Systems and methods for actual gradient waveform estimation

Abstract

The present invention provides a system for MRI. The system can obtain MRI scan data of a scanned object by instructing an MRI scanner to perform an MRI scan on the scanned object according to a first gradient waveform. The system can also determine a second gradient waveform based on a first gradient waveform and a gradient waveform determination model. The gradient waveform determination model is trained according to a machine learning algorithm. The system can also generate a target reconstructed image of the scanned object based on the second gradient waveform and the MRI scan data.

Description

[0001] Cross-reference

[0002] This application claims priority to U.S. Application No. 17 / 243,587, filed on April 29, 2021, the entire content of which is incorporated herein by reference. Technical Field

[0003] This application generally relates to magnetic resonance imaging (MRI), and more particularly, to systems and methods for estimating actual gradient waveforms in MRI. Background Art

[0004] MRI systems have been widely used for disease diagnosis and / or treatment. Generally, during an MRI scan of a subject (e.g., an ultra-short echo scan, a spiral MRI scan), due to factors such as hardware limitations, the actual gradient waveform applied to the subject may be different from the preset gradient waveform planned to be applied to the subject. The deviation of the actual gradient waveform from the preset gradient waveform may affect the quality of the resulting image of the MRI scan. One way to eliminate or reduce the effect of the deviation is to determine or estimate the actual gradient waveform and perform image reconstruction based on the actual gradient waveform. Thus, there is a need to develop systems and methods for estimating actual gradient waveforms. Summary of the Invention

[0005] According to one aspect of the present application, an MRI method is provided. The method includes obtaining MRI scan data of a subject by instructing an MRI scanner to perform an MRI scan on the subject according to a first gradient waveform. The method further includes determining a second gradient waveform based on the first gradient waveform and a gradient waveform determination model, wherein the gradient waveform determination model is trained according to a machine learning algorithm. The method further includes generating a target reconstructed image of the subject based on the second gradient waveform and the MRI scan data.

[0006] According to another aspect of the present application, an MRI method is provided. The method includes obtaining a first gradient waveform planned to be applied to a subject during an MRI scan, and determining a second gradient waveform based on the first gradient waveform and a gradient waveform determination model. The gradient waveform determination model is trained according to a machine learning algorithm. The method further includes instructing an MRI scanner to perform an MRI scan on the subject based on the second gradient waveform.

[0007] According to yet another aspect of the present application, an MRI system is provided, including at least one storage device for storing computer instructions, and at least one processor. The at least one processor is configured to execute the computer instructions to implement the MRI methods disclosed in the present application.

[0008] Some additional features of the present application can be illustrated in the following description. Through the study of the following description and the corresponding drawings, or the understanding of the production or operation of the embodiments, some additional features of the present application will be apparent to those skilled in the art. The features and implementations of the present application can be achieved and realized by practicing or using various aspects of the methods, tools, and combinations set forth in the detailed examples discussed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The present application will be further described by way of exemplary embodiments. These exemplary embodiments will be described in detail with reference to the drawings. The drawings are not drawn to scale. These embodiments are non-limiting exemplary embodiments, in which the same reference numerals in the various figures represent similar structures, where:

[0010] Figure 1 is a schematic diagram showing an exemplary MRI system according to some embodiments of the present application;

[0011] Figure 2 is a schematic diagram of exemplary hardware and / or software components of a computing device according to some embodiments of the present application;

[0012] Figure 3 is a schematic diagram of exemplary hardware and / or software components of a mobile device according to some embodiments of the present application;

[0013] Figure 4A and 4B is a block diagram showing an exemplary processing device according to some embodiments of the present application;

[0014] Figure 5 is a flowchart showing an exemplary process for generating a target reconstruction image of a scanned object according to some embodiments of the present application;

[0015] Figure 6 is a flowchart showing an exemplary process for determining a second gradient waveform based on a first gradient waveform and a gradient waveform determination model according to some embodiments of the present application;

[0016] Figure 7 is a schematic diagram illustrating an exemplary process for determining a second gradient waveform according to some embodiments of the present application;

[0017] Figure 8 is a flowchart showing an exemplary process for generating an amplitude determination model according to some embodiments of the present application;

[0018] Figure 9 is a schematic diagram showing an exemplary process for generating an amplitude determination model according to some embodiments of the present application;

[0019] Figure 10is a flowchart showing an exemplary process for generating a phase determination model according to some embodiments of the present application;

[0020] Figure 11 is a flowchart showing an exemplary process for jointly generating an amplitude determination model and a phase determination model according to some embodiments of the present application; and

[0021] Figure 12 is a flowchart showing an exemplary process for performing an MRI scan on a scanned object according to some embodiments of the present application. Detailed Description

[0022] To more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. However, those skilled in the art should understand that the present application can be implemented without these details. In other cases, well-known methods, processes, systems, components, and / or circuits have been described at a higher level to avoid unnecessarily obscuring aspects of the present application. For those of ordinary skill in the art, it is obvious that various changes can be made to the disclosed embodiments, and the general principles defined in the present application can be applied to other embodiments and application scenarios without departing from the principles and scope of the present application. Therefore, the present application is not limited to the disclosed embodiments, but conforms to the broadest scope consistent with the scope of the patent application.

[0023] The terms used in the present application are for the purpose of describing specific example embodiments only and are not restrictive. As used in the present application, the singular forms "a", "an", and "the" may also include the plural forms unless the context clearly indicates otherwise. It should also be understood that as used in the specification of the present application, the terms "comprising", "including" only indicate the presence of the described features, integers, steps, operations, components, and / or parts, but do not exclude the presence or addition of one or more other features, integers, steps, operations, components, parts, and / or their combinations.

[0024] It should be understood that the terms "system", "engine", "unit", "module", and / or "block" used herein are a way to distinguish different components, elements, parts, portions, or components at different levels in ascending order. However, these terms can also be replaced by other expressions if the same purpose can be achieved.

[0025] Generally, the terms "module", "unit", or "block" as used herein refer to logic embodied in hardware or firmware, or a collection of software instructions. The modules, units, or blocks described herein can be implemented as software and / or hardware and can be stored in any type of non-transitory computer-readable medium or other storage device. In some embodiments, software modules / units / blocks can be compiled and linked into an executable program. It should be understood that software modules can be called from other modules / units / blocks or from themselves, and / or can be called in response to detected events or interrupts. Software modules / units / blocks configured to execute on a computing device (e.g., a processor 210 as shown in Figure 2 FIG. 210) can be provided on a computer-readable medium, such as a compact disc, digital video disc, flash drive, magnetic disk, or any other tangible medium, or as a digital download (which may initially be stored in a compressed or installable format and require installation, decompression, or decryption before execution). The software code herein can be stored, in part or in whole, in the storage device of the computing device performing the operations and applied in the operation of the computing device. Software instructions can be embedded in firmware, such as an EPROM. It should also be understood that hardware modules / units / blocks can include connected logic components, such as gates and flip-flops, and / or can include programmable units, such as programmable gate arrays or processors. The modules / units / blocks or computing device functions described herein can be implemented as software modules / units / blocks, but can be represented in hardware or firmware. Generally, the modules / units / blocks described herein refer to logical modules / units / blocks, which can be combined with other modules / units / blocks or divided into sub-modules / sub-units / sub-blocks, although they are physically organized or stored devices. This description can apply to a system, an engine, or a part thereof.

[0026] It should be understood that when a unit, engine, module, or block is referred to as being "on", "connected to", or "coupled to" another unit, engine, module, or block, it can be directly on, connected, or coupled to the other unit, engine, module, or block, or communicate with intervening units, engines, modules, or blocks, and there may be or may not be intervening units, engines, modules, or blocks, unless the context clearly indicates otherwise. In this application, the term "and / or" can include any one or more of the related listed items or a combination thereof.

[0027] From the following description of the drawings, these and other features, characteristics, and the functions and operating methods of the related structural elements of the present application, as well as the combination of components and manufacturing economy, may become more apparent. These drawings all form a part of the specification of the present application. However, it should be understood that the drawings are for illustrative and descriptive purposes only and are not intended to limit the scope of the present application. It should be understood that the drawings are not drawn to scale.

[0028] The present disclosure provides systems and components for non-invasive imaging and / or treatment, such as for disease diagnosis, treatment, or research purposes. In some embodiments, the system may include a radiotherapy (RT) system, a computed tomography (CT) system, an emission computed tomography (ECT) system, an X-ray imaging system, a positron emission tomography (PET) system, etc., or any combination thereof. For illustrative purposes, the present invention describes systems and methods for radiotherapy.

[0029] The term "image" in the present invention is used to collectively refer to various forms of image data (e.g., scan data, projection data) and / or images, including two-dimensional (2D) images, three-dimensional (3D) images, four-dimensional (4D) images, etc. The terms "pixel" and "voxel" in the present invention may be used interchangeably to refer to the elements of an image. The term "anatomical structure" in the present invention may refer to gas (e.g., air), liquid (e.g., water), solid (e.g., stone), cells, tissues, organs of a subject, or any combination thereof, which can be shown in an image (e.g., a planning image or a treatment image, etc.) and actually exist in or on the body of the subject. The terms "region", "location", and "area" in the present invention may refer to the location of an anatomical structure shown in an image or the actual location of an anatomical structure existing in or on the body of the subject, as the image may indicate the actual location of a specific anatomical structure existing in or on the body of the subject.

