Magnetic resonance image processing method and reconstruction method

Abstract

The embodiment of the invention is suitable for the technical field of medical images, and provides a magnetic resonance image processing method and reconstruction method, and the method comprises the steps: obtaining a plurality of layers of first aliasing images of a target object, and obtaining a plurality of first single-layer images of the target object; performing translation and aliasing processing on the first single-layer image in sequence to obtain a second aliasing image; the aliasing modes of the first aliasing image and the second aliasing image are the same; extracting a channel compression matrix from the second aliasing image; and compressing the first aliasing image by using a channel compression matrix to obtain a first compressed aliasing image, and compressing the second aliasing image to obtain a second compressed aliasing image. By adopting the method, the consistency of the coil sensitivity of the first compression aliasing image and the coil sensitivity of the second compression aliasing image can be ensured.

Description

Technical Field

[0001] The present application belongs to the field of medical imaging technology, and in particular relates to a magnetic resonance image processing method and a reconstruction method. Background Art

[0002] In magnetic resonance diffusion imaging, simultaneous multi-slice echo planar imaging (EPI) is commonly used. Specifically, a coil and multi-slice simultaneous EPI sequence are used to acquire multi-slice aliased imaging data from the target object's k-space, while simultaneously acquiring single-slice reference imaging data from the target object's k-space. Furthermore, to reduce the amount of data required to process the multi-slice aliased imaging data, a coil compression algorithm is typically used to compress the multi-slice aliased imaging data. Subsequently, the multi-slice reference imaging data is de-aliased and reconstructed based on the single-slice reference imaging data to produce the magnetic resonance image.

[0003] During dealiasing and image reconstruction, it is necessary to ensure the consistency of coil sensitivity between single-layer reference imaging data and multi-layer reference imaging data to improve the imaging quality of magnetic resonance images. However, using coil compression algorithms to compress multi-layer aliased imaging data will not ensure that the coil sensitivity of the multi-layer aliased imaging data remains consistent with the coil sensitivity of the single-layer reference imaging data, resulting in poor magnetic resonance image reconstruction.

[0004] Based on this, in the prior art, it is impossible to reduce the amount of data to be processed for multi-layer aliased imaging data while ensuring the consistency of the coil sensitivity of the compressed aliased imaging data with the coil sensitivity of the single-layer reference imaging data. Summary of the Invention

[0005] The embodiments of the present application provide a magnetic resonance image processing method and a reconstruction method, which can solve the problem that the coil sensitivity of compressed multi-layer aliasing imaging data is inconsistent with the coil sensitivity of single-layer reference imaging data.

[0006] In a first aspect, an embodiment of the present application provides a method for processing a magnetic resonance image, the method comprising:

[0007] Acquire multiple layers of first aliased images of the target object, and acquire multiple first single-layer images of the target object;

[0008] performing translation and aliasing processing on the first single-layer image in sequence to obtain a second aliased image; the first aliased image and the second aliased image have the same aliasing pattern;

[0009] extracting a channel compression matrix from the second aliased image;

[0010] The first aliased image is compressed using a channel compression matrix to obtain a first compressed aliased image, and the second aliased image is compressed to obtain a second compressed aliased image.

[0011] In a second aspect, an embodiment of the present application provides a magnetic resonance image reconstruction method, the method comprising:

[0012] Acquire a first aliased image of the target object, where the first aliased image is reconstructed by scanning the target object using a multi-layer simultaneous excitation pulse sequence;

[0013] Acquire a second aliased image, where the second aliased image has a lower resolution than the first aliased image, and the first aliased image and the second aliased image have the same aliasing pattern;

[0014] extracting a channel compression matrix from the second aliased image;

[0015] compressing the first aliased image using a channel compression matrix to obtain a first compressed aliased image, and compressing the second aliased image to obtain a second compressed aliased image;

[0016] obtaining a multi-layer reference image of the target object according to the second compressed aliased image;

[0017] The first compressed aliased image is dealiased and reconstructed based on the multi-layer reference image to obtain a multi-layer magnetic resonance image of the target object.

