Artifact elimination method, system, terminal and storage medium

Abstract

The present invention discloses an artifact elimination method, system, terminal and storage medium, wherein the artifact elimination method is applied to a magnetic resonance imaging device having at least two receivers, and is characterized in that it comprises: obtaining at least two magnetic resonance signals from different receivers, wherein the magnetic resonance signal comprises a magnetic resonance imaging signal and an electromagnetic interference signal; obtaining weight information of the electromagnetic interference signal according to the at least two magnetic resonance signals; optimizing the magnetic resonance signal according to the weight information of the electromagnetic interference signal and a deep learning algorithm to obtain the magnetic resonance imaging signal; generating magnetic resonance imaging according to the magnetic resonance imaging signal, being able to separate the electromagnetic interference signal and the magnetic resonance imaging signal, reducing magnetic resonance imaging artifacts, and improving the accuracy of magnetic resonance imaging.

Description

Technical Field

[0001] The present invention relates to the technical field of magnetic resonance imaging, and in particular to an artifact elimination method, system, terminal and storage medium. Background Art

[0002] With the development of medical technology, magnetic resonance imaging (MRI) technology is becoming more and more important. MRI is a routine procedure for diagnosing and treating various diseases and injuries. It can accurately reflect the pathological tissue conditions, allowing doctors to make accurate diagnoses and provide targeted treatment for patients based on the results of MRI.

[0003] Currently, imaging signals obtained through magnetic resonance imaging are affected by electromagnetic interference (EMI) and noise, which may cause magnetic resonance imaging artifacts and reduce the accuracy of magnetic resonance imaging. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention proposes an artifact elimination method that can separate electromagnetic interference signals and magnetic resonance imaging signals, reduce magnetic resonance imaging artifacts, and improve the accuracy of magnetic resonance imaging.

[0005] According to an embodiment of the first aspect of the present invention, there is provided an artifact elimination method, which is applied to a magnetic resonance imaging device having at least two receivers, and is characterized in that it includes: acquiring at least two magnetic resonance signals from different receivers, the magnetic resonance signals including magnetic resonance imaging signals and electromagnetic interference signals; acquiring weight information of the electromagnetic interference signal based on the at least two magnetic resonance signals; optimizing the magnetic resonance signal based on the weight information of the electromagnetic interference signal and a deep learning algorithm to acquire the magnetic resonance imaging signal; and generating magnetic resonance imaging based on the magnetic resonance imaging signal.

[0006] In some embodiments, acquiring the weight information of the electromagnetic interference signal according to the at least two magnetic resonance signals comprises: applying a blind source separation algorithm to separate the magnetic resonance imaging signal and the electromagnetic interference signal.

[0007] In some embodiments, the blind source separation algorithm includes: acquiring magnetic resonance signals from at least two of the receivers; generating a mixing matrix according to the magnetic resonance signals; and parsing the mixing matrix to acquire the magnetic resonance imaging signal and the electromagnetic interference signal.

[0008] In some embodiments, the deep learning algorithm includes applying a neural network for deep learning; the architecture of the neural network includes: a convolutional neural network model, a generative adversarial network model, a recurrent neural network model, a long short-term memory network model, an autoencoder network model, a deep belief network model, a deep residual network model, a gated recurrent unit network model or an echo state network model. One or more.

[0009] In some embodiments, the artifact removal method further comprises: optimizing the magnetic resonance imaging signal according to a spatiotemporal multi-dimensional coupled linear matrix model.

[0010] In some embodiments, the spatiotemporal multidimensional coupled linear matrix model includes: grouping the receivers; setting the receivers in different groups to have different signal sensitivities; acquiring the magnetic resonance signal; acquiring parameter information of the magnetic resonance imaging signal and the electromagnetic interference signal based on the magnetic resonance signal and the signal sensitivity of the receivers; and optimizing the magnetic resonance imaging signal based on the parameter information of the magnetic resonance imaging signal and the electromagnetic interference signal.