[0030] MRI scans are often used to collect MRI scan data of a scanned object for disease diagnosis and / or treatment. However, due to the hardware limitations of the MRI scanner (e.g., power amplifier, gradient delay, eddy current), the actual gradient waveform applied to the scanned object may be different from the preset gradient waveform planned to be applied to the scanned object. If an MRI image is reconstructed based on the preset gradient waveform, the image quality may be poor (e.g., including artifacts). Traditional methods generally assume that the gradient system of the MRI scanner conforms to a linear model and use the linear model to predict the actual gradient waveform. However, non-linear factors (e.g., non-linear effects of power amplifiers, non-linear turbulence, etc., or any combination thereof) are often ignored by traditional methods. Using a linear model for actual gradient waveform prediction may have limited accuracy (e.g., resulting in image artifacts), especially when the gradient field changes rapidly during the MRI scan.

[0031] One aspect of the present invention relates to systems and methods for actual gradient waveform estimation in MRI. Such systems and methods can obtain MRI scan data of a scanned object by instructing an MRI scanner to perform an MRI scan on the scanned object according to a first gradient waveform (also referred to as a preset gradient waveform or an ideal gradient waveform). The system and method can determine a second gradient waveform based on the first gradient waveform and a gradient waveform determination model. The gradient waveform determination model can be generated by training according to a machine learning algorithm. The second gradient waveform can be regarded as an estimated value of the actual gradient waveform actually applied to the scanned object during the MRI scan. The system and method can further generate a target reconstructed image of the scanned object based on the second gradient waveform and the MRI scan data.

[0032] In some embodiments, the system and method can perform actual gradient waveform estimation before the MRI scan and implement the MRI scan based on the determination result. Specifically, the system and method can obtain the first gradient waveform planned to be applied to the scanned object. The system and method can also determine a second gradient waveform based on the first gradient waveform and the gradient waveform determination model. The system and method can further determine an adjusted gradient waveform by adjusting the first gradient waveform according to the second gradient waveform, and instruct the MRI scanner to perform an MRI scan on the scanned object according to the adjusted gradient waveform.

[0033] According to some embodiments of the present invention, the gradient waveform determination model can be used to determine the second gradient waveform for estimating the actual gradient waveform. Compared with the traditional linear model, the gradient waveform determination model can improve the accuracy, and this model learns the best mechanism for predicting the second gradient waveform from the training data. The gradient waveform determination model can consider linear factors, non-linear factors, and complex factors (usually undetectable by human or traditional actual gradient waveform determination methods). The application of the gradient waveform determination model can improve the accuracy of the determined second gradient waveform, and further improve the accuracy of image reconstruction performed based on the second gradient waveform or the accuracy of the MRI scan performed based on the second gradient waveform.

[0034] Figure 1 is a schematic diagram showing an exemplary MRI system 100 according to some embodiments of the present application. As Figure 1As shown, the MRI system 100 may include an MRI scanner 110, a processing device 120, a storage device 130, one or more terminals 140, and a network 150. In some embodiments, the MRI scanner 110, the processing device 120, the storage device 130, and / or the terminal 140 may be interconnected and / or communicate with each other via a wireless connection, a wired connection, or a combination thereof. The connections between the components in the MRI system 100 may be variable. For example, the MRI scanner 110 may be connected to the processing device 120 via the network 150. As another example, the MRI scanner 110 may be directly connected to the processing device 120.

[0035] The MRI scanner 110 may be configured to scan an object (or a part of the object) to obtain image data, such as echo signals (or MRI signals) associated with the scanned object. In some embodiments, the MRI scanner 110 may include, for example, a main magnet, gradient coils (or also referred to as spatial encoding coils), radio frequency coils, etc.

[0036] In some embodiments, the MRI scanner 110 may include gradient coils that are configured to apply a preset gradient waveform to the scanned object. However, due to device limitations, the actual gradient waveform applied to the scanned object during MRI scanning may be different from the preset gradient waveform. In some embodiments, depending on the type of the main magnet, the MRI scanner 110 may be a permanent magnet MRI scanner, a superconducting electromagnet MRI scanner, a resistive electromagnet MRI scanner, etc. In some embodiments, depending on the magnetic field strength, the MRI scanner 110 may be a high-field MRI scanner, a mid-field MRI scanner, a low-field MRI scanner, etc.

[0037] The subject scanned by the MRI scanner 110 may be biological or non-biological. For example, the subject may include a patient, an artificial object, etc. As another example, the subject may include a specific part, organ, tissue, and / or body part of a patient. By way of example only, the subject may include the head, brain, neck, body, shoulders, arms, chest, heart, stomach, blood vessels, soft tissue, knees, feet, etc., or a combination thereof.

[0038] The processing device 120 may process data and / or information obtained from the MRI scanner 110, the storage device 130, and / or the terminal 140. For example, the processing device 120 may determine a second gradient waveform (i.e., the estimated actual gradient waveform) by applying a gradient waveform determination model. As another example, the processing device 120 may generate a gradient waveform determination model through model training.

[0039] In some embodiments, a trained model (e.g., an amplitude determination model and / or a phase determination model) can be generated by a processing device, and the application of the trained model can be executed on a different processing device. In some embodiments, the trained model can be generated by a processing device of another system different from the MRI system 100, or a server different from the processing device 120 on which the trained model is applied. For example, the trained model can be generated by a first system of a vendor that provides and / or maintains such a trained model, and the actual gradient waveform estimation based on the provided trained model can be executed on a second system of the vendor's customer. In some embodiments, in response to a request to determine a second gradient waveform, the trained model can be applied online. In some embodiments, the trained model can be determined or generated offline.

[0040] In some embodiments, the trained model can be determined and / or updated (or maintained) by the manufacturer or vendor of the MRI scanner 110. For example, the manufacturer or vendor can load the amplitude determination model and / or the phase determination model into the MRI system 100 or a part thereof (e.g., the processing device 120) before or during the installation of the MRI scanner 110 and / or the processing device 120, and maintain or update the amplitude determination model and / or the phase determination model from time to time (regularly or irregularly). The maintenance or update can be achieved by installing a program stored on a storage device (e.g., a CD, a USB drive, etc.) or by obtaining a program from an external source (e.g., a server maintained by the manufacturer or vendor) via the network 150. The program may include a new model (e.g., a newly trained model) or a part of the model, and the new model or the part of the model can be used to replace or supplement the corresponding part of the existing model.

[0041] In some embodiments, the processing device 120 can be a single server or a group of servers. The group of servers can be centralized or distributed. In some embodiments, the processing device 120 can be local or remote. For example, the processing device 120 can access information and / or data from the MRI scanner 110, the storage device 130, and / or the terminal 140 via the network 150. For another example, the processing device 120 can be directly connected to the MRI scanner 110, the terminal 140, and / or the storage device 130 to access information and / or data. In some embodiments, the processing device 120 can be implemented on a cloud platform. For example, the cloud platform can include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an inter-cloud, a multi-cloud, etc., or a combination thereof. In some embodiments, the processing device 120 can be implemented by a computing device 200 having one or more components as Figure 2 described.

[0042] The storage device 130 can store data, instructions, and / or any other information. In some embodiments, the storage device 130 can store data obtained from the MRI scanner 110, the processing device 120, and / or the terminal 140. In some embodiments, the storage device 130 can store data and / or instructions that the processing device 120 can execute or use to execute the exemplary methods described in this application. In some embodiments, the storage device 130 can include a mass storage device, a removable storage device, a volatile read-write memory, a read-only memory (ROM), etc., or a combination thereof. Exemplary mass storage devices can include magnetic disks, optical disks, solid state drives, etc. Exemplary removable storage devices can include flash drives, floppy disks, optical disks, memory cards, compact disks, magnetic tapes, etc. The volatile read-write memory can include random access memory (RAM). The RAM can include dynamic RAM (DRAM), double data rate synchronous dynamic RAM (DDR SDRAM), static RAM (SRAM), T-RAM (T-RAM), and zero-capacitor random access memory (Z-RAM), etc. The ROM can include Mask ROM (MROM), programmable ROM (PROM), erasable programmable ROM (EPROM), electrically erasable programmable read-only memory (EEPROM), compact disk ROM (CD-ROM), and digital versatile disk ROM, etc. In some embodiments, the storage device 130 can be implemented on a cloud platform as described elsewhere in the present invention.

[0043] In some embodiments, the storage device 130 can be connected to the network 150 to communicate with one or more other components in the MRI system 100 (e.g., the MRI scanner 110, the processing device 120, and / or the terminal 140). One or more components of the MRI system 100 can access the data or instructions stored in the storage device 130 via the network 150. In some embodiments, the storage device 130 can be part of the processing device 120 or the terminal 140.