[0018] In a third aspect, an embodiment of the present application provides a magnetic resonance image processing device, the device comprising:

[0019] a first acquisition module, configured to acquire a plurality of first aliased images of the target object and a plurality of first single-layer images of the target object;

[0020] a first processing module, configured to sequentially perform translation and aliasing processing on the first single-layer image to obtain a second aliased image; the first aliased image and the second aliased image have the same aliasing mode;

[0021] a first extraction module, configured to extract a channel compression matrix from the second aliased image;

[0022] The first compression module is configured to compress the first aliased image using a channel compression matrix to obtain a first compressed aliased image, and to compress the second aliased image to obtain a second compressed aliased image.

[0023] In a fourth aspect, an embodiment of the present application provides a magnetic resonance image reconstruction device, comprising:

[0024] A second acquisition module is used to acquire a first aliased image of the target object, where the first aliased image is reconstructed by scanning the target object using a multi-layer simultaneous excitation pulse sequence;

[0025] a third acquisition module, configured to acquire a second aliased image, wherein the resolution of the second aliased image is lower than that of the first aliased image, and the first aliased image and the second aliased image have the same aliasing mode;

[0026] a second extraction module, configured to extract a channel compression matrix from the second aliased image;

[0027] a second compression module, configured to compress the first aliased image using a channel compression matrix to obtain a first compressed aliased image, and to compress the second aliased image to obtain a second compressed aliased image;

[0028] A second processing module, configured to obtain a multi-layer reference image of the target object according to the second compressed aliased image;

[0029] The third processing module is configured to perform dealiasing and reconstruction on the first compressed aliased image based on the multi-layer reference image to obtain a multi-layer magnetic resonance image of the target object.

[0030] In a fifth aspect, an embodiment of the present application provides a terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method of the first aspect described above when executing the computer program.

[0031] In a sixth aspect, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the method of the first aspect and / or the second aspect mentioned above.

[0032] In a seventh aspect, an embodiment of the present application provides a computer program product, which, when executed on a terminal device, enables the terminal device to execute the method of the first aspect and / or the second aspect described above.

[0033] Compared to the prior art, the embodiments of the present application offer the following advantages: after acquiring multiple layers of first mixed images and multiple first single-layer images of a target object, the first single-layer images can be sequentially shifted and aliased based on the target object's image position in the first mixed images to obtain a second mixed image. In this case, because the first and second mixed images have the same aliasing pattern, the target object's image position in the first and second mixed images can be considered consistent, thereby maintaining consistency in the image data of the target object in the first and second mixed images. In other words, the first and second mixed images are consistent in the image domain. Subsequently, a channel compression matrix can be extracted from the second mixed image, and the first and second mixed images can be simultaneously compressed using the channel compression matrix to obtain a first compressed aliased image and a second compressed aliased image. Because the target object's image data in the first and second mixed images is consistent, the coil sensitivities of the first and second compressed aliased images should also be consistent after the first and second mixed images are compressed using the same channel compression matrix, thereby reducing the amount of data required for processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or descriptions of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0035] Figure 1 This is a schematic diagram of an application scenario of an EPI sequence in a magnetic resonance image processing method provided in one embodiment of the present application;

[0036] Figure 2 This is a schematic diagram of filling k-space data for single excitation and multiple excitations in a magnetic resonance image processing method provided by an embodiment of the present application;

[0037] Figure 3 This is a flowchart of an implementation method of a magnetic resonance image processing method provided in one embodiment of the present application;

[0038] Figure 4 This is a schematic diagram of an application scenario of a first aliased image and a first single-layer image in a magnetic resonance image processing method provided by an embodiment of the present application;

[0039] Figure 5 This is a schematic diagram of a process for processing a first aliased image and a first single-layer image in a magnetic resonance image processing method provided by an embodiment of the present application;

[0040] Figure 6 This is a flowchart of an implementation method of a magnetic resonance image reconstruction method provided in one embodiment of the present application;

