Magnetic resonance imaging apparatus, method and program

Abstract

To properly partition a scan sequence to process a hardware instruction.SOLUTION: A magnetic resonance imaging apparatus according to an embodiment includes a processing circuit. The processing circuit is configured to acquire a scan sequence for performing magnetic resonance scan; partition the scan sequence into a plurality of kernels by determining partition time points defining boundaries of the plurality of kernels in the scan sequence, wherein the partition time points are not equally spaced in time; convert the one kernel of the plurality of kernels into a hardware instruction; transmit the hardware instruction to a hardware controller for execution; and reconstruct a magnetic resonance image from received data, including data obtained by executing the hardware instruction.SELECTED DRAWING: Figure 9

Description

Technical Field

[0001] The embodiments disclosed in this specification and the drawings relate to magnetic resonance imaging apparatuses, methods, and programs.

Background Art

[0002] Medical imaging generates images of internal organs and tissues of a patient's body. For example, magnetic resonance imaging (MRI) uses radio waves, magnetic fields, and magnetic field gradients to generate images of internal organs and tissues. Once these images are generated, a doctor can use the images to diagnose a patient's trauma or illness.

[0003] An MRI apparatus performs various functions, including converting a scan sequence (high-level sequence description) into hardware instructions (low-level hardware instructions), attempting to correct system incompleteness, receiving acquired data and storing the data for subsequent reconstruction processing, and applying corrections based on physical characteristics to hardware instructions for controlling other hardware components of an MR system, including a gantry, a magnet, and the like.

[0004] Here, a conventional MRI apparatus executes real-time sequence control software that interprets a scan sequence as one continuous stream consisting of a plurality of events and continuously calculates hardware instructions from the continuous stream of the plurality of events. Here, the correction based on the physical characteristics described above is executed "on the fly" by one or more field programmable gate arrays (FPGAs) immediately before the hardware instructions are transmitted to the hardware for execution. Such a conventional design is suitable for applications that conform to a fixed physical characteristic model that can be expressed in a simple manner using an FPGA and require low latency and timing guarantee.

[0005] However, as mentioned above, when calculating hardware instructions sequentially from a continuous stream of multiple events, high temporal precision is required for the sequence control software. Furthermore, with FPGA-based methods, it is difficult to immediately perform the corrections based on the physical characteristics described above, and it is also difficult to adapt to variable physical characteristic models. For this reason, in the field of MRI devices, there is a need for technology that appropriately divides the scan sequence and processes hardware instructions. [Prior art documents] [Patent Documents]

[0006] [Patent Document 1] U.S. Patent Application Publication No. 2020 / 0003859 [Patent Document 2] U.S. Patent Application Publication No. 2020 / 0025850 [Patent Document 3] U.S. Patent Application Publication No. 2020 / 0363485 [Patent Document 4] U.S. Patent Application Publication No. 2018 / 0267123 [Overview of the Initiative] [Problems that the invention aims to solve]

[0007] One of the problems that the embodiments disclosed herein and in the drawings aim to solve is to appropriately divide the scan sequence and process hardware instructions. However, the problems that the embodiments disclosed herein and in the drawings aim to solve are not limited to the above problem. Problems corresponding to the effects of each configuration shown in the embodiments described later can also be positioned as other problems.

[0008] The above description of "Background Art" is intended to provide an overview of the background to this disclosure. The inventors' research, to the extent described in this Background Art section, is not considered prior art to the present invention, either explicitly or implicitly, similar to the aspects of this specification that cannot constitute prior art at the time of filing. [Means for solving the problem]

[0009] The magnetic resonance imaging apparatus according to the embodiment includes a processing circuit. The processing circuit acquires a scan sequence for performing a magnetic resonance scan, divides the scan sequence into the plurality of kernels by determining division time points that define the boundaries of the plurality of kernels at non-equally spaced time intervals within the scan sequence, converts one of the plurality of kernels into a hardware instruction, transmits the hardware instruction to a hardware control unit for execution, and reconstructs a magnetic resonance image from acquired data including data obtained by executing the hardware instruction. [Brief explanation of the drawing]

[0010] By reviewing the following detailed description along with the attached drawings, and by referring to this description to better understand the disclosure, you will be able to easily grasp a full understanding of the disclosure and the many benefits associated with it.

