A magnetic resonance imaging method, system and computer readable storage medium

Abstract

Embodiments of the present specification provide a magnetic resonance imaging method, system and computer readable storage medium. The method comprises: acquiring imaging data; determining a data block containing an abnormal point in the imaging data, the data block being a part of the imaging data; correcting the imaging data based on the data block containing the abnormal point; and generating a magnetic resonance image based on the corrected imaging data.

Description

Technical Field

[0001] This specification relates to the field of medical technology, and in particular to a magnetic resonance imaging method and system. Background Technology

[0002] Magnetic resonance imaging (MRI) is a medical imaging technique used for medical diagnosis. MRI scanners use strong magnetic fields, magnetic field gradients, and radio waves to generate images of the target object being scanned (e.g., tissues or organs in the human body). However, MRI equipment is prone to artifacts in the acquired images due to intermittent system malfunctions, such as poor contact in electronic switches during high-speed analog-to-digital converter sampling, poor contact in other systems, discharge at wiring tips, or transient radio frequency signal crosstalk from other wireless transmitters. These artifacts can severely affect a doctor's diagnosis.

[0003] Therefore, it is desirable to provide a magnetic resonance imaging method that can improve imaging quality while achieving faster imaging speed. Summary of the Invention

[0004] This specification provides a magnetic resonance imaging method. The method includes: acquiring imaging data; determining data blocks containing anomalous points in the imaging data, the data blocks being a portion of the imaging data; and correcting the imaging data based on the data blocks containing the anomalous points.

[0005] In some embodiments, determining the data block containing anomalies in the imaging data includes: dividing the imaging data into multiple data blocks according to a first preset parameter; and inputting the multiple data blocks into a trained first classification model to obtain the data block containing anomalies.

[0006] In some embodiments, the imaging data includes K-space data; the method further includes: acquiring a reconstructed image corresponding to the K-space data; determining whether the reconstructed image contains artifacts; and determining whether the K-space data contains outliers based on the determination result of the reconstructed image.

[0007] In some embodiments, determining whether the reconstructed image contains artifacts includes: extracting at least one image patch from the reconstructed image according to a second preset parameter; and determining whether the reconstructed image contains artifacts based on the at least one image patch using a trained second classification model.

[0008] In some embodiments, correcting the imaging data based on the data block containing the outliers includes: determining the location information of the outliers; performing artifact correction on the data block containing the outliers based on the location information to obtain a corrected data block; and updating the imaging data based on the corrected data block.

[0009] In some embodiments, determining the location information of the anomaly includes: determining a reference position of the data block containing the anomaly in the imaging data; and using a semantic segmentation neural network to determine the location information of the anomaly in the imaging data based on the reference position.

[0010] In some embodiments, correcting the imaging data based on the data block containing the outliers includes: determining the location information of the outliers using a trained artifact correction model; and performing artifact correction on the data block containing the outliers based on the location information to obtain a corrected data block.

[0011] In some embodiments, the method further includes: generating a magnetic resonance image based on the corrected imaging data; wherein arcing artifacts in the corrected imaging data are suppressed or eliminated.

[0012] Another aspect of this specification provides a magnetic resonance imaging system. The system includes: an acquisition module for acquiring imaging data; a detection module for identifying data blocks containing anomalous points within the imaging data, the data blocks being a portion of the imaging data; and a correction module for correcting the imaging data based on the data blocks containing the anomalous points.

[0013] Another aspect of this specification provides a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, when the processor executes the computer program, the computer program is used to cause the computer device to perform the magnetic resonance imaging method as described above.

[0014] Another aspect of this specification provides a computer-readable storage medium that stores computer instructions, which, when read by a computer, execute the magnetic resonance imaging method as described above. Attached Figure Description

[0015] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting; in these embodiments, the same reference numerals denote the same structures, wherein:

[0016] Figure 1 These are schematic diagrams illustrating application scenarios of exemplary magnetic resonance imaging systems according to some embodiments of this specification;

[0017] Figure 2 This is a schematic diagram of a module of an exemplary magnetic resonance imaging system according to some embodiments of this specification;

[0018] Figure 3This is a schematic flowchart of an exemplary magnetic resonance imaging method according to some embodiments of this specification;

[0019] Figure 4 These are schematic diagrams of exemplary imaging data shown according to some embodiments of this specification;

[0020] Figure 5 These are schematic diagrams of exemplary reconstructed images shown according to some embodiments of this specification;

[0021] Figure 6 This is a schematic diagram illustrating exemplary imaging data partitioning according to some embodiments of this specification;

[0022] Figure 7 This is a schematic diagram illustrating exemplary anomaly point location according to some embodiments of this specification;

[0023] Figures 8-11 This is a schematic flowchart of an exemplary magnetic resonance imaging method according to other embodiments of this specification. Detailed Implementation

[0024] To more clearly illustrate the technical solutions of the embodiments in this specification, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are merely some examples or embodiments of this specification. For those skilled in the art, these drawings can be applied to other similar scenarios without creative effort. Unless obvious from the context or otherwise specified, the same reference numerals in the drawings represent the same structures or operations.

[0025] It should be understood that the terms “system,” “device,” “unit,” and / or “module” used herein are one way to distinguish different components, elements, parts, sections, or assemblies at different levels. However, if other terms can achieve the same purpose, they may be replaced by other expressions.

[0026] 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 on 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 invoked from other modules / units / blocks or from themselves, and / or can be invoked in response to detected events or interrupts. Software modules / units / blocks configured to execute on a computing device can be provided on computer-readable media (e.g., optical discs, digital video discs, flash drives, magnetic disks, or any other tangible media) or as digital downloads (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 operation and applied in the operation of the computing device. Software instructions can be embedded in firmware, such as EPROM. It should also be understood that hardware modules / units / blocks may be included in connected logical components, such as gates and flip-flops, and / or may include programmable units, such as programmable gate arrays or processors. The modules / units / blocks or computing device functions described herein may be implemented as software modules / units / blocks, but can be represented in hardware or firmware. Typically, the modules / units / blocks described herein refer to logical modules / units / blocks that can be combined with other modules / units / blocks or divided into submodules / subunits / subblocks, although they are physical organization or storage devices. This description may apply to a system, an engine, or a part thereof.