[0041] Figure 7 This is a schematic structural diagram of a magnetic resonance image processing device provided by an embodiment of the present application;

[0042] Figure 8 This is a schematic structural diagram of a magnetic resonance image reconstruction device provided by one embodiment of the present application;

[0043] Figure 9 This is a structural diagram of a terminal device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0044] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0045] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0046] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0047] In magnetic resonance diffusion imaging technology, a simultaneous multi-layer excitation echo-planar imaging method is usually used. Specifically, the coil and the multi-layer simultaneous excitation EPI sequence are used to acquire multi-layer aliased imaging data of the target object's k-space, and the single-layer reference imaging data of the target object's k-space is also acquired simultaneously. The coil is used to achieve spatial encoding of the target object's different position information within the slice by spatial encoding along the phase encoding direction and the frequency encoding direction. In addition, in order to reduce the amount of data that needs to be processed for the multi-layer aliased imaging data, it is usually necessary to first use a coil compression algorithm to compress the multi-layer aliased imaging data. Then, based on the single-layer reference imaging data, the multi-layer reference imaging data is subjected to dealiasing and image reconstruction processing to obtain a magnetic resonance image (EPI image).

[0048] K-space data is obtained by switching the readout gradient field's polarity between positive and negative after a single RF pulse, generating a gradient echo (a single signal). For example, signals are collected from the object being treated via a coil. EPI technology can generate k-space data by continuously switching the polarity of the gradient field to obtain multiple signals.

[0049] A single shot is a radiofrequency pulse that fills all k-space data. After a single shot, all signals are collected by continuously switching the gradient field polarity to complete the acquisition of k-space data. Furthermore, a multiple shot is a single shot, but not all signals are collected. Therefore, the radiofrequency pulse excitation process must be repeated.

[0050] Furthermore, since the read gradient field is a signal generated by switching between positive and negative polarities, the directions in which the generated k-space data fills the k-space are also different.

[0051] Specifically, refer to Figure 1 Figure 2 , Figure 1 This is a schematic diagram of an application scenario of an EPI sequence in a magnetic resonance image processing method provided in an embodiment of the present application. Figure 2 This is a schematic diagram of filling k-space data for single excitation and multiple excitations in a magnetic resonance image processing method provided by an embodiment of the present application.

[0052] in, Figure 1 RF in the figure represents radio frequency pulses. Gx, Gy, and Gz represent the gradients applied by the coil in the x, y, and z directions, respectively. That is, the coil includes an x-axis coil for applying a gradient in the x direction, a y-axis coil for applying a gradient in the y direction, and a z-axis coil for applying a gradient in the z direction. Based on this, it can be considered that the above Figure 1 The trapezoid and triangle are the gradients applied by the coil in each direction respectively.

[0053] Figure 2 In the figures corresponding to single excitation and multiple excitation, the arrows indicate the filling direction of the k-space data. For a single excitation, because the readout gradient field is a signal generated by switching between positive and negative polarity phases, the filling direction of the k-space data between adjacent rows is different.

[0054] For multiple excitations, the filling directions of two adjacent rows of k-space data obtained by one excitation are also different. Specifically, taking 4 excitations as an example, for the line segment corresponding to the first excitation of the solid line, the filling directions of the two adjacent rows of k-space data corresponding to the first excitation are also different. Because the first excitation corresponding to the RF pulse cannot collect all the signals, there are subsequent second excitations, third excitations, and fourth excitations represented by the dotted lines. Figure 2 As can be seen from the arrows, the filling directions of the k-space data of two adjacent rows obtained during each excitation are different.

[0055] The generated k-space data can be Figure 2 As shown in the k-space data diagram in , each circle represents the k-space data corresponding to a signal.

[0056] It should be noted that after the k-space data is obtained, an inverse Fourier transform may be performed on the k-space data to reconstruct a first aliased image corresponding to the k-space data.