[0011] According to an embodiment of the second aspect of the present invention, a magnetic resonance imaging system is provided, characterized in that it includes: a signal acquisition module, used to acquire at least two magnetic resonance signals from different receivers, the magnetic resonance signals including magnetic resonance imaging signals and electromagnetic interference signals; a weight calculation module, used to acquire weight information of the electromagnetic interference signal based on the at least two magnetic resonance signals; an imaging optimization module, used to optimize the magnetic resonance signal based on the weight information of the electromagnetic interference signal and a deep learning algorithm to acquire the magnetic resonance imaging signal; and an imaging module, used to generate magnetic resonance imaging based on the magnetic resonance imaging signal.

[0012] In some embodiments, the magnetic resonance imaging system further includes: a signal receiving module, wherein the signal receiving module is used to receive radio frequency signals with different polarization directions and sensitivities.

[0013] According to an embodiment of the third aspect of the present invention, there is provided a magnetic resonance imaging terminal, characterized in that it comprises at least two receivers, a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, it implements: the artifact elimination method as described in the first aspect.

[0014] According to an embodiment of the fourth aspect of the present invention, a computer-readable storage medium is provided for computer-readable storage, wherein the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the artifact elimination method described in the first aspect.

[0015] The embodiment of the present application is based on a magnetic resonance imaging device with at least two receivers, which can separate electromagnetic interference signals and magnetic resonance imaging signals, perform imaging according to the separated magnetic resonance imaging signals, reduce magnetic resonance imaging artifacts, and improve the accuracy of magnetic resonance imaging.

[0016] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0018] Figure 1 is a flow chart of a method for removing artifacts according to an embodiment of the present invention;

[0019] Figure 2 is a flow chart of a method for removing artifacts according to another embodiment of the present invention;

[0020] Figure 3 is a flow chart of a method for removing artifacts according to another embodiment of the present invention;

[0021] Figure 4 is the image space image;

[0022] Figure 5 Based on Figure 4 The corresponding k-space image;

[0023] Figure 6 An image obtained by a method for removing artifacts according to another embodiment of the present invention;

[0024] Figure 7 This is an image obtained by a method for removing artifacts according to another embodiment of the present invention. DETAILED DESCRIPTION

[0025] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0026] In the description of the present invention, it should be understood that descriptions involving orientations, such as up, down, front, back, left, right, etc., and orientations or positional relationships indicated are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0027] In the description of the present invention, "several" means one or more, "more" means more than two, "greater than", "less than", "exceed" etc. are understood as not including the number itself, and "above", "below", "within" etc. are understood as including the number itself. If there is a description of "first" or "second", it is only used for the purpose of distinguishing the technical features, and cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0028] In the description of the present invention, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above terms in the present invention in combination with the specific content of the technical solution.

[0029] Figure 1 FIG. 1 is a flow chart of a method for removing artifacts provided by an embodiment of the present invention. Figure 1 The artifact elimination method shown includes at least the following steps: S110: acquiring at least two magnetic resonance signals from different receivers; S120: acquiring weight information of an electromagnetic interference signal based on at least two magnetic resonance signals; S130: optimizing the magnetic resonance signal based on the weight information of the electromagnetic interference signal and a deep learning algorithm to acquire a magnetic resonance imaging signal; S140 generating a magnetic resonance imaging based on the magnetic resonance imaging signal.

[0030] According to an embodiment of the first aspect of the present invention, there is provided an artifact elimination method, which is applied to a magnetic resonance imaging device having at least two receivers, and is characterized in that it includes: acquiring at least two magnetic resonance signals from different receivers, the magnetic resonance signals including magnetic resonance imaging signals and electromagnetic interference signals; acquiring weight information of the electromagnetic interference signal based on the at least two magnetic resonance signals; optimizing the magnetic resonance signal based on the weight information of the electromagnetic interference signal and a deep learning algorithm to acquire a magnetic resonance imaging signal; and generating magnetic resonance imaging based on the magnetic resonance imaging signal.