[0027] It is understood that, unless the context explicitly states otherwise, when a unit, engine, module, or block is referred to as being "on," "connected," or "coupled to" another unit, engine, module, or block, it may be directly on, connected to, coupled to, or communicate with that other unit, engine, module, or block, or there may be intermediate units, engines, modules, or blocks. In this specification, the term "and / or" may include any one or more of the relevant listed items or a combination thereof. In this specification, the term "image" may refer to a 2D image, a 3D image, or a 4D image.

[0028] These and other features, characteristics, functions and operating methods of related structural elements, as well as the assembly and manufacturing economy of components, will become more apparent from the following description of the accompanying drawings, all of which form part of this specification. However, it should be understood that the drawings are for illustrative purposes only and are not intended to limit the scope of this specification. It should also be understood that the drawings are not drawn to scale.

[0029] As indicated in this specification and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" do not specifically refer to the singular and may also include the plural. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of expressly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0030] Flowcharts are used in this specification to illustrate the operations performed by the system according to embodiments of this specification. The related descriptions are provided to aid in a better understanding of the magnetic resonance imaging method and / or system. It should be understood that preceding or subsequent operations are not necessarily performed precisely in sequence. Instead, steps can be processed in reverse order or simultaneously. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0031] This specification discloses a magnetic resonance imaging method. By dividing the imaging data into multiple data blocks and inputting these data blocks into a trained first classification model to obtain data blocks containing outliers, artifact correction is performed on the imaging data based on the data blocks containing outliers. This method can improve the accuracy and efficiency of artifact correction and save time and costs.

[0032] Figure 1 This is a schematic diagram illustrating an application scenario of an exemplary magnetic resonance imaging system according to some embodiments of this specification.

[0033] like Figure 1 As shown, in some embodiments, the magnetic resonance imaging system 100 may include an imaging device 110, a processing device 120, a terminal 130, a storage device 140, and a network 150. In some embodiments, the various components in the magnetic resonance imaging system 100 may be interconnected via the network 150 or directly connected without using the network 150. For example, the imaging device 110 and the terminal 130 may be connected via the network 150. As another example, the imaging device 110 and the processing device 120 may be connected via the network 150 or directly connected. As yet another example, the imaging device 120 and the terminal 130 may be connected via the network 150 or directly connected.

[0034] Imaging device 110 can be used to scan a target object or a portion thereof located within its detection area and generate an image relating to the target object or a portion thereof. In some embodiments, the target object can be biological or non-biological. For example, the target object can include a patient, a man-made object, etc. In some embodiments, the target object can include a specific part of the body, such as the head, chest, abdomen, etc., or any combination thereof. In some embodiments, the target object can include a specific organ, such as the heart, esophagus, trachea, bronchi, stomach, gallbladder, small intestine, colon, bladder, ureter, uterus, fallopian tubes, etc., or any combination thereof. In some embodiments, the target object can include a region of interest (ROI), such as a tumor, a node, etc.

[0035] In some embodiments, the imaging device 110 may include one or a combination of several of the following: X-ray device, computed tomography (CT) device, three-dimensional (3D) CT, four-dimensional (4D) CT, ultrasound imaging assembly, fluorescence fluoroscopy imaging assembly, magnetic resonance imaging (MRI) device, single-photon emission computed tomography (SPECT) device, positron emission tomography (PET) device, etc.

[0036] In some embodiments, the imaging device 110 may be an MRI device. In some embodiments, the MRI device may include a magnet assembly, a gradient coil assembly, and a radio frequency (RF) coil assembly.

[0037] The magnet assembly can generate a primary magnetic field (also known as the main magnetic field) to polarize the object being scanned. For example, the magnet assembly can include permanent magnets, superconducting electromagnets, resistive electromagnets, etc.

[0038] Gradient coil assemblies can generate a second magnetic field (also called a gradient magnetic field). For example, a gradient coil assembly may include an X-gradient coil, a Y-gradient coil, and a Z-gradient coil. The gradient coil assembly can generate one or more magnetic field gradient pulses in the X (Gx), Y (Gy), and Z (Gz) directions relative to the main magnetic field to encode spatial information of the scanned object. In some embodiments, the X direction can be specified as the frequency encoding direction, and the Y direction can be specified as the phase encoding direction. In some embodiments, Gx can be used for frequency encoding or signal readout, and is generally referred to as the frequency-encoded gradient or readout gradient. In some embodiments, Gy can be used for phase encoding, and is generally referred to as the phase-encoded gradient. In some embodiments, Gz can be used for slicing to obtain two-dimensional K-space data. In some embodiments, Gz can be used for phase encoding to obtain three-dimensional K-space data.

[0039] An RF coil assembly may include at least two RF coils. Each RF coil may include one or more RF transmitting coils and / or one or more RF receiving coils. The RF transmitting coil can transmit RF pulses toward the object to be scanned (e.g., a target object). Under the combined action of the main magnetic field / gradient magnetic field and the RF pulses, a magnetic resonance signal related to the target object can be generated based on the pulse sequence. The RF receiving coil can acquire the magnetic resonance signal from the object based on the pulse sequence. Transform operations (e.g., Fourier transform) can be used to process the magnetic resonance signal to fill the K-space and acquire K-space data.

[0040] Processing device 120 can process data and / or information obtained from imaging device 110, terminal 130, and / or storage device 140. For example, processing device 120 can process K-space data detected by imaging device 110 to obtain corrected imaging data. As another example, processing device 120 can process MRI reconstructed images to obtain imaging data. In some embodiments, processing device 120 can be a single server or a group of servers. The server group can be centralized or distributed. In some embodiments, processing device 120 can be local or remote. For example, processing device 120 can access information and / or data from imaging device 110, terminal 130, and / or storage device 140 via network 150. As another example, processing device 120 can be directly connected to imaging device 110, terminal 130, and / or storage device 140 to access information and / or data. In some embodiments, processing device 120 can be implemented on a cloud platform. For example, the cloud platform can include private cloud, public cloud, hybrid cloud, community cloud, distributed cloud, inter-cloud cloud, multi-cloud, etc., or any combination thereof.