[0057] During dealiasing and image reconstruction, it is necessary to ensure the consistency of coil sensitivities between single-layer reference imaging data and multi-layer reference imaging data to improve the quality of MRI imaging. However, using a coil compression algorithm to compress multi-layer aliased imaging data (and also single-layer reference imaging data) will not ensure consistency between the coil sensitivities of the multi-layer aliased imaging data and the single-layer reference imaging data, resulting in poor MRI image reconstruction.

[0058] If the coil compression algorithm is not used to compress the multi-layer aliased imaging data, although the coil sensitivity of the multi-layer aliased imaging data is consistent with that of the single-layer reference imaging data, due to the larger data volume, a lot of time is required to perform the de-aliasing process on the multi-layer aliased imaging data.

[0059] That is, in the prior art, it is impossible to reduce the amount of data to be processed for multi-layer aliased imaging data while ensuring the consistency of the coil sensitivity of the compressed aliased imaging data with the coil sensitivity of the single-layer reference imaging data.

[0060] Based on this, in order to reduce the amount of data to be processed and ensure the consistency of the coil sensitivity of the compressed aliased imaging data with the coil sensitivity of the single-layer reference imaging data, an embodiment of the present application provides a magnetic resonance image processing method. This method can be applied to terminal devices such as tablet computers, laptop computers, and ultra-mobile personal computers (UMPCs). The embodiment of the present application does not impose any restrictions on the specific type of terminal device.

[0061] See also Figure 3 , Figure 3 The following is a flowchart illustrating an implementation of a magnetic resonance image processing method provided in an embodiment of the present application. The method includes the following steps:

[0062] S301 : Acquire multiple layers of first aliased images of a target object, and acquire multiple first single-layer images of the target object.

[0063] In one embodiment, the target object is a subject to be subjected to magnetic resonance imaging, and the imaging may be performed on a portion of the target object, for example, the head, neck, chest, shoulder, arm, heart, knee, or foot of the target object, without limitation.

[0064] In one embodiment, when acquiring the first aliased image and the first single-slice image, the target object may be excited by radiofrequency pulses corresponding to an EPI sequence. The EPI sequence includes, but is not limited to, a diffusion EPI sequence and an arterial spin labeling (ASL) EPI sequence.

[0065] It should be noted that different EPI sequences are typically used for different regions of the target object to be captured. This is to generate a first aliased image that maximizes the contrast ratio of the image, thereby achieving the best display effect for the captured region of the target object in the image. In this embodiment, the EPI sequence used to capture the target object can be pre-set in the terminal device or determined by the operator, and this is not limited to this.

[0066] The method of acquiring the multi-layer first aliased images of the target object can be described as follows:

[0067] In one embodiment, the terminal device can use a multi-layer simultaneous excitation pulse sequence to excite the target object multiple times to obtain k-space data of multiple excitations; the phase difference between two adjacent rows of data in the k-space data is a preset phase difference; the k-space data is reconstructed to obtain a first aliased image.

[0068] As an example, to eliminate or reduce the effects of phase gradients and frequency encoding on signal acquisition in a multi-slice simultaneous excitation pulse sequence, in this embodiment, the gradients applied by the coil in various directions can be pre-set. Consequently, when the target object is repeatedly excited using multi-slice simultaneous excitation RF pulses and gradients, less-affected k-space data can be obtained, allowing the generation of a first aliased image with a high signal-to-noise ratio and minimal distortion based on the k-space data.

[0069] In one embodiment, the terminal device may determine a ratio of a preset phase to the number of layers of the first aliased image as a preset phase difference, and then determine a gradient based on the preset phase difference.

[0070] Specifically, the preset phase can be set according to actual conditions and is not limited thereto. For example, the preset phase can be 2π. That is, when the number of layers is 2 (the first aliased image is formed by superimposing two single-layer images), the phase difference can be π. When the number of layers is 3, the phase difference can be 120°, and so on.