[0011] [Figure 1] This figure shows a scan control method according to one embodiment of the present disclosure. [Figure 2] This diagram shows an automated method for dividing a scan sequence into kernels. [Figure 3] This figure shows an example of a single-slice FSE scan sequence. [Figure 4] This figure shows an example of a 5-slice FSE scan sequence. [Figure 5] This figure shows an example of kernel delimiter insertion based on the position of an RF excitation pulse according to one embodiment of the present disclosure. [Figure 6]This figure shows an example of kernel delimiter insertion based on the location of RF refocus pulses and readout events according to one embodiment of the present disclosure. [Figure 7] This figure shows an example of kernel delimiter insertion based on the position of an RF inversion pulse according to one embodiment of the present disclosure. [Figure 8] This figure shows a medical imaging system configured to perform an improved method according to an exemplary embodiment of the present disclosure. [Figure 9] This is a schematic block diagram of an MRI device relating to an exemplary implementation of the present disclosure. [Modes for carrying out the invention]

[0012] The embodiments of the magnetic resonance imaging apparatus, method, and program will be described in detail below with reference to the drawings.

[0013] This embodiment relates to an improved magnetic resonance imaging apparatus.

[0014] According to one embodiment, the present disclosure relates to a magnetic resonance imaging apparatus. The magnetic resonance imaging apparatus includes a processing circuit. The processing circuit acquires a scan sequence for performing a magnetic resonance scan, divides the scan sequence into a plurality of kernels by determining division time points that define the boundaries of a plurality of kernels at non-equally spaced time intervals within the scan sequence, converts one of the plurality of kernels into a hardware instruction, transmits the hardware instruction to a hardware control unit for execution, and reconstructs a magnetic resonance image from acquired data including data obtained by executing the hardware instruction.

[0015] According to one embodiment, the present disclosure relates to a method of performing magnetic resonance imaging. The method includes obtaining a scan sequence for performing a magnetic resonance scan, dividing the scan sequence into the plurality of kernels by determining division time points that define boundaries of the plurality of kernels at non-uniform time intervals within the scan sequence, converting one kernel included in the plurality of kernels into hardware instructions, transmitting the hardware instructions to a hardware control unit for execution, and reconstructing a magnetic resonance image from the collected data including the data obtained by executing the hardware instructions.

[0016] According to one embodiment, the present disclosure relates to a program for performing magnetic resonance imaging. The program causes a computer to execute procedures for obtaining a scan sequence for performing a magnetic resonance scan, dividing the scan sequence into the plurality of kernels by determining division time points that define boundaries of the plurality of kernels at non-uniform time intervals within the scan sequence, converting one kernel included in the plurality of kernels into hardware instructions, transmitting the hardware instructions to a hardware control unit for execution, and reconstructing a magnetic resonance image from the collected data including the data obtained by executing the hardware instructions.

[0017] Note that the above paragraph presents an introduction and is not intended to limit the appended claims. The embodiments disclosed in the present application will be best understood with additional advantages by referring to the following detailed description in conjunction with the accompanying drawings.

[0018] As used herein, the term “one” is defined as one or more. The term “multiple” is defined as two or more. The term “other” is defined as at least two. The terms “including” and / or “having” are defined as constituting (i.e., open language). Throughout this specification, any use of “one embodiment,” “a particular embodiment,” “embodiment,” “implementation,” “example,” or similar terms means that a particular characteristic, structure, or feature described in relation to this embodiment is included in at least one embodiment of this application. Therefore, not all such lexical expressions found in various places throughout this specification necessarily refer to the same embodiment. Furthermore, a particular characteristic, structure, or feature may be combined in one or more embodiments in any suitable manner, without limitation.

[0019] Exemplary embodiments are described as methods having predetermined steps. However, such methods and components operate effectively with additional steps and steps in different sequences that are not inconsistent with the exemplary embodiments. Therefore, this disclosure is not limited to the following embodiments, but provides the broadest scope consistent with the principles and features described herein, and is limited only by the appended claims.

[0020] Furthermore, where a range of values ​​is indicated, it should be understood that each value between the upper and lower limits of that range, and any other specified or intervening values ​​within the stated range, are included in this disclosure. Also, if the stated range includes an upper and lower limit, the range excluding either the upper or lower limit is also included. Unless otherwise explicitly stated, the terms used herein are intended to have an obvious and general meaning as understood by those skilled in the art. Any definitions provided are intended to aid the reader's understanding of this disclosure, but are not intended to alter or limit the meaning of such terms unless otherwise explicitly stated.