[0044] The terminal 140 can be configured to enable user interaction between the user and the MRI system 100. For example, the terminal 140 can receive instructions to cause the MRI scanner 110 to scan an object from the user. As another example, the terminal 140 can receive a processing result (e.g., a slice image representing the slice position of the scanned object) from the processing device 120 and display the processing result to the user. In some embodiments, the terminal 140 can be connected to and / or communicate with the MRI scanner 110, the processing device 120, and / or the storage device 130. In some embodiments, the terminal 140 can include a mobile device 140-1, a tablet computer 140-2, a laptop computer 140-3, etc., or a combination thereof. For example, the mobile device 140-1 can include a mobile phone, a personal digital assistant (PDA), a gaming device, a navigation device, a point-of-sale (POS) device, a laptop computer, a tablet computer, a desktop computer, etc., or a combination thereof. In some embodiments, the terminal 140 can include an input device, an output device, etc. The input device can include alphanumeric keys and other keys that can be input via a keyboard, a touch screen (e.g., having tactile or haptic feedback), voice input, eye tracking input, a brain monitoring system, or any other similar input mechanism. The input information received by the input device can be sent to the processing device 120 via, for example, a bus for further processing. Other types of input devices can include cursor control devices, such as a mouse, a trackball, or cursor direction keys, etc. The output device can include a display, a speaker, a printer, etc., or a combination thereof. In some embodiments, the terminal 140 can be a part of the processing device 120 or the MRI scanner 110.

[0045] Network 150 may include any suitable network that may facilitate information and / or data exchange of MRI system 100. In some embodiments, one or more components of MRI system 100 (e.g., MRI scanner 110, processing device 120, storage device 130, terminal 140, etc.) may communicate information and / or data with one or more other components of MRI system 100 via network 150. For example, processing device 120 may obtain image data (e.g., echo signals) from MRI scanner 110 via network 150. As another example, processing device 120 may obtain user instructions from terminal 140 via network 150. Network 150 may include a public network (e.g., the Internet), a private network (e.g., local area network (LAN), wide area network (WAN), etc.), a wired network (e.g., Ethernet), a wireless network (e.g., 802.11 network, Wi-Fi network, etc.), a cellular network (e.g., long term evolution (LTE) network), a frame relay network, a virtual private network (VPN), a satellite network, a telephone network, routers, hubs, switches, server computers, etc., or a combination thereof. For example, network 150 may include a wired network, a wired network, an optical fiber network, a telecommunication network, an intranet, a wireless local area network (WLAN), a metropolitan area network (MAN), a public switched telephone network (PSTN), a BluetoothTM network, a ZigBeeTM network, a near field communication (NFC) network, etc., or a combination thereof. In some embodiments, network 150 may include one or more network access points. For example, network 150 may include wired and / or wireless network access points, such as base stations and / or Internet exchange points, through which one or more components of MRI system 100 may connect to network 150 to exchange data and / or information.

[0046] This description is intended to be illustrative and not to limit the scope of the present application. Many alternatives, modifications, and variations will be apparent to those skilled in the art. The features, structures, methods, and characteristics of the exemplary embodiments described herein may be combined in various ways to obtain additional and / or alternative exemplary embodiments. In some embodiments, MRI system 100 may include one or more additional components and / or may omit one or more of the above-described components. Additionally or alternatively, two or more components of MRI system 100 may be integrated into a single component. For example, processing device 120 may be integrated into MRI scanner 110. As another example, a component of MRI system 100 may be replaced by another component that can implement the function of the component. In some embodiments, storage device 130 may be a data memory including a cloud computing platform, such as a public cloud, a private cloud, a community, and a hybrid cloud, etc. However, these changes and modifications do not depart from the scope of the present application.

[0047] Figure 2FIG. 0 is a schematic diagram showing exemplary hardware and / or software components of a computing device 200 according to some embodiments of the present application. The computing device 200 can be used to implement any component of the MRI system 100 as described herein. For example, the processing device 120 and / or the terminal 140 can be implemented on the computing device 200 through its hardware, software program, firmware, or a combination thereof. Although only one such computing device is shown, for convenience, the computer functions related to the MRI system 100 described herein can be implemented in a distributed manner on multiple similar platforms to distribute the processing load. As Figure 2 shown, the computing device 200 may include a processor 210, a storage device 220, an input / output (I / O) 230, and a communication port 240.

[0048] The processor 210 may execute computer instructions (e.g., program code) according to the techniques described herein and perform the functions of the processing device 120. The computer instructions may include, for example, routines, programs, scan objects, components, data structures, procedures, modules, and functions that perform the specific functions described herein. For example, the processor 210 may process data obtained from the MRI scanner 110, the terminal 140, the storage device 130, and / or any other component of the MRI system 100. In some embodiments, the processor 210 may include one or more hardware processors, such as a microcontroller, a microprocessor, a reduced instruction set computer (RISC), an application-specific integrated circuit (ASIC), an application-specific instruction set processor (ASIP), a central processing unit (CPU), a graphics processing unit (GPU), a physics processing unit (PPU), a microcontroller unit, a digital signal processor (DSP), a field-programmable gate array (FPGA), an advanced RISC machine (ARM), a programmable logic device (PLD), any circuit or processor capable of performing one or more functions, etc., or any combination thereof.

[0049] For illustration only, only one processor is described in the computing device 200. However, it should be noted that the computing device 200 in the present invention may also include multiple processors. Therefore, the operations and / or method operations performed by one processor as described in the present invention may also be performed jointly or separately by multiple processors. For example, if in the present application, the processor of the computing device 200 performs operations A and B simultaneously, it should be understood that operations A and B may also be performed jointly or separately by two or more different processors in the computing device 200 (e.g., the first processor performs operation A, the second processor performs operation B, or the first and second processors jointly perform operations A and B).

[0050] The storage device 220 can store data obtained from one or more components of the MRI system 100. In some embodiments, the storage device 220 may include a mass storage device, a removable storage device, a volatile read-write memory, a read-only memory (ROM), etc., or any combination thereof. In some embodiments, the storage device 220 can store one or more programs and / or instructions to execute the exemplary methods described in this application. For example, the storage device 220 can store a program for the processing device 120 to execute for generating a gradient waveform determination model.

[0051] The I / O 230 can input and / or output signals, data, information, etc. In some embodiments, the I / O 230 can enable user interaction with the processing device 120. In some embodiments, the I / O 230 can include an input device and an output device. The input device can include alphanumeric keys and other keys that can be input via a keyboard, a touch screen (e.g., with haptic or tactile feedback), voice input, eye tracking input, a brain monitoring system, or any other similar input mechanism. The input information received through the input device can be transmitted via, for example, a bus to another component (e.g., the processing device 120) for further processing. Other types of input devices can include cursor control devices, such as a mouse, a trackball, or cursor direction keys, etc. The output device can include a display (e.g., a liquid crystal display (LCD), a light-emitting diode (LED)-based display, a flat panel display, a curved screen, a television device, a cathode ray tube (CRT), a touch screen), a speaker, a printer, etc., or any combination thereof.

[0052] The communication port 240 can be connected to a network (e.g., the network 150) to facilitate data communication. The communication port 240 can establish a connection between the processing device 120 and the MRI scanner 110, the terminal 140, and / or the storage device 130. The connection can be a wired connection, a wireless connection, any other communication connection capable of data transmission and / or reception, and / or any combination of these connections. The wired connection can include, for example, a cable, an optical fiber cable, a telephone line, etc., or any combination thereof. The wireless connection can include, for example, a Bluetooth TM link, a Wi-Fi TM link, a WiMAX TM link, a WLAN link, a ZigBee TM link, a mobile network link (e.g., 3G, 4G, 5G), or the like or any combination thereof. In some embodiments, the communication port 240 can be and / or include a standardized communication port, such as RS232, RS485, etc. In some embodiments, the communication port 240 can be a specifically designed communication port. For example, the communication port 240 can be designed according to the Digital Imaging and Communications in Medicine (DICOM) protocol.

[0053] Figure 3 is a schematic diagram showing exemplary hardware and / or software components of an exemplary mobile device 300 according to some embodiments of the present application. In some embodiments, the terminal 140 and / or the processing device 120 may be implemented on the mobile device 300 respectively. As Figure 3 shown, the mobile device 300 may include a communication platform 310, a display 320, a graphics processing unit (GPU) 330, a central processing unit (CPU) 340, an I / O 350, a memory 360, and a storage 390. In some embodiments, any other suitable components, including but not limited to a system bus or a controller (not shown), may also be included in the mobile device 300. In some embodiments, a mobile operating system 370 (e.g., iOS TM , Android TM , Windows Phone TM ) and one or more application programs 380 may be loaded from the storage 390 into the memory 360 for execution by the CPU 340. The application program 380 may include a browser or any other suitable mobile application for receiving and presenting information related to the MRI system 100. The interaction between the user and the information stream may be implemented via the I / O 350 and provided to the processing device 120 and / or other components of the MRI system 100 via the network 150.

[0054] To implement the various modules, units, and their functions described in the present application, a computer hardware platform may be used as the hardware platform for one or more of the components described herein. A computer with user interface elements may be used to implement a personal computer (PC) or any other type of workstation or terminal device. If the computer is appropriately programmed, the computer may also be used as a server.

[0055] Figure 4A and 4B are block diagrams showing exemplary processing devices 120A and 120B according to some embodiments of the present application. In some embodiments, the processing devices 120A and 120B may be embodiments of the processing device 120 as Figure 1 described. The processing device 120A may be configured to perform actual gradient waveform prediction by applying a gradient waveform determination model. The processing device 120B may be configured to generate a gradient waveform determination model through model training.