[0071] Based on the above description, it can be seen that if the phase difference between two adjacent rows of k-space data needs to be a preset phase difference, the gradient applied by the coil in each direction needs to be adjusted based on the preset phase difference. For example, when the number of layers is two, the terminal device can add the following in Gz: Figure 1 The gradient shown within the black rectangle corresponds one-to-one with the gradient applied in the Gy direction and varies regularly. Therefore, using the gradient corresponding to the above acquisition method to acquire k-space data can better utilize coil sensitivity information, resulting in different phases in the phase encoding direction for k-space data from different layers, which facilitates the subsequent dealiasing of the first aliased image from inter-layer aliasing.

[0072] The terminal device may be pre-configured with a plurality of correspondences between preset phase differences and preset gradients. In this case, after determining the preset phase difference based on the preset phase and the number of layers of the first aliased image, the gradients to be applied by the coil in various directions may be directly determined based on the aforementioned correspondences.

[0073] In another embodiment, the above gradient can also be set by the staff according to actual conditions, which is not limited to this.

[0074] The method of acquiring multiple first-layer images of the target object can be described as follows:

[0075] In one embodiment, the multi-layer simultaneous excitation pulse sequence may be a single-shot EPI imaging sequence or a multi-shot EPI imaging sequence, wherein both the single-shot EPI imaging sequence and the multi-shot EPI imaging sequence have been explained above and will not be further described.

[0076] It should be noted that, at this point, only one image layer (a single-layer image) is acquired each time the target object is acquired. Specifically, the radiofrequency pulses corresponding to the EPI sequence excite one layer of the target object, collecting signals from that layer to populate the k-space data. The k-space data is then inverse Fourier transformed to obtain the first single-layer image.

[0077] However, it should be noted that in order to perform de-aliasing on the first aliased image, the number of first single-layer images must be a multiple of the number of layers in the first aliased image. Specifically, when the number of layers in the first aliased image is M, the number of first single-layer images must be a multiple of M. For example, the number of first single-layer images is N. In this case, when the first single-layer images are combined to form an aliased image, N / M aliased images can be generated, and each aliased image is formed by aliasing M first single-layer images.

[0078] For example, refer to Figure 4 , Figure 4 FIG. 1 is a schematic diagram of an application scenario of a first aliased image and a first single-layer image in a magnetic resonance image processing method provided in an embodiment of the present application, wherein N and M are both 3 as an example. Figure 4 I1, I2 and I3 are the first single-layer images of the target object at different angles, and P1 is the first aliased image obtained by aliasing the three layers of images.

[0079] In another embodiment, the first single-layer image serves as reference data for the first aliased image and is used to perform de-aliasing on the first aliased image. Therefore, to ensure that the first single-layer image can effectively serve as reference data for the first aliased image and improve the accuracy of de-aliasing the first aliased image, the terminal device may capture the first single-layer image that meets one or more of the following preset conditions.

[0080] Specifically, the preconditions may be as follows:

[0081] 1. The distance between two adjacent data points in the k-space data corresponding to each first single-layer image is consistent with the distance between two adjacent data points in the k-space data corresponding to the first aliased image.

[0082] In one embodiment, the k-space data has been explained above and will not be further described. It should be noted that the two adjacent data points mentioned above include being adjacent in both the horizontal and vertical directions. That is, the k-space data corresponding to the first single-slice image and the k-space data corresponding to the first aliased image are spatially aligned.

[0083] It should be added that Figure 2 Each circle in the figure represents a piece of k-space data, corresponding to a single signal reading. That is, when generating the first single-slice image and the first aliased image, the terminal device reads the signal at the same frequency to ensure spatial alignment of the k-space data corresponding to the first single-slice image and the first aliased image. This improves the effectiveness of the first single-slice image as reference data for the first aliased image.

[0084] 2. The image resolution of each first single-layer image is smaller than the image resolution of the first aliased image.

[0085] Reference Figure 2 , Figure 2 The k-space data corresponding to W in can be considered as the a*b-dimensional k-space data corresponding to the first aliased image. Based on this a*b-dimensional k-space data, the terminal device can generate an a*b-dimensional first aliased image. That is, the image resolution of the first aliased image is a*b. Figure 2 The spatial data corresponding to w in can be considered the c*d k-space data corresponding to the first single-layer image. Based on this c*d k-space data, the terminal device can generate a c*d first single-layer image. That is, the image resolution of the first single-layer image is c*d.