[0021] Conventional MRI systems can present various problems. Firstly, the continuous streaming of multiple events in a scan sequence places a high demand on the sequence control software, requiring the sequence interpretation program to maintain microsecond accuracy and synchronization for long periods, sometimes exceeding 10 minutes. Secondly, conventional designs make it difficult to quickly adapt physical property-based corrections to input real-time sensor data (e.g., temperature-based feedback). Similarly, introducing updated or new physical property models (e.g., physical property models for unique clinical applications) is challenging and may require more powerful FPGA processing capabilities. While it is theoretically possible to implement fast-adaptable and updated physical property models using FPGAs, such functionality increasingly requires more powerful FPGAs and / or specialized development skills.

[0022] Therefore, for example, in an MRI system, it is conceivable to divide the scan sequence into multiple linked "chunks," each corresponding to a fixed period of, say, 10 milliseconds. Here, a chunk is a unit of division when dividing large data, and represents a single set of data. When a scan sequence is divided into multiple fixed-length chunks in this way, the scan sequence may be divided within the ecotrain, resulting in the ecotrain being divided into multiple chunks, which can be problematic. For example, if correction or adjustment is performed on the chunks within the ecotrain, the generation of the physical properties of spin may be hindered. Therefore, in order to avoid spin generation errors, the boundaries between chunks must be controlled with high temporal precision.

[0023] In this embodiment, the scan sequence is divided into chunks called "kernels." Here, a kernel means a finite portion of the scan sequence. Specifically, the division time points that define the boundaries of multiple kernels are arranged at generally non-equally spaced but physically significant locations. Here, each kernel represents a consistent arrangement of spin physical characteristics, such as (1) from the excitation RF (Radio Frequency) pulse to the end of the echo train, (2) from the inversion pulse to the excitation RF pulse, and (3) a group of pre-pulses (e.g., T2-prep). Generally, a kernel may also mean a pattern of consecutive events that are repeatedly executed in the scan sequence. A chunk may also mean a set of events that are easily processed by hardware. In such cases, for example, a kernel may be considered to correspond to one chunk as described above, or it may be considered to contain multiple chunks.

[0024] In one embodiment, the boundaries between kernels are automatically determined using a definable heuristic method based on the physical properties of the MR, for example, based on spin excitation, RF pulse position, or termination of the sequence module. Alternatively, the boundaries may be determined manually (e.g., by a pulse sequence engineer) or by a combination of automatic and manual methods.

[0025] Once the boundaries between kernels are determined, the scan control unit processes the kernels one by one in sequence. Specifically, the scan control unit may adaptively apply corrections based on the physical characteristics of each individual kernel by "pre-fetching" each kernel (rather than processing the scan sequence in a single continuous flow). In this way, some or all of the adjustment or correction of the scan sequence, which was conventionally performed by the FPGA, may be performed by sequence control software in a "just-in-time" manner.

[0026] Figure 1 shows a method 100 for performing scan control in an MRI apparatus in an embodiment of the present disclosure. The method in Figure 1 is performed, for example, by the processing circuit of the MRI sequence control unit 984 shown in Figure 9.

[0027] In step 110, the scan sequence is received. For example, Figure 3 shows an example of a one-slice FSE scan sequence including an RF inversion pulse, an RF excitation pulse, and an ecotrain.

[0028] In step 120, the scan sequence is divided into multiple kernels by determining appropriate time boundaries (division time points) for each kernel within the scan sequence based on its physical characteristics. For example, the time boundaries between kernels are automatically determined by a computer program running on the processing circuit using a definable heuristic method, based on the physical characteristics of the MR, for example, based on the position of spin excitations or RF pulses. Figure 2 shows a flowchart of method 200 for dividing the scan sequence into kernels, which will be explained in detail later.

[0029] In step 130, a sequence correction is calculated and applied only to the current kernel in the sequence. Such a correction is made, for example, based on the current state of sensor data received continuously or periodically in step 125. For example, input sensor data related to at least one of temperature data, motion correction, and physiological gating is used when the sequence correction is calculated in step 130.