[0031] In some embodiments, acquiring weight information of an electromagnetic interference signal according to at least two magnetic resonance signals includes: applying a blind source separation algorithm to separate the magnetic resonance imaging signal and the electromagnetic interference signal.

[0032] In some embodiments, the sources of electromagnetic interference signals EMI include internal and external sources. Among them, the sources of electromagnetic interference signals EMI may include power lines or electrical gradients. Electromagnetic interference signals EMI are independent of the encoding gradient of magnetic resonance imaging MRI. Electromagnetic interference signals EMI may also be related to the sequence and parameters used to collect data. That is, the parameters involved in electromagnetic interference signals EMI may include the main magnetic field or bandwidth, and may be sent after a long scan.

[0033] Figure 2FIG. 1 is a flowchart of a method for removing artifacts provided by another embodiment of the present invention. Figure 2 The artifact elimination method shown includes at least the following steps: S210: acquiring magnetic resonance signals from at least two receivers; S220: generating a mixing matrix according to the magnetic resonance signals; S230: analyzing the mixing matrix to acquire magnetic resonance imaging signals and electromagnetic interference signals.

[0034] In some embodiments, the MR signals may be spatially encoded by the magnetic gradient coils, ie, the received MR signals are modulated by coil sensitivity maps.

[0035] In some embodiments, the MRI imaging is affected by the position, orientation and physical characteristics of the radio frequency coil. Electromagnetic interference signals EMI from external sources, i.e., far-field sources, can be provided near the subject with coils in different orientations and multiple receivers.

[0036] In some embodiments, multiple array coils can be used to detect the spatial distribution of near-field sources. The human body is a conductive structure, and the body parts outside the magnet will couple with the environment. Therefore, it is necessary to use multiple coils in different positions and directions to detect the distribution of electromagnetic interference signals EMI, which can further improve the magnetic resonance MR scanning results.

[0037] In some embodiments, three directions, such as X, Y, and Z directions, are selected to set up multiple array coils. Under ultra-low magnetic fields of 0.001-0.1T, the wavelength of the magnetic resonance MR signal is long enough, and the wavelength of the electromagnetic interference signal EMI that can interact with the magnetic resonance imaging MRI acquisition is also long. Therefore, the number of sensors required to characterize the electromagnetic interference signal EMI can be very small. In medium and high magnetic fields, the MR signal wavelength is shorter, and more array coils are required. In high fields, the noise mainly comes from the thermal noise of the sample. In low or ultra-low fields, the noise mainly includes electrical noise generated by the coils and receiving chains, such as RF preamplification, analog-to-digital conversion, demodulation and other links.

[0038] In some embodiments, the artifact elimination method can achieve artifact elimination for ultra-low magnetic field, i.e., 0.001T to 0.1T, low field, i.e., 0.1T to 0.5T, medium field, i.e., 0.5T to 3T, and high field 3T and above magnetic resonance imaging MRI systems by applying radio frequency electronic devices with multiple receivers.

[0039] In some embodiments, the blind source separation algorithm includes: acquiring magnetic resonance signals from at least two receivers; generating a mixing matrix according to the magnetic resonance signals; and analyzing the mixing matrix to acquire a magnetic resonance imaging signal and an electromagnetic interference signal.

[0040] In some embodiments, electromagnetic interference signals EMI signals and magnetic resonance MR signals are separated by using blind signal separation (BSS). Magnetic resonance MR images are modulated by coil sensitivity maps (CSM) and then Fourier transformed into k-space. The modulation of the coil sensitivity map in image space is equivalent to convolution in k-space.

[0041] In some embodiments, it is assumed that the EMI signal is statistically invariant in the entire k-space. The MR signal and the EMI signal detected by the receiver can be linearly modeled as a blind source separation (BSS) problem to separate the MR signal and the EMI signal.

[0042] In some embodiments, when performing blind source separation, the number of measurements should be greater than the number of sources. It can be concluded that: m = Ws, where m is N measurements, s is M sources, and W is an N*M mixing matrix that linearly combines these sources.