[0086] It should be noted that using the low-resolution first single-layer image as reference data can reduce the amount of data that the terminal device needs to process and improve the efficiency of subsequent de-aliasing of the first aliased image, that is, save time.

[0087] 3. The k-space data corresponding to each first single-layer image is not downsampled.

[0088] By not downsampling the k-space data corresponding to the first single-layer image, the first single-layer image obtained after inverse Fourier transform processing based on the k-space data can retain complete image information of the target object. Furthermore, when the terminal device performs de-aliasing processing on the first aliased image based on the clearer first single-layer image, the accuracy of de-aliasing of the first aliased image can be improved.

[0089] S302 : performing translation and aliasing processing on the first single-layer image in sequence to obtain a second aliased image; the first aliased image and the second aliased image have the same aliasing mode.

[0090] The aliasing modes of the first aliasing image and the second aliasing image are the same, and the image position of the target object in the first aliasing image is consistent with the image position in the second aliasing image.

[0091] In one embodiment, the purpose of sequentially performing translation and aliasing processing on the first single-slice image is to ensure that the target object in the second aliased image remains consistent with the target object in the first aliased image in the image domain. That is, the image position of the target object in the first aliased image is consistent with the image position in the second aliased image. Furthermore, when the first and second aliased images, whose image domains are consistent, are subsequently compressed using the same channel compression algorithm, the consistency of the coil sensitivities of the compressed first and second aliased images can be ensured.

[0092] Specifically, refer to Figure 5 , Figure 5FIG. 1 is a schematic diagram of a processing process of a first aliased image and a first single-layer image in a magnetic resonance image processing method provided in an embodiment of the present application. Figure 5 , Figure 5 In the figure, S1, S2, S3, S4, and S5 represent the order of processing steps, and I1, I2, and I3 represent the first single-layer images of the target object at different layers. First, the first single-layer images I1, I2, and I3 are translated (S1) and aliased (S2) in sequence to obtain the second aliased image P2. Figure 5 It can be determined that after image translation and aliasing processing are performed on I1, I2, and I3, the second aliased image P2 generated therefrom is consistent with the first aliased image P1 in the image domain.

[0093] The manner of performing translation and aliasing on the plurality of first single-layer images will not be described in detail.

[0094] It should be noted that when performing overlay, the first single-layer image should be aliased based on the aliasing conditions of the various target objects in the first aliased image. For example, in the first aliased image, if layer I1 is the topmost layer, layer I2 is the middle layer, and layer I3 is the bottommost layer, then during aliasing, the first single-layer image I1 should be used as the topmost layer, the first single-layer image I2 as the middle layer, and the first single-layer image I3 as the bottommost layer to generate the second aliased image.

[0095] S303: Extract a channel compression matrix from the second aliased image.

[0096] In one embodiment, a pre-trained image compression model can be used to extract a channel compression matrix from the second aliased image. The image compression model learns the signal distribution of the image and removes redundant pixels to generate an optimal channel compression matrix for optimal compression of the second aliased image. In this embodiment, the training method and model structure of the compression model are not limited.

[0097] It should be noted that the terminal device can also extract a channel compression matrix from the first aliased image. However, because the second aliased image is directly formed by the aliasing of multiple first single-layer images, the signal-to-noise ratio of a single-layer first single-layer image is usually high, and the brightness of the first single-layer image is relatively uniform. Therefore, the second aliased image obtained by aliasing the first single-layer image with a high signal-to-noise ratio and relatively uniform brightness usually has a higher signal-to-noise ratio than the first aliased image, and the brightness is also usually relatively uniform. Based on this, when the channel compression matrix extracted from the second aliased image with a high signal-to-noise ratio and relatively uniform brightness is used to compress the first aliased image, it can reduce the image data to be processed in the first aliased image while accurately removing redundant noise pixels in the first aliased image, thereby ensuring the image quality of the compressed first aliased image.