[0056] In some embodiments, the processing devices 120A and 120B may be respectively in different processing units (e.g., Figure 2 the processor 210 shown in Figure 3It is implemented on the CPU 340 shown. By way of example only, the processing device 120A can be implemented on the CPU 340 of the terminal device, and the processing device 120B can be implemented on the computing device 200. As another example, the processing device 120A can be implemented on the computing device of the MRI system 100, and the processing device 120B can be a device or system of the manufacturer of the MRI system 100 or a part thereof (e.g., the MRI scanner 110). Alternatively, the processing devices 120A and 120B can be implemented on the same computing device 200 or the same CPU 340. For example, the processing devices 120A and 120B can be implemented on the same computing device 200.

[0057] As Figure 4A shown, the processing device 120A can include an acquisition module 401, a determination module 402, a generation module 403, and a control module 404.

[0058] The acquisition module 401 can be configured to obtain information related to the MRI system 100. For example, the acquisition module 401 can obtain MRI scan data of the scanned object by instructing the MRI scanner to perform an MRI scan on the scanned object according to the first gradient waveform. The first gradient waveform refers to the gradient waveform planned to be applied to the scanned object during the MRI scan. As another example, the acquisition module 401 can be configured to obtain the first gradient waveform planned to be applied to the scanned object before performing the MRI scan.

[0059] The determination module 402 can be configured to determine a second gradient waveform based on the first gradient waveform and the gradient waveform determination model. The second gradient waveform can be an estimated value of the actual gradient waveform applied to the scanned object during the MRI scan. In some embodiments, the determination module 402 can be configured to determine an initial gradient waveform by processing the first gradient waveform using at least one response function. The determination module 402 can be configured to determine the second gradient waveform based on the initial gradient waveform and the gradient waveform determination model. More descriptions regarding the determination of the second gradient waveform can be found elsewhere in this application. See Figure 5 Operation 520 and its related descriptions in. In some embodiments, the determination module 402 can be configured to determine an adjusted gradient waveform by adjusting the first gradient waveform according to the second gradient waveform. More descriptions regarding the determination of the adjusted waveform can be found elsewhere in this application. See, for example Figure 12 Operation 1230 and its related descriptions in.

[0060] The generation module 403 can be configured to generate a target reconstruction image of the scanned object based on the second gradient waveform and the MRI scan data. More descriptions regarding the generation of the target reconstruction image can be found elsewhere in this application. See Figure 5Operation 530 therein and its related description.

[0061] The control module 404 may be configured to control one or more components of the MRI system 100. For example, the control module 404 may instruct the MRI scanner to perform an MRI scan on a scanned object according to an adjusted gradient waveform.

[0062] As Figure 4B shown, the processing device 120B may include an acquisition module 405 and a training module 406.

[0063] The acquisition module 405 may be configured to acquire training data for generating one or more trained models disclosed herein, such as training data for a gradient waveform determination model, an amplitude determination model, and a phase determination model. For example, the acquisition module 405 may be configured to obtain a plurality of first training samples and a first initial model, which may be used to generate an amplitude determination model. As another example, the acquisition module 405 may be configured to obtain a plurality of second training samples and a second initial model, which may be used to generate a phase determination model. As yet another example, the acquisition module 405 may be configured to obtain a plurality of third training samples and a third initial model including a first sub-model and a second sub-model, which may be used to jointly generate an amplitude determination model and a phase determination model. More descriptions of the training data can be found elsewhere in this application. For example, see Figures 8 - 11 and its related description.

[0064] The training module 406 may be configured to generate one or more trained models (e.g., machine learning models) through model training. In some embodiments, one or more trained models may be generated according to machine learning algorithms. Machine learning algorithms may include, but are not limited to, artificial neural network algorithms, deep learning algorithms, decision tree algorithms, association rule algorithms, inductive logic programming algorithms, support vector machine algorithms, clustering algorithms, Bayesian network algorithms, reinforcement learning algorithms, representation learning algorithms, similarity and metric learning algorithms, sparse dictionary learning algorithms, genetic algorithms, rule-based machine learning algorithms, etc., or any combination thereof. The machine learning algorithms for generating one or more machine learning models may be supervised learning algorithms, semi-supervised learning algorithms, unsupervised learning algorithms, etc. More descriptions of generating one or more trained models can be found elsewhere in this application. See, for example, see Figures 8 - 11 and its related description.

[0065] It should be noted that the above description is provided for illustrative purposes only and is not intended to limit the scope of the present application. Those of ordinary skill in the art can make various changes and modifications based on the description of the present application. However, these changes and modifications do not depart from the scope of the present application. In some embodiments, the processing device 120A and the processing device 120B may share two or more modules, and any one module may be divided into two or more units. For example, the processing devices 120A and 120B may share the same acquisition module, that is, the acquisition module 401 and the acquisition module 405 are the same module. In some embodiments, the processing device 120A and / or the processing device 120B may include one or more additional modules, such as a storage module (not shown) for storing data. In some embodiments, the processing device 120A and the processing device 120B may be integrated into one processing device 120.

[0066] [[ID=3 is a flowchart showing an exemplary process for generating a target reconstruction image of a scanned object according to some embodiments of the present application. In some embodiments, the process 500 may be executed by the MRI system 100. For example, the process 500 may be implemented as a set of instructions (e.g., an application program) stored in a storage device (e.g., the storage device 130, the storage device 220, and / or the memory 390). In some embodiments, the processing device 120A (e.g., the processor 210 of the computing device 200, the CPU 340 of the mobile device 300, and / or ​ one or more of the modules shown) may execute the instruction set, and when the instructions are executed, the processing device 120A may be configured to execute the process 500. The operations of the following-described process are for illustrative purposes only. In some embodiments, the process 500 may be completed by one or more additional operations not described and / or one or more operations not discussed. In addition, ​ the order of operations of the process 500 shown and described below is not intended to be limiting.

[0067] In 510, the processing device 120A (e.g., the acquisition module 401) may obtain MRI scan data of the scanned object by instructing the MRI scanner to perform an MRI scan on the scanned object according to the first gradient waveform.

[0068] The scanned object may be biological or non-biological. For example, the subject may include a patient, an artificial object, etc. As another example, the subject may include a specific part, organ, tissue, and / or body part of a patient. By way of example only, the subject may include the head, brain, neck, body, shoulders, arms, chest, heart, stomach, blood vessels, soft tissue, knees, feet, etc., or a combination thereof.

[0069] In some embodiments, the processing device 120A may instruct an MRI scanner (e.g., MRI scanner 110) to perform an MRI scan on a scan object according to a first gradient waveform and obtain MRI scan data from the MRI scanner. Alternatively, the MRI scan data may be pre-collected and stored in a storage device (e.g., storage device 130, storage device 220, external storage device). The processing device 120A may retrieve the MRI scan data from the storage device. In some embodiments, the MRI scan may be an ultra-short echo time MRI or a spiral MRI scan.

[0070] The first gradient waveform refers to the gradient waveform planned to be applied to the scan object during the MRI scan. In some embodiments, the first gradient waveform may be defined by first amplitude information and first phase information of the first gradient waveform. For example, the first amplitude information may include the amplitudes of the first gradient waveform at at least two time points. The first phase information may include the phases of the first gradient waveform at at least two time points. In some embodiments, the first amplitude information may be represented as a first vector, and the first phase information may be represented as a second vector.

[0071] In some embodiments, the first gradient waveform may be determined according to the default settings of the MRI system 100 or manually set by a user of the MRI system 100 via, for example, a terminal (e.g., terminal 140). For example, a doctor may select the first gradient waveform from multiple first gradient waveforms. Alternatively, the first gradient waveform may be determined by the processing device 120A based on the actual situation. For example, the processing device 120A may determine the first gradient waveform based on information related to the scan object, such as the scan region, age, body shape, etc., of the scan object, or any combination thereof. In some embodiments, the processing device 120A may obtain one or more scan parameters (e.g., field of view, bandwidth). The processing device 120A may also determine the first gradient waveform based on one or more scan parameters.

[0072] In an application, due to hardware limitations as described elsewhere in this application, the actual gradient waveform applied to the scan object may be different from the first gradient waveform. Image reconstruction using the first gradient waveform may generate an MRI image with lower image quality (e.g., having artifacts). Therefore, it may be necessary to determine a second gradient waveform as an estimate of the actual gradient waveform and use it to generate a reconstructed image of the scan object to improve the image quality.

[0073] In 520, the processing device 120A (e.g., determination module 402) may determine the second gradient waveform based on the first gradient waveform and a gradient waveform determination model.

[0074] In some embodiments, the second gradient waveform may be defined by second amplitude information and second phase information of the second gradient waveform. For example, the second amplitude information may include the amplitudes of the second gradient waveform at at least two time points. The second phase information may include the phases of the second gradient waveform at at least two time points.