[0030] In step 140, the corrected kernel is converted into hardware instructions. Note that the order in which steps 130 and 140 are performed, as shown in Figure 1, can be changed. Specifically, corrections based on physical characteristics may be performed after the kernel has been converted into hardware instructions.

[0031] In step 150, the hardware instruction is transmitted to an appropriate hardware control board for execution. Here, the hardware control board is an example of a hardware control unit.

[0032] In step 160, it is determined whether or not other kernels exist within the scan sequence being executed. If other kernels exist, the process proceeds to step 170. In step 170, the next kernel is acquired, and then step 130 is performed on the newly acquired kernel. On the other hand, if no other kernels exist, the process ends.

[0033] Figure 2 shows a method 200 for dividing a sequence into kernels according to the method of this disclosure.

[0034] In step 210, after the scan sequence from step 110 has been received, the positions of the RF excitation pulses within the scan sequence are identified. For example, Figure 4 shows a 5-slice FSE scan sequence containing 5 RF excitation pulses.

[0035] In step 220, multiple kernels are generated by inserting kernel delimiters at each RF excitation pulse location identified in step 210. Each of these kernels is then further subdivided into smaller kernels, as will be described later. For example, as shown in Figure 5, the initial multiple kernels are generated by inserting five kernel delimiters at each RF excitation pulse location.

[0036] In step 230, the locations of the RF refocus pulse and readout event are identified within each kernel included in the multiple kernels generated in step 220.

[0037] In step 240, within each of the multiple kernels generated in step 220, a kernel delimiter is inserted after the last readout event following the last RF refocus pulse identified within that kernel. This further subdivides the kernel. If no RF refocus pulse is identified within the kernel, no kernel delimiter is inserted in step 240.

[0038] Specifically, as shown in Figure 6, five additional kernel delimiters are inserted based on the position of the last read event in each of the initial multiple kernels.

[0039] In step 250, the location of the RF inversion pulse is identified in each of the existing kernels.

[0040] In step 260, for each RF inversion pulse location identified in step 250, a kernel delimiter is inserted at the location of the RF inversion pulse if there is no RF refocus pulse or readout event between the RF inversion pulse and the next RF excitation pulse.

[0041] Specifically, as shown in Figure 7, five additional kernel delimiters are inserted based on the position of the RF inversion pulse. In this way, method 200 generates a total of 15 physical characteristic kernels, including five ecotrain kernels, five inversion kernels, and five immobile time kernels.

[0042] In other embodiments, steps 210 to 260 of method 200 may be performed in a different order and do not necessarily have to be performed in the order shown in Figure 2. For example, steps 230 and 240 may be performed before steps 210 and 220.

[0043] This embodiment offers several advantages over conventional MRI. For example, by reducing the need to process continuous streams "on the fly," sequence changes or corrections for the next kernel can be performed "just in time," thereby enabling such changes or corrections to be carried out by the CPU (Central Processing Unit) instead of the FPGA. Furthermore, by using the CPU for correction, it becomes possible to employ more sophisticated sequence / physical property adjustment models, such as processing unique inputs (temperature sensors) and using unique physical property models for specific clinical applications. While some of these improvements can be achieved using an FPGA, it is easier to implement them using a CPU-based approach with the kernel concept described above. In addition, upgrading to a more powerful CPU is easier than upgrading the FPGA.

[0044] Furthermore, in the kernel-based methods of this disclosure, each kernel based on physical properties is expected to be longer than in methods that use a fixed length such as 10 milliseconds, as described above. This provides longer processing time for applying more sophisticated sequence adjustments and for using more sophisticated physical property models. In addition, the kernel-based methods of this disclosure may also lead to a reduction in processing power requirements.

[0045] In addition, the kernel-based methods of this disclosure, particularly software-oriented solutions, can improve the scalability of MRI. For example, high-level "research" scanners can perform more complex physical corrections using more expensive, higher-performance processors, or "value" scanners can perform simpler physical corrections using less expensive, lower-performance processors. In either case, scaling the software complexity and / or CPU processor is easier than designing a unique FPGA for each scanner.