[0041] Terminal 130 may include mobile device 131, tablet computer 132, laptop computer 133, etc., or any combination thereof. In some embodiments, terminal 130 may interact with other components in magnetic resonance imaging system 100 via network 150. For example, terminal 130 may send one or more control commands to imaging device 110 via network 150 to control imaging device 110 to scan a target object according to the commands. As another example, terminal 130 may also receive magnetic resonance images generated by processing device 120 via network 150 and display the magnetic resonance images for analysis and confirmation by the operator. In some embodiments, mobile device 131 may include smart home devices, wearable devices, mobile devices, virtual reality devices, augmented reality devices, etc., or any combination thereof.

[0042] In some embodiments, terminal 130 may be part of processing device 120. In some embodiments, terminal 130 may be integrated with processing device 120 as a control panel for imaging device 110. For example, the user / operator of magnetic resonance imaging system 100 (e.g., a doctor or nurse) may control the operation of imaging device 110 through the control panel, such as scanning a target object.

[0043] Storage device 140 may store data (e.g., k-space data of a target object, reconstructed images, magnetic resonance images, etc.), instructions, and / or any other information. In some embodiments, storage device 140 may store data acquired from imaging device 110, processing device 120, and / or terminal 130. For example, storage device 140 may store k-space data of a target object, reconstructed images, etc., acquired from imaging device 110. In some embodiments, storage device 140 may store data and / or instructions that processing device 120 may execute or use to perform the exemplary methods described herein.

[0044] In some embodiments, storage device 140 may include one or a combination of several of the following: mass storage, removable storage, volatile read-write storage, read-only storage (ROM). In some embodiments, storage device 140 may be implemented using the cloud platform described herein.

[0045] In some embodiments, storage device 140 may be connected to network 150 to communicate with one or more components of magnetic resonance imaging system 100 (e.g., processing device 120, terminal 130). One or more components of magnetic resonance imaging system 100 may read data or instructions from storage device 140 via network 150. In some embodiments, storage device 140 may be part of processing device 120 or may be independent and directly or indirectly connected to processing device 120.

[0046] Network 150 may include any suitable network capable of facilitating information and / or data exchange between the magnetic resonance imaging system 100 and the magnetic resonance imaging system 100. In some embodiments, one or more components of the magnetic resonance imaging system 100 (e.g., imaging device 110, processing device 120, terminal 130, storage device 140) may exchange information and / or data with one or more components of the magnetic resonance imaging system 100 via network 150. For example, processing device 120 may acquire imaging data of a target object from imaging device 110 via network 150. In some embodiments, network 150 may include one or more combinations of public networks (e.g., the Internet), private networks (e.g., local area networks (LANs), wide area networks (WANs)), wired networks (e.g., Ethernet), wireless networks (e.g., 802.11 networks, Wi-Fi networks), cellular networks (e.g., LTE networks), Frame Relay networks, virtual private networks (VPNs), satellite networks, telephone networks, routers, hubs, server computers, etc. 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 switching points, through which one or more components of the magnetic resonance imaging system 100 may connect to network 150 to exchange data and / or information.

[0047] It should be noted that the above description is provided for illustrative purposes only and is not intended to limit the scope of this specification. Various changes and modifications can be made by those skilled in the art based on the content of this specification. Features, structures, methods, and other features of the exemplary embodiments described herein can be combined in various ways to obtain other and / or alternative exemplary embodiments. For example, system 100 may also include a display device for outputting and displaying magnetic resonance images generated by processing device 120. However, these changes and modifications will not depart from the scope of this specification.

[0048] Figure 2 This is a schematic diagram of a module of an exemplary magnetic resonance imaging system according to some embodiments of this specification.

[0049] like Figure 2 As shown, in some embodiments, the magnetic resonance imaging system 200 may include an acquisition module 210, a detection module 220, and a calibration module 230. In some embodiments, one or more modules of the magnetic resonance imaging system 200 may be interconnected. The connection may be wireless or wired. At least a portion of the magnetic resonance imaging system 200 may be configured as follows: Figure 1 It is implemented on the imaging device 110, processing device 120 or terminal 130 shown.

[0050] The acquisition module 210 can be used to acquire imaging data. In some embodiments, the imaging data may include K-space data. In some embodiments, the acquisition module 210 can acquire a reconstructed image corresponding to the K-space data.

[0051] The detection module 220 can be used to identify data blocks in the imaging data that contain anomalous points. In some embodiments, the data block can be a portion of the imaging data.

[0052] In some embodiments, the detection module 220 may further include a first detection unit 223 and a second detection unit 225.

[0053] The first detection unit 223 can be used to determine whether the K-space data contains outliers. In some embodiments, the first detection unit 223 can divide the imaging data into multiple data blocks according to a first preset parameter, and input the multiple data blocks into a trained first classification model to obtain data blocks containing outliers. In some embodiments, the first detection unit 223 can determine whether the K-space data contains outliers based on the judgment result of the reconstructed image.

[0054] The second detection unit 225 can be used to determine whether the reconstructed image contains artifacts. In some embodiments, the second detection unit 225 can extract at least one image patch from the reconstructed image according to a second preset parameter, and use a trained second classification model based on the at least one image patch to determine whether the reconstructed image contains artifacts. In some embodiments, the first classification model and / or the second classification model may include a convolutional neural network.

[0055] The correction module 230 can be used to correct imaging data based on data blocks containing outliers. In some embodiments, the correction module 230 can generate a magnetic resonance image based on the corrected imaging data.

[0056] In some embodiments, the correction module 230 may further include a positioning unit 233 and a correction unit 235.

[0057] In some embodiments, the positioning unit 233 can be used to determine the location information of anomalies. In some embodiments, the correction unit 235 can perform artifact correction on the data block containing the anomalies based on the location information to obtain the corrected data block, and update the imaging data based on the corrected data block.

[0058] In some embodiments, the positioning unit 233 can determine the reference position of the data block containing the anomaly in the imaging data, and use a semantic segmentation neural network to determine the location information of the anomaly in the imaging data based on the reference position.

[0059] In some embodiments, the correction unit 235 can simultaneously determine the location information of outliers and perform artifact correction on the data block based on the location information using a trained artifact correction model to obtain a corrected data block.

[0060] In some embodiments, the artifact correction method may include at least one of the following: setting outliers to be uncollected, filling, interpolation, or neural network assignment.

[0061] In some embodiments, the acquisition module 210, the detection module 220, and the correction module 230 may be modules in the same or different processing devices. For example, the acquisition module 210, the detection module 220, and the correction module 230 may all be modules in the imaging device 110 or the processing device 120. Alternatively, the acquisition module 210 and the detection module 220 may be modules in the processing device 120, while the correction module 230 may be a module in another processing device (such as the imaging device 110) besides the processing device 120.