[0075] As used herein, a gradient waveform determination model refers to a trained model (e.g., a machine learning model) or algorithm that is used to determine information related to a second gradient waveform based on its input. For example, the processing device 120A may input the first amplitude information and the first phase information of the first gradient waveform into the gradient waveform determination model, and the gradient waveform determination model may output the second amplitude information and the second phase information of the second gradient waveform. As another example, the processing device 120A may determine an initial gradient waveform based on the first gradient waveform and at least one response function. The processing device 120A may further input the information of the initial gradient waveform into the gradient waveform determination model, and the gradient waveform determination model may output the second amplitude information and the second phase information of the second gradient waveform. More descriptions about the initial gradient waveform can be found elsewhere in this application. See, for example ​ and its related descriptions.

[0076] In some embodiments, the gradient waveform determination model may include an amplitude determination model and a phase determination model. The amplitude determination model refers to a trained model (e.g., a machine learning model) or algorithm that is used to determine the second amplitude information of the second gradient waveform based on its input. The phase determination model refers to a trained model (e.g., a machine learning model) or algorithm that is used to determine the second phase information of the second gradient waveform based on its input.

[0077] In some embodiments, the amplitude determination model and the phase determination model may be two independent models that are trained separately. In other words, the processing device 120A may obtain two independent models and determine the second gradient waveform based on these two independent models. For example, the processing device 120A may determine the first input of the amplitude determination model based on the first gradient waveform, and determine the second amplitude information of the second gradient waveform based on the first input and the amplitude determination model. The processing device 120A may also determine the second input of the phase determination model based on the first gradient waveform, and determine the second phase information of the second gradient waveform based on the second input and the phase determination model.

[0078] Alternatively, the amplitude determination model and the phase determination model can be two sub-models of a trained hybrid model, where the amplitude determination model and the phase determination model can be jointly trained during the generation of the trained hybrid model. In other words, the processing device 120A can obtain a single trained hybrid model including the amplitude determination model and the phase determination model, and determine the second gradient waveform based on the trained hybrid model. For example, the processing device 120A can determine a third input of the trained hybrid model, and determine the second amplitude information and the second phase information of the second gradient waveform based on the trained hybrid model and the third input. For ease of description, the present application uses the term "gradient waveform determination model" to generically refer to the phase determination model, the amplitude determination model, the trained hybrid model, or any combination of these models. A more detailed description of determining the second gradient waveform based on the gradient waveform determination model can be found elsewhere in the present application. See, for example ​ and its related description.

[0079] In some embodiments, the gradient waveform determination model can be a machine learning model according to a machine learning algorithm. For example, the gradient waveform determination model can include a neural network model, such as a Convolutional Neural Network (CNN) model (e.g., a complete CNN model, a V-net model, a U-net model, an AlexNet model, an Oxford Visual Geometry Group (VGG) model, a ResNet model), a Generative Adversarial Network (GAN) model, etc., or any combination thereof. In some embodiments, the gradient waveform determination model can include one or more components for feature extraction and / or feature combination, such as a fully convolutional block, a skip connection, a residual block, a dense block, etc., or any combination thereof.

[0080] Exemplary machine learning algorithms can include artificial neural network algorithms, deep learning algorithms, decision tree algorithms, association rule algorithms, inductive logic programming algorithms, support vector machine algorithms, clustering algorithms, Bayesian network algorithms, reinforcement learning algorithms, representation learning algorithms, similarity and metric learning algorithms, sparse dictionary learning algorithms, genetic algorithms, rule-based machine learning algorithms, etc., or any combination thereof. The machine learning algorithm for generating the gradient waveform determination model can be a supervised learning algorithm, a semi-supervised learning algorithm, an unsupervised learning algorithm, etc.

[0081] In some embodiments, the processing device 120A may obtain a gradient waveform determination model from one or more components (e.g., the storage device 130, the terminal 140) of the MRI system 100 or an external source through a network (e.g., the network 150). For example, the gradient waveform determination model may be pre-trained by a computing device (e.g., the processing device 120B) and stored in a storage device of the MRI system 100 (e.g., the storage device 130, the storage device 220, and / or the storage device 390). The processing device 120A may access the storage device and retrieve the gradient waveform determination model from the storage device. In some embodiments, the processes 800 disclosed herein may be executed by a computing device (e.g., the processing device 120B) to generate an amplitude waveform determination model. The phase generation model may be generated by a computing device (e.g., the processing device 120B) by executing the process 1000 disclosed herein. The trained hybrid model may be generated by a computing device (e.g., the processing device 120B) by executing the process 1100 disclosed herein. Different models may be generated by the same computing device or different computing devices.

[0082] In 530, the processing device 120A (e.g., the generation module 403) may generate a target reconstruction image of the scanned object based on the second gradient waveform and the MRI scan data.

[0083] In some embodiments, the processing device 120A may generate k-space data by filling the MRI scan data into the k-space according to the second gradient waveform. The processing device 120A may further reconstruct the target reconstruction image of the scanned object based on the k-space data, for example, by performing an inverse Fourier transform on the k-space data. In some embodiments, the processing device 120A may use an MRI image reconstruction algorithm to generate the target reconstruction image.

[0084] It should be noted that the above description of the process 500 is for illustrative purposes only and is not intended to limit the scope of the present application. For those of ordinary skill in the art, various changes and modifications can be made according to the description of the present application. However, these changes and modifications do not depart from the scope of the present application. In some embodiments, the process 500 may be completed by one or more additional operations not described and / or without one or more of the above operations. By way of example only, the process 500 may include an additional operation of sending the target reconstruction image to a terminal for display. Again, for example, the operation 530 may be omitted.

[0085] ​ is a flowchart showing an exemplary process of determining a second gradient waveform based on a first gradient waveform and a gradient waveform determination model according to some embodiments of the present application. In some embodiments, the process 600 may be executed to implement at least a part of the operation 520 as described ​ above.

[0086] In 610, a processing device 120A (e.g., a determination module 402) may determine an initial gradient waveform by processing a first gradient waveform using at least one response function.

[0087] As used herein, the response function may include a linear function or model that can be used for actual gradient waveform prediction. The at least one response function may consider linear factors in actual waveform prediction, such as the linear effect of a power amplifier, the linear effect of turbulence, etc., or any combination thereof. However, the at least one response function may ignore non-linear factors, such as the non-linear effect of a power amplifier, non-linear turbulence, etc., or any combination thereof. The actual gradient waveform prediction using the at least one response function may have limited accuracy (e.g., resulting in image artifacts), especially when the gradient field changes rapidly during an MRI scan. Therefore, the present application uses at least one response function to preprocess the first gradient waveform to determine the initial gradient waveform, and further uses a gradient waveform determination model to determine a second gradient waveform based on the initial gradient waveform. In this way, both linear and non-linear factors can be considered simultaneously, thereby improving the accuracy of the determined second gradient waveform, and further improving the accuracy of image reconstruction performed based on the second gradient waveform. Additionally, by preprocessing the first gradient waveform, the computational amount of the gradient waveform determination model can be reduced, which can improve the efficiency of second gradient waveform determination.

[0088] The initial gradient waveform may be defined by initial amplitude information and initial phase information. For example, the initial amplitude information may include the amplitudes of the initial gradient waveform at at least two time points. The initial phase information may include the phases of the initial gradient waveform at multiple time points.

[0089] In some embodiments, the processing device 120A may input information related to the first gradient waveform into at least one response function, and the at least one response function may output information related to the initial gradient waveform. For example, the at least one response function may include an amplitude response function and a phase response function. The processing device 120A may determine the initial phase information of the initial gradient waveform by processing the first phase information of the first gradient waveform using the phase response function. The processing device 120A may determine the initial amplitude information of the initial gradient waveform by processing the first amplitude information of the first gradient waveform using the amplitude response function.

[0090] In some embodiments, the processing device 120A may determine at least one response function based on experimental data of one or more experimental scans (e.g., actual scans or simulated scans) performed on one or more experimental scan objects. Alternatively, at least one response function may previously have been determined by the processing device 120A or another computing device and stored in a storage device (e.g., storage device 130, storage device 220, and / or memory 390). The processing device 120A may retrieve at least one response function from the storage device.

[0091] In 620, the processing device 120A (e.g., determination module 402) may determine a second gradient waveform based on an initial gradient waveform and a gradient waveform determination model.

[0092] In some embodiments, as ​ described, the gradient waveform determination model may include an amplitude determination model and a phase determination model. The processing device 120A may determine second amplitude information of the second gradient waveform by processing primary amplitude information using the amplitude determination model. The processing device 120A may determine second phase information of the second gradient waveform by processing initial phase information using the phase determination model.

[0093] For illustrative purposes, ​ FIG. shows a schematic diagram of an exemplary process for determining a second gradient waveform according to some embodiments of the present application.

[0094] As ​ shown, the processing device 120A may process information of the first gradient waveform using an amplitude response function and a phase response function. For example, the first amplitude information of the first gradient waveform may be processed by the amplitude response function to determine the initial amplitude information of the initial gradient waveform. The first phase information of the first gradient waveform may be processed by the phase response function to determine the initial phase information of the initial gradient waveform.

[0095] Then, the processing device 120A may process the initial amplitude information using the amplitude determination model to determine the second amplitude information of the second gradient waveform. The processing device 120A may process the initial phase information using the phase determination model to determine the second phase information of the second gradient waveform.