[0046] Figure 8 shows an exemplary embodiment of a medical imaging system 860 capable of performing the method 100 of the present disclosure. The medical imaging system 860 includes at least one scanning device 862, one or more image generating devices 864, each being a specially configured computer device (e.g., a specially configured desktop computer, a specially configured laptop computer, a specially configured server), and a display 866.

[0047] The scanning device 862 is configured to collect scan data by scanning a region (e.g., area, volume, slice) of a subject (e.g., a patient). Here, the means of scanning are, for example, MRI, computed tomography (CT), positron emission tomography (PET), radiography, and ultrasound.

[0048] One or more image generation devices 864 acquire scan data from the scanning device 862 and generate images of the subject's body parts based on the scan data. For example, one or more image generation devices 864 may generate images by performing a reconstruction process on the scan data during the generation of intermediate images or the reconstruction of the final image. Examples of reconstruction processes include GRAPPA (GeneRalized Autocalibrating Partially Parallel Acquisitions), CG-SENSE (Conjugate Gradient-SENSitivity Encoding), SENSE (SENSitivity Encoding), ARC (Autocalibrating Reconstruction for Cartesian imaging), SPIRIT (Iterative Self-consistent Parallel Imaging Reconstruction), and LORAKS (LOw-RAnk modeling of local K-Space neighborhoods).

[0049] In one embodiment, one or more image generating devices 864 generate an image, then transmit the image to a display 866, and the display 866 displays the image.

[0050] In another embodiment, in addition to the above, one or more image generating devices 864 may generate two images from the same scan data. One or more image generating devices 864 may generate two images from the same scan data using different reconstruction processes, and one image may have a lower resolution than the other. Furthermore, one or more image generating devices 864 may generate one image.

[0051] Here, Figure 9 shows a non-limiting example of an MRI apparatus 970. The MRI apparatus 970 shown in Figure 9 includes a pedestal 971 (shown in a schematic cross-sectional view) and various related MRI system components 972 interconnected via interfaces. Typically, at least the pedestal 971 is installed in a shielded room. The geometric structure of the MRI system shown in Figure 9 includes a substantially coaxial cylindrical structure containing a static magnetic field B0 magnet 973, a set of Gx, Gy, and Gz gradient magnetic field coils 974, and a whole-body RF coil (WBC) 975. Along the horizontal axis of each element contained in this cylindrical structure, an imaging volume 976 is shown so as to substantially surround the head of a patient 977 supported by a patient bed 978.

[0052] Within the imaging volume 976, one or more smaller array RF coils 979 are connected to be closer to the patient's head (referred herein, for example, to the “subject being scanned” or “subject”). As those skilled in the art will understand, typically coils and / or arrays are used, such as surface coils, which are relatively smaller than whole-body coils (WBCs) and customized for specific body parts (e.g., arms, shoulders, elbows, wrists, knees, legs, chest, spine, etc.). These smaller RF coils are referred herein to as array coils (ACs) or phased-array coils (PACs). They include at least one coil configured to transmit RF signals to the imaging volume 976, and a plurality of receiver coils configured to receive RF signals from a subject, such as the patient's head, within the imaging volume 976.

[0053] The MRI apparatus 970 includes an MRI system control unit 983 having input / output ports connected to a display 980, a keyboard 981, and a printer 982. As those skilled in the art will understand, the display 980 may be a touchscreen that also provides control inputs. Alternatively, a mouse or other I / O device may be used.

[0054] The MRI system control unit 983 is connected to the MRI sequence control unit 984 via an interface. The MRI sequence control unit 984 successively controls the Gx, Gy, and Gz gradient coil drivers 985, as well as the RF transmitter 986 and (if the same RF coil is used for both transmission and reception) the transmit / receive switch 987. The MRI sequence control unit 984 includes a suitable program code structure 988 for performing MRI (also known as nuclear magnetic resonance imaging or NMR (Nuclear Magnetic Resonance) imaging) techniques, including parallel imaging. Furthermore, the MRI sequence control unit 984 includes a processing circuit that performs the method shown in Figure 1. The MRI sequence control unit 984 is configured for MRI with or without parallel imaging. Furthermore, the MRI sequence control unit 984 executes one or more pre-scan sequences and scan sequences to acquire main scan magnetic resonance (MR) images (also called diagnostic images). The MR data obtained by pre-scanning is used, for example, to measure the sensitivity maps (sometimes called coil sensitivity maps or spatial sensitivity maps) of the whole-body RF coil 975 and / or array RF coil 979, and to measure the unfolded maps for parallel imaging.