[0062] It should be noted that the above description of the magnetic resonance imaging system 200 is for illustrative purposes only and is not intended to limit the scope of this specification. Various modifications and variations can be made based on this specification by those skilled in the art. However, these changes and modifications do not depart from the scope of this specification. For example, one or more modules of the magnetic resonance imaging system 200 described above may be omitted or integrated into a single module. As another example, the magnetic resonance imaging system 200 may include one or more additional modules, such as a storage module for data storage.

[0063] Figure 3 This is a schematic flowchart of an exemplary magnetic resonance imaging method according to some embodiments of this specification.

[0064] In some embodiments, process 300 may be executed by imaging device 110, processing device 120, or magnetic resonance imaging system 200. For example, process 300 may be implemented as instructions (e.g., an application program) and stored in a memory external to, for example, storage device 140 or the magnetic resonance imaging system (e.g., magnetic resonance imaging system 100 or magnetic resonance imaging system 200) and accessible by imaging device 110, processing device 120, or magnetic resonance imaging system 200. Imaging device 110, processing device 120, or magnetic resonance imaging system 200 may execute the instructions, and when executing the instructions, may be configured to execute process 300. The operational schematic diagram of process 300 presented below is illustrative. In some embodiments, the process may be accomplished using one or more additional operations not described and / or one or more operations not discussed. Additionally, Figure 3 The sequence of operations shown in the diagram and described below for process 300 is non-restrictive.

[0065] Step 310: Acquire imaging data. In some embodiments, step 310 may be performed by imaging device 110, processing device 120, or acquisition module 210.

[0066] Imaging data can refer to data that has defects and needs to be processed, such as K-space data containing bright spots, artifacts, or white spots.

[0067] In some embodiments, imaging data may include K-space data of the target object. Images generally have two representations: the pixel domain (image domain) and the frequency domain. The conversion between these two forms is called the Fourier transform. The Fourier transform is a transformation from the pixel domain to the frequency domain, and the inverse Fourier transform is a transformation from the frequency domain to the pixel domain. K-space can refer to the space used to store information in the frequency domain of the image. As an example only, during the imaging process, magnetic resonance imaging (MRI) equipment uses pulses of corresponding frequencies with three gradients to spatially encode the MRI signal (slice selection, frequency encoding, phase encoding), and then converts the acquired analog echo signal containing spatial information into a digital signal to fill the K-space, obtaining K-space data, such as... Figure 4 As shown on the left side of the middle line; finally, the MR signals of different frequencies, phases, and amplitudes are analyzed by inverse Fourier transform. The MR digital signals of different frequencies, phases, and signal intensities are then assigned to the corresponding pixels to obtain the entity image of the target object (i.e., the pixel-domain image), as shown below. Figure 4 The image domain shown to the right of the central line. Different frequencies and phases represent different spatial locations, while amplitude represents the MR signal intensity. In some embodiments, K-space data may include 2D K-space data and 3D K-space data. For example, 3D K-space data corresponding to the three coordinate axes of layer selection, frequency encoding, and phase encoding, and 2D K-space data corresponding to any two of the coordinate systems of layer selection, frequency encoding, and phase encoding. In some embodiments, K-space data may include fully sampled K-space data and downsampled K-space data. For example, Figure 4 The K-space data shown to the left of the central line is a downsampled image, where the white box corresponds to the uncollected K-space data.

[0068] In some embodiments, imaging data may include a pixel-domain image of the target object, such as a reconstructed image. In some embodiments, imaging data may be acquired by an imaging device, such as K-space data generated by scanning with imaging device 110, K-space data stored by imaging device 110, etc. In some embodiments, imaging data may be acquired from a storage device (such as storage device 140). In some embodiments, imaging data may be obtained through Fourier transform.

[0069] Step 320: Identify data blocks in the imaging data that contain anomalous points. In some embodiments, step 320 may be performed by the imaging device 110, the processing device 120, or the detection module 220.

[0070] Anomalies refer to points in imaging data that differ from normal data, such as straight lines, bright spots, ring artifacts, white spots, and black spots. As an example, in MRI equipment, occasional system malfunctions such as poor contact in the electronic switch of the high-speed analog-to-digital converter (ADC), poor contact in other system modules, discharge at wiring tips, or transient radio frequency signal crosstalk from other wireless transmitters can easily lead to artifacts in the clinical images obtained by MRI equipment, which manifest as randomly occurring bright spots in K-space data. For example... Figure 4 As shown, due to an occasional system malfunction, spark bright spots (shown in the white ellipse) are present in the K-space data acquired by the MRI imaging equipment, and diagonal spark artifacts are present in the corresponding image domain image (to the right of the straight line).

[0071] In some embodiments, a data block may be a portion of imaging data. In some embodiments, the imaging data may be divided into multiple data blocks according to a first preset parameter; the multiple data blocks are input into a first classification model to obtain data blocks containing outliers.

[0072] In some embodiments, the first preset parameter can reflect parameters such as the size (e.g., length and width, diameter, perimeter, etc.) and shape (e.g., square, rectangle, polygon, circle, etc.) of the data block. For example, according to the first preset parameter 32*32, the K-space data of 426*1024 can be sliced into 14*32 = 448 data blocks of size 32*32. In some embodiments, the first preset parameter can be determined according to at least one of properties such as the dimension of the imaging data, the data point acquisition and arrangement method, etc. Only as an example, when the imaging data is one-dimensional linear data (such as 1*N), the data block can be selected as a line segment of a certain length (such as 1*M, M < N); when the imaging data is two-dimensional data (such as N*N), the data block can be selected as a rectangle (such as M*M, M < N); when the imaging data is three-dimensional data (such as N*N*N), the data block can be selected as a three-dimensional cube (such as M*M*M, M < N). In some embodiments, the first preset parameter can be determined according to the K-space. For example, according to the size of the K-space corresponding to the MRI imaging device, the data block size is determined to be 32*32. In some embodiments, the first preset parameter can be determined according to the clinical scan parameters, such as determining according to the K-space size set in the clinical scan. In some embodiments, the first preset parameter can be adjusted in real time according to the actual situation. For example, different first preset parameters can be set according to different scan scenarios each time. In some embodiments, the first preset parameter can be automatically set or adjusted. For example, the system can automatically determine the first preset parameter or adjust it in real time according to the target object information, big data statistical information, the scanning parameter setting habits of medical staff, etc.