[0043] In some embodiments, when N>M, the mixing matrix W and M sources s can be calculated by independent component analysis (ICA) using the measurement value m, where s contains at most M-1 sources related to the magnetic resonance MR signal and 1 source related to the electromagnetic interference signal EMI.

[0044] In some embodiments, the mixing matrix may be estimated by independent component analysis. The measured mixed signal is decomposed into additive subcomponents to obtain corresponding unmixing matrices and mixing matrices.

[0045] In some embodiments, blind source separation can be performed by a fixed-point algorithm, such as a fixed-point algorithm based on projection pursuit, an information maximization algorithm (infomax) based on entropy maximization, or a fast independent component analysis (fast ICA) algorithm.

[0046] In some embodiments, the mixing matrix may also be estimated by principal component analysis (PCA), which may be performed by eigenvalue decomposition of a data covariance matrix or singular value decomposition of a data matrix.

[0047] In some embodiments, the k-space of the magnetic resonance MR data acquired by a plurality of receiver elements may be reconstructed into a matrix or a tensor and decomposed using singular value decomposition.

[0048] In some embodiments, the Welch method can also be used to perform electromagnetic interference EMI and noise analysis. Due to the different power spectrum distribution characteristics of MR signals and EMI signals, the Welch method is used to perform non-parametric classical spectrum estimation. The signal is divided into partially overlapping data segments, and the periodogram is calculated by windowing respectively, and finally the average is taken to complete the analysis.

[0049] In some embodiments, an autoregressive moving average (ARMA) model can also be used for electromagnetic interference EMI and noise analysis. The application of the autoregressive moving average model can improve the problems of poor variance performance and low resolution of classical spectrum estimation. Assume that the source signal s is the output of an input sequence through a linear non-time-varying system H(z) with both zeros and poles. The parameters of H(z) are estimated by s or its autocorrelation function r, and the power spectrum is estimated by H(z). Specifically, the modified Yule-Walker method can be used. After the power spectral density of each source is obtained, it is classified. If the power spectral density of the source is higher than a preset threshold within a preset frequency band, the source is identified as electromagnetic interference EMI. Magnetic resonance MR signals can be identified by peak detection, that is, checking whether the peak corresponding to the center of k-space is higher than the threshold.

[0050] In some embodiments, the source may be identified in image space.EMI typically appears as bright noise bands in image space that overlap with background or object regions.

[0051] In some embodiments, coil sensitivity information may be integrated into an independent component analysis algorithm.

[0052] In some embodiments, the deep learning algorithm includes applying a neural network for deep learning; the architecture of the neural network includes: a convolutional neural network model, a generative adversarial network model, a recurrent neural network model, a long short-term memory network model, an autoencoder network model, a deep belief network model, a deep residual network model, a gated recurrent unit network model, or an echo state network model. One or more of the models.

[0053] In some embodiments, a convolutional neural network is used as an example. The input is complex image data, for example, complex image data of size 256×256×2×2, the length of the second-to-last dimension is 2, representing 2 receivers, and the length of the last dimension is 2, representing the real channel and the imaginary channel respectively. The output of the convolutional neural network model is 256×256×2×2 residual image data, corresponding to the real and imaginary parts of the residual image data of the 2 receivers. The convolutional neural network model includes five convolutional layers and activation functions, specifically, including 9×9 convolutional layers, ReLu activation layers, 7×7 convolutional layers, ReLu activation layers; 5×5 convolutional layers, ReLu activation layers, 5×5 convolutional layers, ReLu activation layers, 3×3 convolutional layers, and the number of channels of the output of each corresponding convolutional layer is 128, 64, 32, 32 and 4 respectively. When training the model, the ADAM or SGD optimizer can be used to minimize the loss function, and the loss function selects the mean squared error (MSE).

