Fitting a circular function to the calculated centroid; determining the radiometric isocenter associated with each of the first image, the second image, and the third image based on the center of the circular function; and The motion in the plurality of radiation beams is detected based on the determined difference in radiation isocenters associated with each of the first image, the second image, and the third image.

8. The method according to claim 7, wherein: processing a fourth image of the seven images to identify an open field, wherein the configuration information indicates an open position for a blade of the MLC, the method further comprising, for each edge in the open field of the fourth image: generating an initial contour based on pixel intensity values ​​of the open field edge; modifying points of the initial contour to minimize the internal energy of the contour; moving the points towards the strongest gradient and maintaining regular spacing between the points; and The second derivative of the profile is calculated to identify the corner of the open field edge.

9. The method according to claim 8, wherein: processing a fifth image of the seven images to detect extended blades of the MLC, wherein the configuration information indicates a first position for a first group of blades and a second group of blades of the MLC and alternates between indicating a second position and a third position for a third group of blades between the first group and the second group, the method further comprising: identifying a center of each of the third set of blades based on an edge of the field in the fifth image; and Blade edges of the third set of blades are identified.

10. The method according to claim 9, wherein: processing a sixth image and a seventh image of the seven images to detect blade and jaw positions, wherein, for the sixth image, the configuration information indicates a first position for a first blade stack of the MLC and a second position for a second blade stack of the MLC, and wherein, for the seventh image, the configuration information indicates the first position for a second blade stack of the MLC and the second position for a first blade stack of the MLC, the method further comprising: generating intensity distributions of the sixth image and the seventh image; identifying a blade position based on a distance between a horizontal centerline of the field in the sixth image and the seventh image and the identified blade edge; and The jaw position is identified based on field edge distributions of the sixth image and the seventh image.

11. The method according to any one of claims 1 to 3, wherein: The selected validation process comprises a linear accelerator flatness symmetry validation process, wherein the plurality of images comprises four images acquired at four different gantry angles of the MR-Linac, wherein the determined characteristics comprise symmetry and flatness of the images, and wherein the allowable threshold associated with the selected validation process comprises a specified percentage value.

12. The method according to claim 11 further comprises performing the following operations for each image: determining locations of field edges of fields in the image; averaging the edge positions of the field to determine the center of the field; Generates an average distribution from nine vertical distributions, where A middle distribution of the nine vertical distributions passes through the center of the field; identifying a position in the nine vertical distributions corresponding to 80% intensity; determining a difference between a first vertical distribution and a second vertical distribution of the nine vertical distributions that are equidistant from the middle distribution to determine symmetry; as well as The variation between the values ​​of the nine vertical distributions is determined to determine the flatness.

13. A method for determining the image quality of a radiotherapy apparatus comprising an integrated linear accelerator and an MRI imaging device MR-Linac, the method comprising: selecting a quality assessment process associated with the MR-Linac MRI imaging device; obtaining a plurality of images of a phantom including at least one module from the MRI imaging device; processing features in the plurality of images representing components in the at least one module based on the selected quality assessment process; calculating an image quality attribute associated with the MRI imaging device based on the processed features; determining whether the image quality attributes meet acceptance criteria; and In response to determining that the image quality attribute satisfies the acceptance criteria, determining whether the MR-Linac is operating effectively, The quality assessment process is a first quality assessment process for calculating the position of the phantom relative to the axial field of view and calculating the size of the phantom.

14. The method according to claim 13, further comprising: In response to determining that the image quality attribute does not meet the allowable criteria, one or more characteristics of the MR-Linac are adjusted.

15. The method according to claim 13 or 14, further comprising: selecting a first image from the plurality of images; Binarizing the first image so that the phantom displayed in the image has a uniform pixel value; identifying the column numbers of the first pixel and the last pixel of the phantom in each row of pixels; generating a region of interest (ROI) around the phantom based on the identified column number; identifying a centroid of the ROI based on the identified column number; identifying a center of the first image; calculating a difference between the coordinates of the phantom and the center of the first image to determine a misalignment of the phantom; as well as The height and width of the phantom are calculated based on the ROI.

16. The method according to claim 13 or 14, wherein: The selected quality assessment process also includes one of multiple quality assessment processes, and the multiple quality assessment processes include at least one of the quality assessment processes for performing the following operations: checking the rotation of the phantom relative to the image space, calculating slice thickness accuracy, calculating high contrast spatial resolution, checking slice position accuracy, calculating image uniformity percentage, calculating signal artifact percentage, calculating low contrast detectability or calculating geometric accuracy.

17. A method for performing a quality assessment on a radiotherapy apparatus comprising an integrated linear accelerator and an MRI imaging device MR-Linac, the method comprising: selecting a validation process associated with the MR-Linac's linear accelerator and a quality assessment process associated with the MR-Linac's MRI imaging device; obtaining from an imaging detector a first plurality of images, each image being associated with a given radiation beam of a plurality of radiation beams that have passed through a multi-leaf collimator MLC of the MR-Linac that has been adjusted based on the selected validation procedure; obtaining a second plurality of images of a phantom including at least one module from the MRI imaging device; processing the first plurality of images based on the selected verification process, wherein processing the first plurality of images comprises: calculating a centroid of each of a first image, a second image, and a third image in the first plurality of images to determine a radiometric isocenter associated with each of the first image, the second image, and the third image; and detecting motion in the plurality of radiation beams associated with the first image, the second image, and the third image based on the determined differences in radiation isocenters associated with each of the first image, the second image, and the third image; processing the second plurality of images based on the selected quality assessment process; and Based on attributes of the first and second pluralities of images determined from processing the first and second pluralities of images, it is determined whether the MR-Linac is operating within a valid range.

18. The method of claim 17, further comprising adjusting one or more characteristics of the MR-Linac in response to determining that the MR-Linac is operating outside of the valid range.

19. The method according to claim 18, wherein: The one or more characteristics are determined based on a selected validation process or quality assessment process.

Images