[0055] The MRI system component 972 includes an RF receiver 989 that provides input to the MRI data processor 990 to generate processed image data to be transmitted to the display 980. The MRI data processor 990 is also configured to access pre-generated MR data, images and / or maps (e.g., coil sensitivity map, parallel image unfolding map, strain map, etc.), and / or system configuration parameters 991, as well as program code structures 992 and stored programs 993 for image reconstruction.

[0056] In one embodiment, the MRI data processor 990 includes a processing circuit. The processing circuit includes, for example, devices such as an Application Specific Integrated Circuit (ASIC), a programmable logic device (e.g., a Simple Programmable Logic Device (SPLD), a Complex Programmable Logic Device (CPLD), and an FPGA), and other circuit components configured to perform the functions described in this disclosure.

[0057] The MRI data processor 990 executes one or more sequences of one or more instructions contained in the program code structure 992 and the stored program 993, such as in the method 100 described herein. Alternatively, the instructions may be readable from another computer-readable medium, such as a hard disk or a removable media drive. Furthermore, one or more processors included in a multiprocessing configuration may also be available to execute sequences of instructions contained in the program code structure 992 and the stored program 993. In another embodiment, hardwired circuitry may be used instead of or in combination with software instructions. Thus, embodiments of this disclosure are not limited to any particular combination of hardware circuitry and software.

[0058] In addition, the term “computer-readable medium” as used herein means any non-temporary medium that supplies instructions to the MRI data processor 990 for execution. Computer-readable medium can be, but is not limited to, many forms, including non-volatile and volatile media. Examples of non-volatile media include optical disks, magnetic disks and magneto-optical disks, or removable media drives. Examples of volatile media include dynamic memory.

[0059] Furthermore, Figure 9 shows a generalized representation of the MRI system program memory in which the stored program 993 described above is stored. Here, the stored program 993 is stored in a non-temporary computer-readable storage medium that can be called by various data processing components included in the MRI device 970. As those skilled in the art will understand, the MRI system program memory may be divided and configured to be directly connected, at least partially, to different processing computers that most immediately require the stored program 993 during normal operation within the MRI system component 972 (i.e., the stored program 993 is not normally stored and directly connected to the MRI system control unit 983).

[0060] In addition, the MRI apparatus 970 shown in Figure 9 is used to perform the exemplary embodiments described below. The system components may be divided into different logical sets of "boxes," which typically include a number of digital signal processors (DSPs), microprocessors, and special-purpose processing circuits (e.g., processing circuits for high-speed analog-to-digital (A / D) conversion, fast Fourier transform, array processing, etc.). Each of these processing circuits is typically a clock-controlled "state machine" in which a physical data processing circuit progresses from one physical state to another every clock cycle (or a predetermined number of clock cycles).

[0061] Furthermore, during operation, not only do the physical states of processing circuits (e.g., CPU, registers, buffers, arithmetic units, etc.) change progressively from one clock cycle to another, but the associated physical states of data storage media (e.g., bit storage in a magnetic storage medium) also change from one state to another during the operation of such a system. For example, at the end of image reconstruction processing, and / or sometimes image reconstruction map generation processing (e.g., coil sensitivity map, unfolded map, ghost map, strain map, etc.), an array of computer-readable and accessible data value storage units contained in the physical storage medium changes from some previous state (e.g., all uniform "0" values ​​or all "1" values) to a new state. Here, the physical state of the physical parts of the array changes between a minimum and a maximum value to represent real-world physical events and physical states (e.g., the internal physical structure of a patient across imaging volume space). As those skilled in the art will understand, the array in which data values ​​are stored represents and constitutes a physical structure, such that when a specific structure of computer control program code is sequentially loaded into instruction registers by one or more CPUs of the MRI device 970 and executed, it generates a specific sequence of operable states within the MRI device 970 and changes those states.

[0062] Based on the above, many modifications and variations are possible. Therefore, it is understood that the technology disclosed herein may be implemented in a manner other than that specifically described herein, within the scope of the attached claims.