[0054] In some embodiments, the output of the neural network may be a pure magnetic resonance MR signal, which is then connected to another network responsible for reconstructing the magnetic resonance MR image, that is, the magnetic resonance MR signal is input and the magnetic resonance MR image is output.

[0055] In some embodiments, a neural network can be designed that does not perform an image reconstruction process, that is, a single neural network directly outputs a reconstructed image. In addition, the EMI elimination process can occur in a hybrid space between the original data k-space and the image space, such as k x -y space, xk y Space, k x -yz space and k x -k y -z, etc.

[0056] In some embodiments, the output of the neural network may be an electromagnetic interference signal EMI, and a pure magnetic resonance MR signal may be obtained by subtracting the electromagnetic interference signal EMI output from the contaminated magnetic resonance MR signal. This neural network training method can obtain training data as quickly as possible.

[0057] In some embodiments, pure magnetic resonance MR signals may be acquired for training by a magnetic resonance system with a good shielding setting.

[0058] In some embodiments, the electromagnetic interference signal EMI can be obtained by simulation. The electromagnetic interference signal EMI can be generated by randomizing the phase, center frequency and bandwidth of the electromagnetic interference signal EMI. Noise such as additive white Gaussian can also be added to simulate the real situation.

[0059] In some embodiments, the electromagnetic interference signal EMI can be obtained by turning off the NMR excitation radio frequency pulse or removing the magnetic resonance MR signal source from the radio frequency coil. The detected electromagnetic interference signal EMI can also be related to the sequence and parameters of the collected data. For different sequences and different parameter settings, it may be necessary to use matching parameters to obtain the electromagnetic interference signal EMI.

[0060] In some embodiments, electromagnetic interference signal EMI training data can be collected during the patient MRI scan. The electromagnetic interference signal EMI can be collected during each NMR excitation repetition cycle when the magnetic resonance MR signal is expected to be zero.

[0061] In some embodiments, the MRI data contaminated by the EMI signal can be obtained by adding the simulated or collected EMI signal or noise signal to the pure MR signal, so as to enable further machine learning.

[0062] In some embodiments, the position and direction of the array coils may be optimized specifically.

[0063] In some embodiments, a clean magnetic resonance MR signal may be obtained by simulation.Magnetic resonance imaging MRI data may be generated by modulating a complex image with the coil sensitivity of a map of array coils, and then converted to the frequency domain by Fourier transformation.

[0064] In some embodiments, the artifact removal method further comprises: optimizing the magnetic resonance imaging signal according to a spatiotemporal multi-dimensional coupled linear matrix model.

[0065] Figure 3 FIG. 1 is a flowchart of a method for removing artifacts provided by another embodiment of the present invention. Figure 3 The artifact elimination method shown includes at least the following steps: S310: grouping receivers; S320: setting receivers in different groups to have different signal sensitivities; S330: acquiring magnetic resonance signals; S340: acquiring parameter information of magnetic resonance imaging signals and electromagnetic interference signals according to the magnetic resonance signals and the signal sensitivities of the receivers;

[0066] S350: Optimizing the magnetic resonance imaging signal according to parameter information of the magnetic resonance imaging signal and the electromagnetic interference signal

[0067] In some embodiments, the spatiotemporal multidimensional coupled linear matrix model includes: grouping receivers; setting different groups of receivers to have different signal sensitivities; acquiring magnetic resonance signals; acquiring parameter information of magnetic resonance imaging signals and electromagnetic interference signals based on the magnetic resonance signals and the signal sensitivities of the receivers; and optimizing the magnetic resonance imaging signals based on the parameter information of the magnetic resonance imaging signals and the electromagnetic interference signals.