[0073] In some embodiments, the sizes of multiple data blocks can be equal to each other. For example Figure 6 as shown in, when dividing the 426*1024 K-space data according to the first preset parameter 32*32, since 426 is not an integer multiple of 32, one side length of the data block corresponding to the last row obtained after slicing the K-space data will not reach 32. In this case, the slicing line of the last row of the K-space data can be moved up, such as moving to the position of the white straight line in the figure, and based on this, multiple data blocks with equal sizes are obtained. In some embodiments, at least two of the multiple data blocks can include overlapping content. For example, for the part corresponding to the spark bright spot in the K-space data, it can be divided into multiple data blocks that include all or part of the bright spot.

[0074] In some embodiments, preprocessing can be performed on the data block. For example, the preprocessing can include normalization processing, etc. For example, the multiple data blocks obtained by division can be respectively subjected to normalization processing to convert each data block into an image with pixel values between 0 and 1.

[0075] In some embodiments, the first classification model may include a Convolutional Neural Network (CNN). In some embodiments, the input to the first classification model may be a data block, and the output may be the probability that the data block contains outliers or whether the data block contains outliers. For example, multiple data blocks can be input simultaneously or separately into a trained first classification model, and the first classification model outputs the probability that each data block contains an outlier, such as a value between 0 and 1. Alternatively, multiple data blocks can be input simultaneously or separately into a trained first classification model, and the first classification model outputs whether each data block contains an outlier, such as 0 indicating no outlier and 1 indicating an outlier. In some embodiments, when the first classification model outputs a probability value containing an outlier, whether the data block contains an outlier can be determined based on a preset threshold. For example, the preset threshold can be set to 0.5; if the probability value of the data block is greater than 0.5, it is considered to contain an outlier; if it is less than 0.5, it is considered not to contain an outlier. In some embodiments, the preset threshold can be determined automatically or manually. For example, medical personnel can manually adjust the preset threshold based on the quality of the clinical entity image. Alternatively, the system can automatically determine the value of the preset threshold based on historical preset threshold values ​​or user feedback on the recognition results.

[0076] In some embodiments, for data blocks in the K-space center region, the presence or absence of outliers can be determined through methods such as individual identification (e.g., inputting data blocks in the K-space center region into another classification model for outlier identification), default settings (e.g., defaulting to whether data blocks in the K-space center region contain outliers), or mixed judgment (e.g., mixing data blocks in the K-space center region with other data blocks during training of the first classification model). For example, for Figure 6 The six data blocks within the central white rectangle can be left unincluded by default, or their inclusion of outliers can be determined through other classification models, threshold adjustments, or other means.

[0077] In some embodiments, the first classification model can be trained based on a training set. As an example only, the locations of ignition artifacts in K-space data from a large amount of clinical data can be statistically analyzed. Based on this, K-space data containing ignition artifacts can be obtained through simulation (e.g., setting ignition bright spots in simulated normal K-space data). The K-space data is then divided into multiple sample data blocks according to a first preset parameter, labeled with whether the sample data block contains anomalies (such as ignition bright spots). The sample data blocks and labels are input into an initial convolutional neural network for training to obtain a trained first classification model. In some embodiments, the first classification model can be trained in any reasonable manner, and this specification does not impose any limitations on this.

[0078] In some embodiments, a separate classification model can be trained on data blocks in the central region of K-space. In some embodiments, for data blocks in the central region of K-space, the data blocks in the central region can be trained together with other data blocks during the training of the first classification model. For example, for sample data blocks in the central region of K-space, the corresponding probability threshold can be adjusted, such as being higher than the probability threshold of sample data blocks at other locations, to avoid incorrectly identifying bright spots or other abnormalities in the central region data blocks as outliers, thereby improving the accuracy of the judgment results. Alternatively, sample data blocks in the central region of K-space can be labeled during the training of the first classification model to improve the model's ability to recognize such data blocks.

[0079] In some embodiments, a reconstructed image corresponding to the K-space data can be obtained; it can be determined whether the reconstructed image contains artifacts; and based on the determination result of the reconstructed image, it can be determined whether the K-space data contains outliers.

[0080] In some embodiments, the reconstructed image corresponding to the K-space data can be obtained through inverse Fourier transform. For example, if the K-space is fully sampled, the K-space data can be directly reconstructed into the image domain through inverse Fourier transform to obtain the corresponding reconstructed image; if the K-space is downsampled, the downsampled portion can be filled with 0s first, such as... Figure 4 The white rectangle represents downsampling based on partial Fourier transform technology. First, the downsampled data can be filled in using conjugate symmetric data in K space. After filling, the remaining unsampled parts are filled with 0. Then, the filled K space data is reconstructed into the image domain through inverse Fourier transform to obtain the corresponding reconstructed image.

[0081] In some embodiments, at least one image patch can be extracted from the reconstructed image according to a second preset parameter; based on the at least one image patch, a second classification model is used to determine whether the reconstructed image contains artifacts.

[0082] In some embodiments, the second preset parameter may reflect parameters such as the shape and size of the image block. In some embodiments, the second preset parameter may be the same as or different from the first preset parameter. In some embodiments, the second preset parameter may be any reasonable value, such as 224*224. In some embodiments, the second preset parameter may be determined based on image domain parameters, standard image size, etc. In some embodiments, the second preset parameter may be determined manually or automatically. For example, the user may manually set the second preset parameter.

[0083] In some embodiments, the second classification model may include a convolutional neural network. In some embodiments, the input to the second classification model may be an image patch, and the output may be whether the image patch contains artifacts. For example, a cropped image patch can be input into a trained second classification model, and the second classification model can output whether the image patch contains artifacts, such as 0 indicating no artifacts and 1 indicating artifacts.

[0084] In some embodiments, the second classification model can be trained based on a training set. In some embodiments, the label trained on the second classification model can be whether or not artifacts are present in the sample image patch, such as... Figure 5 As shown, label 0 indicates no artifacts, and label 1 indicates artifacts. The training process of the second classification model is similar to that of the first classification model; for more details, please refer to the relevant description of the first classification model, which will not be repeated here.