[0063] Embodiments of this disclosure can also be described in addition as follows:

[0064] (1) A device for magnetic resonance imaging, the device comprising a processing circuit, the processing circuit acquiring a set of sequence instructions for performing a magnetic resonance scan, dividing the acquired set of sequence instructions into a plurality of kernels by determining divided time points that define the boundaries of a plurality of kernels, the divided time points not being equally spaced in time, converting a first kernel among the plurality of kernels into a first hardware instruction set, transmitting the first hardware instruction set to an execution hardware board control unit, and reconstructing a magnetic resonance image from received data including data acquired by executing the first kernel.

[0065] (2) The apparatus according to (1), wherein the processing circuit is further configured to apply sequence corrections based on physical characteristics after converting the first kernel into a first hardware instruction set.

[0066] (3) The apparatus according to (1), wherein the processing circuit is further configured to apply sequence corrections based on physical characteristics before converting the first kernel to the first hardware instruction set.

[0067] (4) The apparatus described in any one of (1) to (3), wherein the processing circuit is further configured to divide the acquired set of sequence instructions into multiple kernels using a predetermined heuristic method.

[0068] (5) The apparatus according to any one of (1) to (4), wherein the processing circuit is further configured to divide the acquired set of sequence instructions into a plurality of kernels based on the position of at least one RF excitation pulse in the acquired set of sequence instructions.

[0069] (6) The apparatus according to any one of (1) to (5), wherein the processing circuit is further configured to divide the acquired set of sequence instructions into a plurality of kernels based on the position of at least one RF inversion pulse in the acquired set of sequence instructions.

[0070] (7) The apparatus according to (6), wherein the processing circuit is further configured to divide the acquired set of sequence instructions into multiple kernels by inserting a kernel delimiter at the position of the RF inversion event if no RF refocus pulse or readout event is detected between the RF inversion event and the next RF excitation pulse.

[0071] (8) The apparatus according to any one of (1) to (7), wherein the processing circuit is further configured to divide the acquired set of sequence instructions into multiple kernels based on the positions of the RF refocus pulses and readout events in the acquired set of sequence instructions.

[0072] (9) The apparatus according to any one of (1) to (8), wherein the processing circuit is further configured to repeat the conversion and transmission steps for each kernel of a plurality of kernels following the first kernel.

[0073] (10) A method for performing magnetic resonance imaging, comprising: obtaining a set of sequence instructions for performing a magnetic resonance scan; dividing the obtained set of sequence instructions into a plurality of kernels by determining division time points that define the boundaries of a plurality of kernels, wherein the division time points are not equally spaced in time; converting a first kernel among the plurality of kernels into a first hardware instruction set; transmitting the first hardware instruction set to an execution hardware board control unit; and reconstructing a magnetic resonance image from received data including data obtained by executing the first kernel.

[0074] (11) The method of (10), further comprising applying a sequence correction based on physical characteristics to the first kernel after the conversion step.

[0075] (12) The method of (10), further comprising applying a sequence correction based on physical characteristics to the first kernel before the conversion step.

[0076] (13) The method according to any one of (10) to (12), wherein the splitting step further comprises splitting the acquired set of sequence instructions into multiple kernels using a predetermined heuristic method.

[0077] (14) The method according to any one of (10) to (13), wherein the splitting step further comprises splitting the acquired set of sequence instructions into multiple kernels based on the location of at least one RF excitation pulse in the acquired set of sequence instructions.

[0078] (15) The method according to any one of (10) to (14), wherein the splitting step further comprises splitting the acquired set of sequence instructions into multiple kernels based on the location of at least one RF inversion pulse in the acquired set of sequence instructions.

[0079] (16) The method of (15), wherein the splitting step further comprises splitting the acquired set of sequence instructions into multiple kernels by inserting a kernel delimiter at the location of the RF inversion event if no RF refocus pulse or readout event is detected between the RF inversion event and the next RF excitation pulse.

[0080] (17) The method of any one of (10) to (16), wherein the splitting step further comprises splitting the acquired set of sequence instructions into multiple kernels based on the location of RF refocus pulses and readout events in the acquired set of sequence instructions.

[0081] (18) The method according to any one of (10) to (17), further comprising repeating the translation and transmission steps for each kernel of a plurality of kernels following the first kernel.