[0096] It should be noted that the above regarding ​ and ​The description is for illustrative purposes only and is not intended to limit the scope of the present invention. Those of ordinary skill in the art can make various changes and modifications based on the description of this application. However, these changes and modifications do not depart from the scope of this application. In some embodiments, the processing device 120A can directly determine the second gradient waveform based on the gradient waveform determination model without using at least one response function. For example, the second amplitude information of the second gradient waveform can be determined by processing the first amplitude information of the first gradient waveform using an amplitude determination model; and the second phase information of the second gradient waveform can be determined by processing the first phase information of the first gradient waveform using a phase determination model.

[0097] ​ is a flowchart showing an exemplary process for generating an amplitude determination model according to some embodiments of the present application. In some embodiments, process 800 can be executed by the MRI system 100. For example, process 800 can be implemented as a set of instructions (e.g., an application program) stored in a storage device (e.g., storage device 130, storage device 220, and / or storage device 390). In some embodiments, the processing device 120B (e.g., the processor 210 of the computing device 200, the CPU 340 of the mobile device 300, and / or ​ one or more modules shown therein) can execute the instruction set and is accordingly instructed to execute process 800.

[0098] In some embodiments, one or more operations of process 800 can be executed to implement ​ at least a part of the operation 520 described above. In some embodiments, process 800 can be executed by another device or system other than the MRI system 100, e.g., a device or system of a manufacturer's supplier. For illustrative purposes, the implementation process of process 800 is described below by taking the processing device 120B as an example.

[0099] In 810, the processing device 120B (e.g., the acquisition module 405) can obtain a plurality of first training samples. Each of the plurality of first training samples can include the sample first amplitude information of the sample first gradient waveform planned to be applied to the sample scan object during the sample MRI scan, and the true amplitude information of the true gradient waveform applied to the sample scan object during the sample MRI scan.

[0100] In some embodiments, the sample scan object of the first training sample may be of the same type or a different type from the scan object described in step 510. For example, the scan object can be a patient's head, and the sample scan object may be the head of another patient or an artificial object (e.g., a phantom). The sample MRI scan can be an actual MRI scan or a simulated MRI scan applied to the sample scan object. The sample first gradient waveform of the first training sample refers to the first gradient waveform planned to be applied to the sample scan object of the first training sample. For example, the sample first gradient waveform can be defined by sample first amplitude information and sample first phase information. In some embodiments, multiple sample MRI scans with different scan parameters (e.g., field of view, bandwidth, resolution) can be performed on the sample scan object so that multiple different sample first gradient waveforms with different shapes can be applied to the sample scan object in the sample MRI scans.

[0101] The ground truth gradient waveform of the first training sample refers to the measured value of the actual gradient waveform applied to the sample scan object in the sample MRI scan. For example, during the sample MRI scan, the ground truth gradient waveform can be measured by a magnetic field detection device. As another example, the sample scan object can be a water phantom, and the ground truth gradient waveform can be determined based on the MRI signals collected during the sample MRI scan of the water phantom. In some embodiments, the ground truth gradient waveform can be defined by the ground truth amplitude information and the ground truth phase information of the ground truth gradient waveform. The ground truth amplitude information of the ground truth gradient waveform can include the amplitudes of the ground truth gradient waveform at at least two time points during the sample MRI scan. The ground truth phase information of the ground truth gradient waveform can include the phases of the ground truth gradient waveform at at least two time points during the sample MRI scan.

[0102] In some embodiments, the first training sample (or a part thereof) can be pre-generated by a computing device (e.g., processing device 120B) and stored in a storage device (e.g., storage device 130, storage device 220, storage 390, or an external database). The processing device 120B can retrieve the first training sample (or a part thereof) from the storage device. Alternatively, the first training sample (or a part thereof) can be generated by the processing device 120B.

[0103] In 820, the processing device 120B (e.g., acquisition module 405) can obtain the first initial model.

[0104] In some embodiments, the first initial model can be any type of model (e.g., a machine learning model), such as, for example, a neural network model (e.g., a CNN model, a GAN model), etc. The first initial model can include one or more model parameters. For example, the first initial model can be a CNN model, and exemplary model parameters of the initial model can include the number (or count) of layers, the number (or count) of kernels, the kernel size, the stride, the padding of each convolutional layer, the loss function, etc., or any combination thereof. Before training, the model parameters of the first initial model may have their respective initial values. For example, the processing device 120B can initialize the parameter values of the model parameters of the first initial model.

[0105] In 830, for each of the plurality of first training samples, the processing device 120B (e.g., the training module 406) can generate sample initial amplitude information of the sample first gradient waveform of the first training sample by using the amplitude response function.

[0106] For example, the sample first amplitude information of the sample first gradient waveform can be input into the amplitude response function, and the amplitude response function can output the sample initial amplitude information. In some embodiments, the manner of generating the sample initial amplitude information can be similar to the manner of generating the initial amplitude information of the initial gradient waveform described in operation 610, and will not be elaborated herein.

[0107] In 840, the processing device 120B (e.g., the training module 406) can generate an amplitude determination model by training the first initial model by using the initial amplitude information and the true amplitude information of each sample among the plurality of first training samples.

[0108] In some embodiments, the first initial model can be trained according to the machine learning algorithms (e.g., ​ and related descriptions) described elsewhere in the present invention. For example, the processing device 120B can iteratively update the model parameters of the first initial model by executing one or more iterations according to the supervised machine learning algorithm, so as to generate an amplitude determination model. For illustrative purposes, ​ an exemplary current iteration is shown. The current iteration can be performed based on at least a part of the first training samples. In some embodiments, the same or different sets of first training samples can be used in different iterations of training the first initial model. For the sake of brevity, the first training samples used in the current iteration are referred to as target training samples.

[0109] As ​As shown, for each target training sample, the sample amplitude information can be determined by processing the sample first amplitude information of the sample first gradient waveform of the target training sample. In the current iteration, the updated first initial model generated in the previous iteration can be evaluated. For example, for each target training sample, the processing device 120B can determine the predicted amplitude information of the true gradient waveform of the target training sample by inputting the sample initial amplitude information of the target training sample into the updated first initial model. Then, the processing device 120B can determine the value of the first loss function of the updated first initial model based on the predicted amplitude information and the true amplitude information of the true gradient waveform of each target training sample.

[0110] The first loss function can be used to evaluate the accuracy and reliability of the updated first initial model. For example, the smaller the first loss function, the more reliable the updated first initial model. Exemplary first loss functions can include the L1 loss function, the focal loss function, the logarithmic loss function, the cross-entropy loss function, the Dice loss function, etc. The processing device 120B can further update the values of the model parameters of the updated first initial model based on the value of the loss function according to algorithms such as backpropagation for the next iteration.

[0111] In some embodiments, if the termination condition is met in the current iteration, one or more iterations can be terminated. An exemplary termination condition can be that the value of the loss function obtained in the current iteration is less than a predetermined threshold. Other exemplary termination conditions can include the iteration count having reached a specific number of executions, the loss function converging, the difference in the values of the loss function obtained in consecutive iterations being within a threshold, etc. If the termination condition is met in the current iteration, the processing device 120B can designate the updated first initial model as the amplitude determination model.

[0112] In some embodiments, step 830 can be omitted, and the amplitude determination model can be generated by training the first initial model using the sample first amplitude information and the true amplitude information of each first training sample. For example, during the iteration for generating the amplitude determination model, the predicted amplitude information can be determined based on the sample first amplitude information and the updated first initial model.

[0113] ​ is a flowchart showing an exemplary process for generating a phase determination model according to some embodiments of the present application. In some embodiments, process 1000 can be executed by the MRI system 100. For example, process 1000 can be implemented as a set of instructions (e.g., an application program) stored in a storage device (e.g., storage device 130, storage device 220, and / or storage device 390). In some embodiments, the processing device 120B (e.g., the processor 210 of the computing device 200, the CPU 340 of the mobile device 300, and / or ​One or more modules (as shown) may execute an instruction set and be correspondingly instructed to execute process 1000. In some embodiments, one or more operations of process 1000 may be performed to implement at least a portion of operation 520 as ​ described.

[0114] In 1010, processing device 120B (e.g., acquisition module 405) may obtain a plurality of second training samples.

[0115] Each of the plurality of second training samples may include sample first phase information of a sample first gradient waveform planned to be applied to a sample scan object during a sample MRI scan, and ground truth phase information of a ground truth gradient waveform applied to the sample scan object during the sample MRI scan. More descriptions regarding the sample subject, the sample first gradient waveform, the sample first phase information, the ground truth gradient waveform, and the ground truth phase information can be found elsewhere in this application. See, for example, operation 810 and its related description.

[0116] In step 1020, processing device 120B (e.g., acquisition module 405) may obtain a second initial model.

[0117] The second initial model may be similar to the first initial model described in operation 820. In some embodiments, the first and second initial models may be of the same type or different types of models.

[0118] In step 1030, for each of the plurality of second training samples, processing device 120B (e.g., training module 406) may generate sample initial phase information of the sample first gradient waveform of the second training sample by using a phase response function.

[0119] For example, the sample first phase information of the sample first gradient waveform may be input into the phase response function, and the phase response function may output the sample initial phase information. In some embodiments, the manner of generating the sample initial phase information may be similar to the manner of generating the initial phase information of the initial gradient waveform described in step 610, which will not be elaborated herein.