[0085] In some embodiments, outliers in the K-space data can be determined to exist when artifacts are present in the reconstructed image; and outliers in the K-space data can be determined not to exist when artifacts are not present in the reconstructed image. In some embodiments, a preset threshold corresponding to the K-space data can be adjusted based on the determination result of the reconstructed image. For example, when artifacts are not present in the reconstructed image, the preset threshold corresponding to outlier determination in the data block can be adaptively increased, such as increasing the preset threshold to 0.9. In some embodiments, the first classification model can be adjusted based on the determination result of the reconstructed image, such as adjusting the threshold for determining whether a point is an outlier, or adjusting the label, etc.

[0086] In some embodiments, if it is determined from the reconstructed image that no outliers are present in the K-space data, then step 330 is not performed.

[0087] Step 330: Correct the imaging data based on the data block containing the outliers. In some embodiments, step 330 may be performed by the imaging device 110, the processing device 120, or the correction module 230.

[0088] In some embodiments, artifact correction can be performed on the imaging data based on the data block containing outliers determined in step 320. In some embodiments, the location information of the outliers can be determined, and artifact correction can be performed on the data block containing the outliers based on the location information to obtain a corrected data block; the imaging data is then updated based on the corrected data block.

[0089] In some embodiments, a reference location can be determined in the imaging data for the data block containing the outlier. This is merely an example, such as... Figure 6As shown, by determining whether each divided data block contains outliers, data blocks P1, P2, P3, P4, and P5 containing outliers can be selected. The positions of these outlier-containing data blocks P1, P2, P3, P4, and P5 in K-space can be determined based on the position of each input data block in the K-space data. For example, the coordinates of P5 are (256, 64). In some embodiments, the approximate position of the outlier in the imaging data (such as K-space data) can be determined based on a reference position. For instance, the coordinates of the data block containing the outlier in K-space can be used as the coordinates of the outlier in K-space.

[0090] In some embodiments, a semantic segmentation neural network can be used to determine the location information of outliers in a data block. For example, the semantic segmentation neural network may include U-Net, SeeNet, Deconv-NET, etc. In some embodiments, a semantic segmentation neural network can be used to segment outliers in a data block containing outliers. A semantic segmentation neural network can also classify pixels in an image. (This is just an example.) Figure 7 As shown, a data block P5 containing outliers in K-space data can be input into a semantic segmentation neural network. The semantic segmentation neural network performs Spark point segmentation on the data block P5, generating a 0 / 1 mask, where 0 represents points without Spark and 1 represents points with Spark. The coordinates of outlier 1 in the data block are also shown.

[0091] In some embodiments, the semantic segmentation neural network can be trained using sample data blocks. For example, the labels corresponding to the semantic segmentation neural network can be the locations of outliers in the sample data blocks, with a size of 0 / 1 Mask equal to the input.

[0092] In some embodiments, a semantic segmentation neural network can be used to determine the location information of anomalies in the imaging data based on a reference location. In some embodiments, the location information of the anomalies can reflect the precise coordinates of the anomalies in the imaging data (such as K-space data). This is merely an example. Figure 6 and 7 As shown, the semantic segmentation neural network can first use the top-left corner of the data block P5 containing outliers as the anchor point. Then, the approximate location of the outlier in the entire imaging data (such as K-space data) is determined by the coordinates of the top-left anchor point of P5. a ,ro a ) Decision, such as (pe a ,ro a= (256, 64); then obtain the position coordinates (row, col) of the segmented outlier in data block P5: when the anchor point coordinates are defined as (0, 0), the position of the segmented outlier in data block P5 can be determined as [14, 14], [14, 15], [14, 16], [14, 18], [14, 19]; finally, based on the position coordinates of the anchor point in the K-space data and the position coordinates of the outlier in the data block, calculate the position coordinates of the outlier in K-space as (pe). a +row,ro a +col), for example, the outlier in the 15th row and 15th column of data block P5 has the coordinates [14,14] in data block P5, then the coordinates of the outlier in the entire K space are (256+14,64+14)=(270,78).

[0093] In some embodiments, artifact correction can be performed on data blocks containing outliers based on the location information of the outliers to obtain corrected data blocks. In some embodiments, artifact correction can be performed on data blocks containing outliers based on a reference position to obtain corrected data blocks. In some embodiments, artifact correction can be performed on data blocks containing outliers using at least one of the following methods: setting outliers to uncollected values, padding, interpolation, or neural network assignment to obtain corrected data blocks. For example, the values ​​corresponding to outliers in the data block can be filled with 0 based on the location information of the outliers to obtain corrected data blocks.

[0094] In some embodiments, a trained artifact correction model can be used to determine the location information of outliers and perform artifact correction on data blocks containing outliers based on the location information to obtain corrected data blocks. That is, the trained artifact correction model simultaneously locates outliers and corrects artifacts on data blocks containing outliers. In some embodiments, the input to the artifact correction model can be a data block containing outliers, and the output can include corrected data blocks. In some embodiments, the input to the artifact correction model can be a data block containing outliers, and the output can include corrected imaging data. In some embodiments, the trained artifact correction model can include a neural network consisting of at least an encoder layer and a decoder layer, such as the U-Net network. In some embodiments, the artifact correction model can be trained in any feasible manner, for example, by training on multiple sample pairs consisting of data blocks containing outliers and data blocks not containing outliers; this specification does not limit this.

[0095] In some embodiments, imaging data can be updated based on corrected data blocks to obtain corrected imaging data, for example, by filling the corrected data blocks into the corresponding positions in the K-space data. In some embodiments, artifact correction can be performed on the imaging data (such as K-space data) based on the location information of outliers to obtain corrected imaging data. For example, based on the location information of outliers, outliers in the K-space data can be filled with 0 to obtain corrected imaging data. In some embodiments, outliers in the corrected imaging data, such as arcing artifacts, are suppressed or eliminated. In some embodiments, magnetic resonance images can be generated based on the corrected imaging data. For example, based on the corrected K-space data, clinical entity images without arcing artifacts can be output through image reconstruction and / or post-processing processes for clinical diagnostic use.