[0098] S304 : Using a channel compression matrix to compress the first aliased image to obtain a first compressed aliased image, and compress the second aliased image to obtain a second compressed aliased image.

[0099] In one embodiment, the compression method of compressing the first aliased image and the second aliased image using a channel compression matrix is ​​an existing method, which will not be described in detail.

[0100] It should be noted that the channel compression matrix carries image spatial information as well as information redundancy and correlation between channels, ensuring image quality during compression. Therefore, during the compression process, the channel compression matrix can be considered to transform the spatial information of each channel image. In this case, if the channel compression matrix is ​​directly used to compress the first single-layer image and the first aliased image, the change in coil sensitivity corresponding to the first single-layer image (reflected in the image as the brightness trend of the first single-layer image) will be inconsistent with the change in coil sensitivity corresponding to the first aliased image (the brightness trend of the first aliased image). Consequently, when performing de-aliasing on the first compressed aliased image, the compressed first single-layer image cannot serve as valid reference data.

[0101] Similarly, if the second aliased image obtained in S302 is inconsistent with the first aliased image in the image domain, the coil sensitivity of the second compressed aliased image cannot be guaranteed to be consistent with the coil sensitivity of the first compressed aliased image after compression.

[0102] Based on this, since the target object is consistent in the first aliased image and the second aliased image in the image domain, when the first aliased image and the second aliased image are subsequently compressed, the consistency of the coil sensitivities of the compressed first aliased image and the second aliased image can also be guaranteed, and the amount of data to be processed is reduced.

[0103] In this embodiment, after acquiring multiple layers of first mixed images and multiple first single-layer images of the target object, the first single-layer images can be sequentially shifted and aliased to obtain a second mixed image. Since the first and second mixed images have the same aliasing pattern, the image position of the target object in the first and second mixed images can be assumed to be consistent, thereby maintaining consistency in the image data of the target object in the first and second mixed images. In other words, the first and second mixed images are consistent in the image domain. Subsequently, a channel compression matrix can be extracted from the second mixed image, and the first and second images can be simultaneously compressed using the channel compression matrix to obtain first and second compressed mixed images. Since the image data of the target object in the first and second mixed images are consistent, the coil sensitivities of the first and second compressed mixed images should also be consistent after the first and second mixed images are compressed using the same channel compression matrix, thereby reducing the amount of data required for processing.

[0104] It should be added that the purpose of ensuring the consistency of the coil sensitivities of the compressed first aliased image and the second aliased image is to provide a reliable data basis for the subsequent de-aliasing of the first compressed aliased image as reference data, thereby improving the accuracy of the de-aliasing of the first aliased image.

[0105] Specifically, the terminal device can perform de-aliasing processing on the first compressed aliasing image based on the second compressed aliasing image to generate a multi-layer magnetic resonance image of the target object. Exemplarily, the terminal device can first obtain a multi-layer reference image of the target object based on the second compressed aliasing image. For example, the second compressed aliasing image is sequentially subjected to inverse aliasing translation processing to obtain a multi-layer reference image, so that the image position of the target object in each reference image is consistent with the initial image position in the first single-layer image. Then, based on the multi-layer reference image, the first compressed aliasing image is de-aliased and reconstructed to obtain a multi-layer magnetic resonance image of the target object, so that the image position of the target object in each de-aliased magnetic resonance image is consistent with the image position in each reference image.

[0106] It is understood that the magnetic resonance image is obtained by processing the first aliased image, and as described above, the resolution of the first aliased image is higher than that of the first single-slice image. Furthermore, the reference image is obtained by performing an inverse aliasing shift on the second compressed aliased image, which is also obtained by performing a shift, aliasing, and compression process based on the first single-slice image. Therefore, it can be considered that the resolution of the magnetic resonance image is higher than that of the first single-slice image or the reference image.