[0120] In 1040, processing device 120B (e.g., training module 406) may generate a phase determination model by training the second initial model by using the initial phase information and the ground truth phase information of each sample in the plurality of second training samples.

[0121] The generation method of the phase determination model can be similar to the generation method of the amplitude determination model described in step 840. For example, in the iteration for generating the phase determination model, the updated second initial model generated in the previous iteration can be evaluated. The updated second initial model can be used to determine the predicted phase information of the second training sample. The predicted phase information and the true value phase information of the second training sample can be used to determine the value of the second loss function associated with the second initial model. The second loss function can be similar to ​ the first loss function described above. The updated second initial model can be updated based on the value of the second loss function.

[0122] In some embodiments, step 1030 can be omitted, and a phase determination model can be generated by training a second initial model using the sample first phase information and the true value phase information of each second training sample. For example, during the iteration for generating the phase determination model, the predicted phase information can be determined based on the sample first phase information and the updated second initial model.

[0123] ​ is a flowchart showing an exemplary process for jointly generating an amplitude determination model and a phase determination model according to some embodiments of the present application. In some embodiments, process 1100 can be executed by the MRI system 100. For example, process 1100 can be implemented as a set of instructions (e.g., an application) stored in a storage device (e.g., storage device 130, storage device 220, and / or storage device 390). In some embodiments, the processing device 120B (e.g., the processor 210 of the computing device 200, the CPU 340 of the mobile device 300, and / or ​ one or more modules shown in) can execute the instruction set and is accordingly instructed to execute process 1100. In some embodiments, one or more operations of process 1100 can be executed to implement at least a portion of operation 520 as described in ​ the above.

[0124] In step 1110, the processing device 120B (e.g., the acquisition module 405) can obtain a plurality of third training samples.

[0125] Each of the plurality of third training samples can include sample first amplitude information and sample first phase information of a sample first gradient waveform planned to be applied to a sample scan object during a sample MRI scan, as well as true value amplitude information and true value phase information of a true value gradient waveform applied to the sample scan object during the sample MRI scan. More descriptions about the sample scan object, the sample first gradient waveform, the sample first amplitude information, the sample first phase information, the true value gradient waveform, the true value amplitude information, and the true value phase information can be found elsewhere in the present application. See, for example, step 810 and its related descriptions.

[0126] In 1120, a processing device 120B (e.g., the acquisition module 405) may obtain a third initial model including a first sub-model and a second sub-model.

[0127] In some embodiments, the third initial model may be an initial hybrid model including a first sub-model and a second sub-model. The first sub-model may be trained as an amplitude determination model, and the second sub-model may be trained as a phase determination model. In some embodiments, the first and second sub-models may be of the same type or different types of models. In some embodiments, the first sub-model may be similar to the first initial model described in operation 820, and the second sub-model may be similar to the second initial model described in operation 1020.

[0128] In 1130, a processing device 120B (e.g., the acquisition module 405) may generate a trained hybrid model by training the third initial model using a plurality of third training samples.

[0129] The trained hybrid model is generated in a manner similar to the generation manner of the amplitude determination model described in step 840. For example, in an iteration for generating the trained hybrid model, the updated third initial model generated in the previous iteration may be evaluated. The updated third initial model may be used to determine the predicted phase information and predicted amplitude information of the third training sample. The predicted phase information, predicted amplitude information, true value phase information, and true value phase information of the third training sample may be used to determine the value of a third loss function associated with the third initial model. The third loss function may be similar to ​ the first loss function described above. The updated third initial model may be updated based on the value of the third loss function. In some embodiments, the third loss function may include a first component for measuring the difference between the true value amplitude information and the predicted amplitude information, and a second component for measuring the difference between the true value phase information and the predicted phase information. The first component may be used for the first sub-model of the updated third initial model, and the second component may be used for the second sub-model of the updated third initial model.

[0130] In step 1140, a processing device 120B (e.g., the training module 406) may designate the trained first sub-model of the trained hybrid model as the amplitude determination model.

[0131] In step 1150, a processing device 120B (e.g., the training module 406) may designate the trained second sub-model of the trained hybrid model as the phase determination model.

[0132] It should be noted that the above regarding ​For illustrative purposes only and not intended to limit the scope of the present invention. Those of ordinary skill in the art can make various changes and modifications according to the description of this application. However, these changes and modifications do not depart from the scope of this application. In some embodiments, process 800 can be completed with one or more additional operations not described and / or without one or more of the above operations.

[0133] For example, after generating a trained model (amplitude determination model, phase determination model, trained hybrid model), processing device 120B can further test the trained model using a set of test samples. As another example, processing device 120B can update the trained model periodically or irregularly based on one or more newly generated training samples (e.g., new sample first gradient waveform in an MRI scan). As yet another example, the training samples (or a portion thereof) can be pre - processed before model training. By way of example only, one or more waveform signal processing operations (e.g., linearization, denoising, filtering, sharpening, etc.) can be performed on the sample first gradient waveform. As yet another example, before step 1130, processing device 120B can determine the sample initial amplitude information and sample initial phase information of each third training sample, and train the third initial model using the sample initial amplitude information, sample initial phase information, true amplitude information, and true phase information of each third training sample.

[0134] ​ is a flowchart showing an exemplary process for performing an MRI scan on a scanned object according to some embodiments of the present application. In some embodiments, process 1200 can be executed by MRI system 100. For example, process 1200 can be implemented as a set of instructions (e.g., an application program) stored in a storage device (e.g., storage device 130, storage device 220, and / or storage device 390). In some embodiments, processing device 120A (e.g., processor 210 of computing device 200, CPU 340 of mobile device 300, and / or ​ one or more modules shown in) can execute the instruction set and be correspondingly instructed to execute process 1200. The operations of the steps shown below are for illustrative purposes only. In some embodiments, process 1200 can be completed with one or more additional operations not described and / or omitting one or more of the operations described below. Additionally, ​ the order of operations of process 1200 shown and described below is not intended to be limiting.

[0135] In operation 1210, a processing device 120A (e.g., the acquisition module 401) may obtain a first gradient waveform that is planned to be applied to a scanned object. For example, the processing device 120A may obtain first amplitude information and first phase information of the first gradient waveform. More descriptions about the first gradient waveform can be found elsewhere in this application. See, for example, operation 510 and its related descriptions.

[0136] In operation 1220, the processing device 120A (e.g., the determination module 402) may determine a second gradient waveform based on the first gradient waveform and a gradient waveform determination model, where the gradient waveform determination model is trained according to a machine learning algorithm.

[0137] When performing an MRI scan according to the first gradient waveform, the actual gradient waveform actually applied to the scanned object may be different from the first gradient waveform. The second gradient waveform may be regarded as an estimated value of the actual gradient waveform. Operation 1220 may be performed in a manner similar to the described operation 520, which will not be elaborated here. ​ Operation 1220 may be performed in a manner similar to the described operation 520, which will not be elaborated here.

[0138] In 1230, the processing device 120A (e.g., the control module 404) may, based on the second gradient waveform, instruct the MRI scanner to perform an MRI scan on the scanned object.

[0139] In some embodiments, the processing device 120A may determine one or more scan parameters for implementing the second gradient waveform, and instruct the MRI scanner to perform an MRI scan on the scanned object based on the one or more scan parameters.

[0140] In some embodiments, the processing device 120A (e.g., the determination module 404) may determine an adjusted gradient waveform by adjusting the first gradient waveform according to the second gradient waveform. In addition, the processing device 120A (e.g., the control module 404) may instruct the MRI scanner to perform an MRI scan on the scanned object according to the adjusted gradient waveform.

[0141] As described above, due to hardware limitations, the actual gradient waveform applied to the scanned object during an MRI scan is usually different from the ideal first gradient waveform. The processing device 120A may determine an adjusted gradient waveform according to the second gradient waveform (i.e., the estimated actual gradient waveform when performing an MRI scan according to the first gradient waveform), so that when performing an MRI scan according to the adjusted gradient waveform, the actual gradient waveform applied to the scanned object under the influence of hardware limitations can be as close as possible to the first gradient waveform.

[0142] For example, by comparing the first gradient waveform and the second gradient waveform, the processing device 120A can determine the rule that the hardware limitation affects the application of the first gradient waveform, and adjust the first gradient waveform based on this rule. Merely as an example, if the rule shows that the second gradient waveform rapidly decreases during a certain period while the first gradient waveform is in a stable state during this period, the processing device 120A can adjust the first gradient waveform to an ascending state during this period. In this case, when performing an MRI scan according to the adjusted gradient waveform, even if the adjusted part during the period decreases due to hardware limitations, the actual gradient waveform may be close to the first gradient waveform.

[0143] By determining the adjusted gradient waveform based on the second gradient waveform and performing an MRI scan according to the adjusted gradient waveform, the actual gradient waveform applied to the scanned object during the MRI scan can be close to the ideal first gradient waveform, which can achieve the desired scanning effect. In some embodiments, the MRI scan can be an ultra-short echo time MRI or a spiral MRI scan. In some embodiments, the processing device 120A can also generate a target reconstruction image of the scanned object based on the first gradient waveform and the scan data collected during the MRI scan. As described in connection with operation 1230, when performing an MRI scan according to the adjusted gradient waveform, the actual gradient waveform applied to the scanned object during the scan can be close to the first gradient waveform. Reconstructing the target reconstruction image based on the first gradient waveform can improve the reconstruction accuracy.