[0068] In some embodiments, training data that does not contain magnetic resonance MR signals and only contains electromagnetic interference signals EMI signals is first obtained. It can be obtained from a specific subject, a group of subjects, a phantom, or no subject or phantom, and can also be collected at the time when the magnetic resonance MR signal is expected to be zero within each NMR excitation repetition cycle during the magnetic resonance imaging MRI scan of a specific patient. Assume that there are C receiver elements in total, of which the 1st to Kth receiver elements are adjusted to have a certain sensitivity to both magnetic resonance MR and electromagnetic interference signals EMI signals, and the sensitivity of the K+1th to Cth receiver elements to the magnetic resonance MR signal is adjusted to be close to 0, and the sensitivity to the electromagnetic interference signal EMI is adjusted to the maximum, that is, only the electromagnetic interference signal EMI is received. For each receiver element i, the collected training data containing only the electromagnetic interference signal EMI is ni, where i = 1, 2, 3, ..., C. Using the data of the last CK receiver elements to represent the first K receivers, the following formula can be established:

[0069] [n K+1 ,n K+2 ,…,n C ][w K+1 ,w K+2 ,…,w C ] T =[n1,n2,…,n K ]

[0070] Simplifying the formula, we can get:

[0071] N EMI W=N mix

[0072] Among them, N EMI、 W, N mix The matrix sizes are Rx(CK), (CK)xK, and RxK, respectively, where R is the number of repeated measurements.

[0073] It can be concluded that: W = N EMI +N mix , where + represents the Moore–Penrose generalized inverse.

[0074] Assume that the data collected by the first K and last CK receivers during the actual scanning is X mix and X EM , the data YMR after the first K receiver elements eliminate the electromagnetic interference signal EMI signal can be expressed as:

[0075] Y MR =X mix -X EMI W

[0076] In some embodiments, Y MRThe data of the K receivers can be combined by using the Root Sum Of Squares (RSS) method, or combined by using the Singular Value Decomposition (SVD) method before image reconstruction.

[0077] According to an embodiment of the second aspect of the present invention, a magnetic resonance imaging system is provided, characterized in that it includes: a signal acquisition module, used to acquire at least two magnetic resonance signals from different receivers, the magnetic resonance signals including magnetic resonance imaging signals and electromagnetic interference signals; a weight calculation module, used to acquire weight information of the electromagnetic interference signal based on the at least two magnetic resonance signals; an imaging optimization module, used to optimize the magnetic resonance signal based on the weight information of the electromagnetic interference signal and a deep learning algorithm to acquire a magnetic resonance imaging signal; and an imaging module, used to generate magnetic resonance imaging based on the magnetic resonance imaging signal.

[0078] In some embodiments, the magnetic resonance imaging system further includes: a signal receiving module, the signal receiving module being used to receive radio frequency signals with different polarization directions and sensitivities.

[0079] In some embodiments, the RF coils may be designed or placed with different purposes. Some coils may be designed to minimize their sensitivity to magnetic resonance MR signals or to detect only electromagnetic interference signals EMI. This can be achieved by placing the coils away from the imaging area or in an orientation with minimal sensitivity to magnetic resonance MR signals. Some coils are designed to detect magnetic resonance MR signals while having minimal electromagnetic interference signal EMI contamination. This optimization can be used to maximize the efficiency of eliminating electromagnetic interference signals EMI, and at the same time, this can minimize thermal noise amplification in the process of electromagnetic interference signal EMI elimination.

[0080] Figure 4 is an image space image. Figure 5 Based on Figure 4 The corresponding k-space image. Figure 4 From left to right in the figure are pure magnetic resonance MR signals, electromagnetic interference signals EMI, noise signals, and spatial images of magnetic resonance MR signals contaminated by electromagnetic interference and noise. Figure 5 From left to right in the figure are the pure MR signal, the EMI signal, the noise signal, and the k-space image of the MR signal contaminated by EMI and noise. Figure 4 and Figure 5 As shown, when interfered by electromagnetic interference signal EMI and / or noise signal, both the image in image space and the image in k space will be affected.

[0081] Figure 6 is an image obtained by a method for removing artifacts according to another embodiment of the present invention. Figure 6 As shown in the images, the upper image is a magnetic resonance imaging MRI with electromagnetic interference signals EMI, and the lower image is an image obtained after the electromagnetic interference signals EMI are eliminated using the artifact elimination method. By comparison, it can be concluded that the image obtained after the electromagnetic interference signals EMI are eliminated using the artifact elimination method has a higher signal-to-noise ratio and a clearer image.