[0096] It should be noted that the above description of process 300 is provided for illustrative purposes only and is not intended to limit the scope of this specification. Various changes and modifications can be made based on the description in this specification by those skilled in the art. However, these changes and modifications do not depart from the scope of this specification. In some embodiments, process 300 may include one or more additional operations, or one or more of the aforementioned operations may be omitted. For example, process 300 may include one or more additional operations for MRI imaging.

[0097] Figures 8-11 This is a schematic flowchart of an exemplary magnetic resonance imaging method according to other embodiments of this specification.

[0098] like Figure 8 As shown, in some embodiments, K-space data can be input into a magnetic resonance imaging system (such as magnetic resonance imaging system 200). The detection module (such as detection module 220) in the magnetic resonance imaging system divides the K-space data into multiple data blocks and inputs these data blocks individually or simultaneously into a first classification model to determine whether the input data blocks contain outliers and to determine the reference position of the data blocks containing outliers in the K-space data. Further, after the magnetic resonance imaging system obtains the data blocks containing outliers based on the output of the first classification model, the correction module (such as correction module 230) can input them into a semantic segmentation network. Based on the location information of the outliers in the K-space data obtained from the segmentation results, artifact correction is performed on the data blocks containing outliers based on the location information to obtain corrected data blocks. The K-space data is then updated based on the corrected data blocks, and finally, the corrected K-space data is output.

[0099] like Figure 9As shown, in some embodiments, K-space data can be input into a magnetic resonance imaging system (such as magnetic resonance imaging system 200). The detection module (such as detection module 220) in the magnetic resonance imaging system divides the K-space data into multiple data blocks and inputs these data blocks individually or simultaneously into a first classification model to determine whether the input data blocks contain outliers and to determine the reference position of the data blocks containing outliers in K-space. Further, after the magnetic resonance imaging system obtains the data blocks containing outliers based on the output of the first classification model, the correction module (such as correction module 230) can input them into a trained artifact correction model. Based on the artifact correction model, the position information of the outliers in the K-space data is determined, and artifact correction is performed on the data blocks containing outliers based on the position information. The corrected data blocks are then output. The correction module updates the K-space data based on the corrected data blocks and finally outputs the corrected K-space data.

[0100] like Figure 10 As shown, in some embodiments, K-space data and corresponding reconstructed images can be simultaneously input into a magnetic resonance imaging system (such as magnetic resonance imaging system 200). In the magnetic resonance imaging system, a detection module (such as the first detection unit 223) divides the K-space data into multiple data blocks and inputs these blocks individually or simultaneously into a first classification model to determine whether the input data blocks contain outliers and to determine the reference position of the data blocks containing outliers in the K-space data. Simultaneously, a detection module (such as the second detection unit 225) extracts at least one image block from the reconstructed image and inputs it into a second classification model to determine whether the image block contains artifacts. Further, the detection module determines whether the K-space data contains outliers based on the judgment result of the reconstructed image. When it is determined that the K-space data contains outliers, the magnetic resonance imaging system acquires the data block containing the outliers. A correction module (such as correction module 230) inputs this data block into a semantic segmentation network. Based on the location information of the outliers in the K-space data obtained from the segmentation results, artifact correction is performed on the data block containing the outliers based on the location information to obtain a corrected data block. The K-space data is then updated based on the corrected data block, and finally, the corrected K-space data is output.

[0101] like Figure 11As shown, in some embodiments, K-space data and corresponding reconstructed images can be simultaneously input into a magnetic resonance imaging system (such as magnetic resonance imaging system 200). In the magnetic resonance imaging system, a detection module (such as the first detection unit 223) divides the K-space data into multiple data blocks and inputs these data blocks individually or simultaneously into a first classification model to determine whether the input data blocks contain outliers and to determine the reference position of the data blocks containing outliers in the K-space data. Simultaneously, a detection module (such as the second detection unit 225) extracts at least one image block from the reconstructed image and inputs the image block into a second classification model to determine whether the image block contains artifacts. Furthermore, the detection module can determine whether the K-space data contains outliers based on the judgment result of the reconstructed image. Once it is determined that there are outliers in the K-space data, the magnetic resonance imaging system acquires the data block containing the outliers. The correction module (such as correction module 230) can input the data block into the trained artifact correction model. The artifact correction model determines the location information of the outliers in the K-space data and performs artifact correction on the data block containing the outliers based on the location information, thereby obtaining the corrected data block. The K-space data is then updated based on the corrected data block, and finally the corrected K-space data is output.

[0102] The beneficial effects that the embodiments of this specification may bring include, but are not limited to: (1) dividing the imaging data into data blocks and judging whether the imaging data contains outliers based on multiple data blocks can reduce misjudgments and improve the accuracy of artifact judgment results; (2) performing artifact correction based on data blocks containing outliers can improve the efficiency and accuracy of artifact correction and save time costs; (3) combining the artifact judgment results of the reconstructed image to judge whether the K-space data contains outliers can filter out cases where the artifacts are not obvious and do not need to be corrected, reduce unnecessary correction operations, and improve the accuracy of outlier identification and artifact correction efficiency; (4) through Dividing K-space data into data blocks containing overlapping parts can avoid missing K-space data content, thereby improving the accuracy of identifying outliers in the data blocks; (5) By distinguishing different regions of the K-space where the data blocks are located and performing normalization processing on each data block separately, misjudgments can be reduced and the accuracy of artifact judgment results can be improved; (6) By determining the precise location of outliers (such as arcing points) in the imaging data (such as K-space data) and correcting the data blocks containing outliers, an MRI image without arcing artifacts can be generated based on the corrected imaging data, thereby improving diagnostic efficiency and the accuracy of diagnostic results. It should be noted that different embodiments may produce different beneficial effects. In different embodiments, the beneficial effects that may be produced can be any one or a combination of the above, or any other possible beneficial effects.

[0103] The basic concepts have been described above. Obviously, for those skilled in the art, the detailed disclosure above is merely illustrative and does not constitute a limitation of this specification. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and corrections to this specification. Such modifications, improvements, and corrections are suggested in this specification and therefore remain within the spirit and scope of the exemplary embodiments described herein.

[0104] Furthermore, this specification uses specific terms to describe embodiments thereof. For example, "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that references to "an embodiment," "one embodiment," or "an alternative embodiment" in different locations throughout this specification do not necessarily refer to the same embodiment. Moreover, certain features, structures, or characteristics in one or more embodiments of this specification can be appropriately combined.