[0144] The basic concepts have been described above. Obviously, for those of ordinary skill in the art after reading this application, the above invention disclosure is only an example and does not constitute a limitation to this application. Although not explicitly stated here, those of ordinary skill in the art may make various modifications, improvements, and corrections to this application. Such modifications, improvements, and corrections are proposed in this application, so such modifications, improvements, and corrections still fall within the spirit and scope of the exemplary embodiments of this application.

[0145] Meanwhile, this application uses specific terms to describe the embodiments of this application. For example, "one embodiment", "an embodiment", and / or "some embodiments" mean a certain feature, structure, or characteristic related to at least one embodiment of this application. Therefore, it should be emphasized and noted that the "one embodiment" or "an embodiment" or "an alternative embodiment" mentioned two or more times in different positions in this specification is not necessarily referring to the same embodiment. In addition, certain features, structures, or characteristics in one or more embodiments of this application can be appropriately combined.

[0146] In addition, those of ordinary skill in the art will understand that various aspects of the present application can be illustrated and described by several patentable types or situations, including any new and useful process, machine, product, or composition of matter, or any new and useful improvement thereof. Thus, various aspects of the present invention can be implemented entirely in hardware, software (including firmware, resident software, microcode, etc.), or a combination of software and hardware implementations, which are generally referred to herein as "units", "modules", or "systems". In addition, various aspects of the present invention can take the form of a computer program product that includes computer-readable program code in one or more computer-readable media.

[0147] A computer-readable signal medium can include a propagated data signal that contains computer program code, for example, on a baseband or as part of a carrier wave. Such propagated signals can come in many forms, including electromagnetic, optical, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium that can communicate, propagate, or transport a program for use by connecting to an instruction execution system, apparatus, or device. The program code located on the computer-readable signal medium can be propagated through any suitable medium, including radio, cable, fiber optic cable, RF, etc., or any combination of the above media.

[0148] The computer program code for operating various aspects of the present application can be written in any combination of one or more programming languages, including object-oriented programming languages such as Java, Scala, Smalltalk, Eiffel, Jade, Emerald, C++, C#, VB, NET, Python, etc., traditional procedural programming languages such as the "C" programming language, Visual Basic, Fortran2103, Perl, Cobol2102, PHP, ABAP, dynamic programming languages such as Python, Ruby, and Groovy, or other programming languages. The program code can execute entirely on the user's computer, as a stand-alone software package, partially on the user's computer, partially on a remote computer, or entirely on a remote computer or server. In the latter case, the remote computer can be connected to the user's computer through any type of network connection, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., through the Internet using an Internet service provider), or in a cloud computing environment, or as a service offering, such as software as a service (SaaS).

[0149] Moreover, unless explicitly stated in the claims, the order of the processing elements and sequences, the use of numerical and alphabetical characters, or the use of other names in this application are not used to limit the order of the processes and methods of this application. Although some currently useful embodiments of the invention are discussed through various examples in the above disclosure, it should be understood that such details are only for illustrative purposes, and the appended claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that conform to the essence and scope of the embodiments of this application. For example, although the implementation of the various components described above can be embodied in a hardware device, it can also be implemented as a software-only solution, for example, an installation on an existing server or mobile device.

[0150] Similarly, it should be noted that, in order to simplify the presentation of the disclosure of this application and thus help the understanding of one or more embodiments of the invention, in the foregoing description of the embodiments of this application, sometimes multiple features are merged into one embodiment, drawing, or description thereof. However, this method of this application should not be construed as reflecting an intention that the claimed subject matter of the object to be scanned requires more features than are expressly recited in each claim. On the contrary, the subject matter of the invention should have fewer features than the above single embodiment.

[0151] In some embodiments, the numbers representing the amounts or characteristics used to describe and claim certain embodiments of this application should be understood to be modified in certain cases by the terms "about", "approximate", or "substantially". For example, "about", "approximate", or "substantially" may represent a variation of ±1%, ±5%, ±10%, or ±20% of the value they describe, unless otherwise stated. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, and these approximate values may change according to the characteristics required by individual embodiments. In some embodiments, the numerical parameters should consider the specified significant digits and adopt the method of retaining the general number of digits. Although the numerical ranges and parameters used to confirm the breadth of their scope in some embodiments of this application are approximate values, in specific embodiments, the setting of such numerical values is as precise as possible within the feasible range.

[0152] Each patent, patent application, patent application publication, and other materials cited herein, such as articles, books, specifications, publications, documents, articles, and / or the like, are hereby incorporated by reference in their entirety for all purposes, except for any prosecution file history associated therewith, any content inconsistent with or conflicting with this document, or any content that has a limiting effect on the broadest scope of the claims currently or hereafter related to this document. For example, if there is any inconsistency or conflict between the description, definition, and / or use of terms associated with any incorporated material and the terms associated with this document, the terms described, defined, and / or used in this document shall prevail.

[0153] Finally, it should be understood that the embodiments described in this application are only used to illustrate the principles of the embodiments of this application. Other variations may also fall within the scope of this application. Therefore, by way of example and not limitation, alternative configurations of the embodiments of this application may be regarded as consistent with the teachings of this application. Accordingly, the embodiments of this application are not limited to the embodiments explicitly presented and described in this application.

Claims

1. A magnetic resonance imaging (MRI) method, executed by at least one processor, characterized in that, The method includes: Obtaining MRI scan data of the scanned object by instructing an MRI scanner to perform an MRI scan on the scanned object according to a first gradient waveform, where the first gradient waveform is an ideal gradient waveform; Determining a second gradient waveform based on the first gradient waveform and a gradient waveform determination model, where the second gradient waveform is an estimate of the actual gradient waveform actually applied to the scanned object, and the gradient waveform determination model is trained according to a machine learning algorithm; and Generating a target reconstructed image of the scanned object based on the second gradient waveform and the MRI scan data.

2. The method according to claim 1, characterized in that, The determining the second gradient waveform based on the first gradient waveform and the gradient waveform determination model includes: Determining an initial gradient waveform by processing the first gradient waveform using at least one response function; and Determining the second gradient waveform based on the initial gradient waveform and the gradient waveform determination model.

3. The method according to claim 2, wherein The at least one response function includes an amplitude response function and a phase response function, and The determining the initial gradient waveform by processing the first gradient waveform using the at least one response function includes: Determining initial amplitude information of the initial gradient waveform by processing first amplitude information of the first gradient waveform using the amplitude response function; and Determining initial phase information of the initial gradient waveform by processing first phase information of the first gradient waveform using the phase response function.

4. The method according to claim 3, wherein The gradient waveform determination model includes an amplitude determination model and a phase determination model, and The determining the second gradient waveform based on the initial gradient waveform and the gradient waveform determination model includes: Determining second amplitude information of the second gradient waveform by processing the initial amplitude information using the amplitude determination model; and Determining second phase information of the second gradient waveform by processing the initial phase information using the phase determination model.

5. The method according to claim 1, wherein The generating the target reconstructed image of the scanned object based on the second gradient waveform and the MRI scan data includes: Filling the MRI scan data into k-space according to the second gradient waveform to generate k-space data; Performing an inverse Fourier transform on the k-space data to reconstruct the target reconstructed image of the scanned object.

6. A magnetic resonance imaging (MRI) method, executed by at least one processor, characterized in that, The method includes: Obtaining a first gradient waveform planned to be applied to a scanned object during an MRI scan, where the first gradient waveform is an ideal gradient waveform; Determining a second gradient waveform based on the first gradient waveform and a gradient waveform determination model, where the second gradient waveform is an estimate of the actual gradient waveform actually applied to the scanned object, and the gradient waveform determination model is trained according to a machine learning algorithm; and Based on the second gradient waveform, instructing an MRI scanner to perform an MRI scan on the scanned object.

7. The method according to claim 6, wherein The instructing the MRI scanner to perform an MRI scan on the scanned object based on the second gradient waveform includes: Adjusting the first gradient waveform according to the second gradient waveform to determine an adjusted gradient waveform; and Instructing the MRI scanner to perform the MRI scan according to the adjusted gradient waveform.

8. The method according to claim 6, wherein Determining the second gradient waveform based on the first gradient waveform and the gradient waveform determination model includes: Determining an initial gradient waveform by processing the first gradient waveform using at least one response function; and Determining the second gradient waveform based on the initial gradient waveform and the gradient waveform determination model.

9. The method according to claim 8, wherein The gradient waveform determination model includes an amplitude determination model and a phase determination model, and determining the second gradient waveform based on the initial gradient waveform and the gradient waveform determination model includes: Determining second amplitude information of the second gradient waveform by processing initial amplitude information of the initial gradient waveform using the amplitude determination model; and Determining second phase information of the second gradient waveform by processing initial phase information of the initial gradient waveform using the phase determination model.

10. A magnetic resonance imaging (MRI) system, comprising: At least one storage device for storing computer instructions; At least one processor for executing the computer instructions to implement the method according to any one of claims 1-9.

Images