[0082] (19) A non-temporary computer-readable storage medium storing computer-readable instructions, wherein, when executed by a processing circuit, the computer-readable instructions cause the processing circuit to perform a method comprising: obtaining a set of sequence instructions for performing a magnetic resonance scan; dividing the obtained set of sequence instructions into a plurality of kernels by determining division time points that define the boundaries of a plurality of kernels, wherein the division time points are not equally spaced in time; converting a first kernel among the plurality of kernels into a first hardware instruction set; transmitting the first hardware instruction set to an execution hardware board control unit; and reconstructing a magnetic resonance image from received data including data obtained by executing the first kernel.

[0083] According to at least one embodiment described above, the scan sequence can be appropriately divided and hardware instructions can be processed.

[0084] While several embodiments have been described, these embodiments are presented as examples only and are not intended to limit the scope of the invention. These embodiments can be implemented in a variety of other forms, and various omissions, substitutions, modifications, and combinations of embodiments are possible without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents. [Explanation of Symbols]

[0085] 970 Magnetic Resonance Imaging (MRI) System 984 MRI Sequence Control Unit

Claims

1. acquiring a scan sequence for performing a magnetic resonance scan; Dividing the scan sequence into a plurality of kernels by determining division time points within the scan sequence that define boundaries of the plurality of kernels at non-equidistant time intervals; converting a kernel included in the plurality of kernels into a hardware instruction; transmitting the hardware instructions to a hardware controller for execution to acquire magnetic resonance scan data; Reconstructing a magnetic resonance image from the acquired magnetic resonance scan data by executing the hardware instructions. A magnetic resonance imaging device comprising a processing circuit.

2. the processing circuit converts the one kernel into the hardware instruction, and then applies a sequence correction based on a physical characteristic to the hardware instruction.

2. The magnetic resonance imaging apparatus according to claim 1.

3. the processing circuitry applies a physics-based sequence correction to the one kernel before converting the one kernel into the hardware instruction.

2. The magnetic resonance imaging apparatus according to claim 1.

4. the processing circuitry divides the scan sequence into the plurality of kernels using a predetermined heuristic.

4. The magnetic resonance imaging apparatus according to claim 1.

5. the processing circuitry divides the scan sequence into the plurality of kernels based on a position of at least one RF excitation pulse within the scan sequence.

5. The magnetic resonance imaging apparatus according to claim 4.

6. the processing circuitry divides the scan sequence into the plurality of kernels based on a position of at least one RF inversion pulse within the scan sequence.

5. The magnetic resonance imaging apparatus according to claim 4.

7. the processing circuit divides the scan sequence into the plurality of kernels by inserting a kernel breakpoint at a location of an RF reversal event when no RF refocus pulse or readout event is detected between the RF reversal event and a next RF excitation pulse in the scan sequence.

5. The magnetic resonance imaging apparatus according to claim 4.

8. the processing circuit divides the scan sequence into the plurality of kernels based on a position of an RF refocus pulse and a position of a readout event within the scan sequence.

5. The magnetic resonance imaging apparatus according to claim 4.

9. the processing circuit repeats, for each kernel subsequent to the one kernel included in the plurality of kernels, a process of converting the kernel into a hardware instruction and a process of transmitting the hardware instruction to the hardware control unit; 4. The magnetic resonance imaging apparatus according to claim 1.

10. 1. A method of performing magnetic resonance imaging, comprising: acquiring a scan sequence for performing a magnetic resonance scan; Dividing the scan sequence into a plurality of kernels by determining division time points within the scan sequence that define boundaries of the plurality of kernels at non-equidistant time intervals; converting a kernel included in the plurality of kernels into a hardware instruction; transmitting the hardware instructions to a hardware controller for execution to acquire magnetic resonance scan data; reconstructing a magnetic resonance image from the acquired magnetic resonance scan data by executing the hardware instructions; A method comprising:

11. 1. A program for performing magnetic resonance imaging, comprising: acquiring a scan sequence for performing a magnetic resonance scan; dividing the scan sequence into a plurality of kernels by determining division time points within the scan sequence at non-equidistant time intervals that define boundaries of the plurality of kernels; converting a kernel included in the plurality of kernels into a hardware instruction; transmitting the hardware instructions to a hardware controller for execution to acquire magnetic resonance scan data; reconstructing a magnetic resonance image from the acquired magnetic resonance scan data by executing the hardware instructions; A program that causes a computer to execute the following.

Images