[0105] Furthermore, unless expressly stated in the claims, the order of processing elements and sequences, the use of numbers and letters, or other names described in this specification are not intended to limit the order of the processes and methods described herein. Although various examples have been discussed in the foregoing disclosure of some embodiments of the invention that are currently considered useful, it should be understood that such details are for illustrative purposes only, and the appended claims are not limited to the disclosed embodiments; rather, the claims are intended to cover all modifications and equivalent combinations that conform to the spirit and scope of the embodiments described herein. For example, while the system components described above can be implemented using hardware devices, they can also be implemented solely using software solutions, such as installing the described system on existing servers or mobile devices.

[0106] Similarly, it should be noted that, in order to simplify the description disclosed herein and thus aid in the understanding of one or more embodiments of the invention, the foregoing description of embodiments in this specification may sometimes combine multiple features into a single embodiment, drawing, or description thereof. However, this method of disclosure does not imply that the subject matter of this specification requires more features than those mentioned in the claims. In fact, the embodiments contain fewer features than all the features of a single embodiment disclosed above.

[0107] In some embodiments, numbers describing the quantity of components and attributes are used. It should be understood that such numbers used in the description of embodiments are modified in some examples with the terms "approximately," "approximately," or "generally." Unless otherwise stated, "approximately," "approximately," or "generally" indicates that the numbers are allowed to vary by ±20%. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, which may be changed depending on the characteristics required by individual embodiments. In some embodiments, numerical parameters should take into account specified significant digits and employ a general method of digit reservation. Although the numerical ranges and parameters used to confirm their breadth of range in some embodiments of this specification are approximate values, in specific embodiments, such values ​​are set as precisely as feasible.

[0108] For each patent, patent application, patent application publication, and other material, such as articles, books, specifications, publications, and documents, referenced in this specification, the entire contents of which are incorporated herein by reference. This excludes historical application documents that are inconsistent with or conflict with the content of this specification, as well as documents that limit the broadest scope of the claims in this specification (currently or subsequently appended to this specification). It should be noted that in the event of any inconsistency or conflict between the descriptions, definitions, and / or terminology used in the supplementary materials to this specification and the content of this specification, the descriptions, definitions, and / or terminology used in this specification shall prevail.

[0109] Finally, it should be understood that the embodiments described in this specification are merely illustrative of the principles of the embodiments described herein. Other variations may also fall within the scope of this specification. Therefore, alternative configurations of the embodiments described herein are intended to be illustrative rather than limiting, and should be considered consistent with the teachings of this specification. Accordingly, the embodiments described herein are not limited to those explicitly introduced and described herein.

Claims

1. A magnetic resonance imaging method, characterized by, The method comprises: acquiring imaging data, the imaging data being K-space data; dividing the imaging data into a plurality of data blocks according to a first preset parameter, the first preset parameter being a parameter reflecting the size and shape of the data blocks; inputting the plurality of data blocks into a first classification model to obtain data blocks containing abnormal points, the output of the first classification model including the probability of each data block containing abnormal points or whether the data block contains abnormal points; correcting the imaging data based on the data blocks containing abnormal points; wherein the inputting the plurality of data blocks into the first classification model to obtain data blocks containing abnormal points comprises: for data blocks in a K-space center region among the plurality of data blocks, inputting the data blocks into a third classification model for abnormal point identification; for data blocks not in the K-space center region among the plurality of data blocks, inputting the data blocks into the first classification model for abnormal point identification, wherein the third classification model is different from the first classification model.

2. The method of claim 1, wherein, The first classification model is obtained by mixed training of sample data blocks corresponding to a K-space center region and sample data blocks corresponding to other regions of the K-space, and in the training process of the first classification model, the probability threshold of the sample data blocks corresponding to the K-space center region is greater than the probability threshold of the sample data blocks corresponding to other regions of the K-space.

3. The method of claim 1, wherein, The inputting the plurality of data blocks into the first classification model to obtain data blocks containing abnormal points comprises: acquiring a reconstructed image corresponding to the K-space data; determining whether the reconstructed image contains artifacts; in response to determining that the reconstructed image contains artifacts, adjusting a threshold value corresponding to abnormal point determination in the first classification model; inputting the plurality of data blocks into the adjusted first classification model to obtain data blocks containing abnormal points.

4. The method of claim 3, wherein, The determining whether the reconstructed image contains artifacts comprises: according to a second preset parameter, extracting at least one image block from the reconstructed image; based on the at least one image block, determining whether the reconstructed image contains artifacts by using a second classification model.

5. The method of claim 1, wherein, The correcting the imaging data based on the data blocks containing abnormal points comprises: determining position information of the abnormal points; based on the position information, performing artifact correction on the data blocks containing abnormal points to obtain corrected data blocks; updating the imaging data based on the corrected data blocks.

6. The method of claim 5, wherein, The determining position information of the abnormal points comprises: determining a reference position of the data blocks containing abnormal points in the imaging data; based on the reference position, determining position information of the abnormal points in the imaging data by using a semantic segmentation neural network.

7. The method of claim 5, wherein, The correcting the imaging data based on the data blocks containing abnormal points comprises: by using a trained artifact correction model, determining position information of the abnormal points; and based on the position information, performing artifact correction on the data blocks containing abnormal points to obtain corrected data blocks.

8. A magnetic resonance imaging system, characterized by The method comprises: an acquisition module configured to acquire imaging data, the imaging data being K-space data; a detection module configured to According to a first preset parameter, the imaging data is divided into a plurality of data blocks, the first preset parameter being a parameter reflecting the size and shape of the data blocks; The plurality of data blocks are input into a first classification model to obtain data blocks containing abnormal points, the output of the first classification model including the probability of each data block containing abnormal points or whether it contains abnormal points; wherein the input of the plurality of data blocks into the first classification model to obtain data blocks containing abnormal points comprises: For data blocks in the K-space center region among the plurality of data blocks, the data blocks are input into a third classification model for abnormal point identification; For data blocks not in the K-space center region among the plurality of data blocks, the data blocks are input into a first classification model for abnormal point identification, wherein the third classification model is different from the first classification model; A correction module is configured to correct the imaging data based on the data blocks containing abnormal points.

9. A computer-readable storage medium, the storage medium storing computer instructions, when the computer reads the computer instructions in the storage medium, the computer executes the method of any one of claims 1-7.

Images