[0082] Figure 7 is an image obtained by a method for removing artifacts according to another embodiment of the present invention. Figure 7 As shown in the images, the upper image is a human brain magnetic resonance imaging MRI with electromagnetic interference signals EMI, and the lower image is a human brain image obtained after the electromagnetic interference signals EMI are eliminated using the artifact elimination method. By comparison, it can be concluded that the image obtained after the electromagnetic interference signals EMI are eliminated using the artifact elimination method is clearer and the noise caused by the electromagnetic interference signals EMI can be removed.

[0083] According to an embodiment of the third aspect of the present invention, there is provided a magnetic resonance imaging terminal, characterized in that it comprises at least two receivers, a memory, a processor and a computer program stored in the memory and executable on the processor, and when the processor executes the program, it implements: the artifact elimination method of the first aspect.

[0084] According to an embodiment of the fourth aspect of the present invention, a computer-readable storage medium is provided for computer-readable storage, wherein the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the artifact removal method of the first aspect.

[0085] The embodiments of the present invention are described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments. Various changes can be made within the knowledge scope of ordinary technicians in the technical field without departing from the purpose of the present invention.

[0086] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0087] It will be appreciated by those skilled in the art that all or some of the steps and systems in the disclosed method above may be implemented as software, firmware, hardware and appropriate combinations thereof. Some physical components or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor or a microprocessor, or may be implemented as hardware, or may be implemented as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or a non-transitory medium) and a communication medium (or a temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, disk storage or other magnetic storage devices, or any other medium that may be used to store desired information and may be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0088] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0089] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

Claims

1. An artifact removal method, applied to a magnetic resonance imaging device having at least two receivers, characterized in that: include: Grouping receivers; Setting different groups of receivers to have different signal sensitivities; acquiring magnetic resonance signals; Acquiring parameter information of the magnetic resonance imaging signal and the electromagnetic interference signal according to the magnetic resonance signal and the signal sensitivity of the receiver; Optimizing the magnetic resonance imaging signal according to parameter information of the magnetic resonance imaging signal and the electromagnetic interference signal; It also includes: optimizing the magnetic resonance imaging signal according to a spatiotemporal multi-dimensional coupled linear matrix model; The time-space multi-dimensional coupled linear matrix model includes: C receiver elements are grouped, wherein the 1st to Kth receiver elements are adjusted to have certain sensitivity to both the magnetic resonance signal and the electromagnetic interference signal, and the sensitivity of the K+1th to Cth receiver elements to the magnetic resonance signal is adjusted to be close to 0, and the sensitivity to the electromagnetic interference signal is adjusted to be maximum; For each receiver element i, the training data collected only contains electromagnetic interference signals is n i , where i = 1, 2, 3, ..., C, using the data of the last CK receiver elements to represent the first K receiver elements, the following formula is established: [n K+1 ,n K+2 ,...,n C ][w K+1 ,w K+2 ,...,w C ] T =[n1,n2,...,n K ]; Simplifying the formula we can get: N EMI W=N mix ; Among them, N EMI , W, N mix The matrix sizes are R×(CK), (CK)×K, and R×K respectively; R is the number of repeated measurements; It can be concluded that: W = N EMI +N mix , where + represents the Moore-Penrose generalized inverse; The data collected by the first K receiver elements during actual scanning is X mix , the data collected by the last CK receiver elements is X EMI , the data Y after the first K receiver elements eliminate the electromagnetic interference signal MR is represented as: Y MR =X mix -X EMI W。 2. A magnetic resonance imaging terminal, characterized in that: The invention comprises at least two receivers, a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the following is achieved: The method for removing artifacts as claimed in claim 1.

3. A computer-readable storage medium for computer-readable storage, wherein the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the artifact removal method according to claim 1.

Images