[0107] In one embodiment, the above-described S302 sequentially performs translation and aliasing processing on the first single-layer image. Based on this, the above-described inverse aliasing and translation processing can be considered to be first de-aliasing the second compressed aliased image to obtain de-aliased images. Subsequently, each de-aliased image is inversely translated to obtain a reference image, such that the image position of the target object in each reference image is consistent with the initial image position of the first single-layer image. At this point, the reference image can be used as reference data for the first compressed aliased image to perform de-aliasing processing on the first compressed aliased image.

[0108] Specifically, refer to Figure 5 First, the first single-layer images I1, I2, and I3 are sequentially shifted (S1) and aliased (S2) to obtain a second aliased image P2. The terminal device can then extract a channel compression matrix (S3) from the second aliased image P2 and apply it to the first aliased image (S4) to obtain a first compressed aliased image (not shown). Furthermore, the terminal device can apply it to the shifted first single-layer image (S4) and perform an inverse shift (S5) to obtain a reference image (not shown). Finally, the first compressed aliased image is de-aliased based on the reference image.

[0109] The method for performing de-aliasing on the aliased image (the first compressed aliased image) based on the reference data (each reference image) is an existing method and will not be described in detail. For example, a pre-trained de-aliasing model can be used to take the first compressed aliased image and each reference image as input to obtain a de-aliased multi-slice magnetic resonance image.

[0110] It should be noted that when the number of the first aliased images is N, the magnetic resonance images obtained will also be N. That is, N single-layer magnetic resonance images are obtained.

[0111] It should be noted that the purpose of acquiring the first aliased image is to obtain a single-layer MRI image. However, it should be noted that in this embodiment, the first single-layer image acquired in S301 cannot be used as an MRI image. This is because if a single-layer MRI image that meets the requirements is to be acquired directly (the image contrast of the MRI image must be greater than a preset contrast ratio to ensure that the target object is displayed in the MRI image as required), the EPI sequence corresponding to the target object must be used to acquire one layer of image at a time. Furthermore, if the number of single-layer MRI images to be acquired is N, the EPI sequence must be used to perform the above acquisitions N times, which consumes a considerable amount of acquisition time and may significantly degrade image quality due to variations in parameters such as TE and TR.

[0112] However, in this embodiment, the first single-layer image acquired as reference data in S301 need only include the first single-layer image corresponding to the target object. In other words, the image contrast of the first single-layer image is not limited and can be significantly lower than the preset contrast, eliminating the need for optimal display of the target object in the first single-layer image. In this case, the EPI sequence used to acquire the first single-layer image can be different from the EPI sequence used to acquire the first aliased image.

[0113] For example, an EPI sequence that can quickly complete acquisition of the target object can be selected. In this case, the generated first single-layer image, which has poor or average display quality, can include the target object and serve as reference data. That is, in S301 above, the image contrast of the first aliased image is greater than the preset contrast, and the image contrast of the first single-layer image is less than the preset contrast. Alternatively, the image contrast of the first aliased image is greater than the image contrast of the first single-layer image.

[0114] Furthermore, as described above, simultaneous multi-slice excitation imaging can simultaneously acquire multiple layers of a target object to produce a first aliased image. Therefore, acquisition efficiency is significantly higher than when a single layer of the target object is excited at a time. Based on this, employing steps S301-S304 and performing a dealiasing process on the first compressed aliased image to produce a magnetic resonance image can improve the efficiency of magnetic resonance imaging while maintaining the quality of the resulting magnetic resonance image.

[0115] The preset contrast ratio can be set according to actual conditions and is not limited thereto.

[0116] In another embodiment, the terminal device may first obtain a receive coil sensitivity distribution map based on a multi-layer reference image. Then, based on the receive coil sensitivity distribution map, the terminal device may perform dealiasing and reconstruction on the first compressed aliased image to obtain a multi-layer magnetic resonance image of the target object. The resolution of the magnetic resonance image is higher than that of the first single-layer image or the reference image.

[0117] In one embodiment, as described above regarding coils, coil sensitivity indicates the receiving coil's response to input signals. A higher value indicates a stronger ability to detect weak signals. The terminal device may first obtain a coil sensitivity distribution map based on a multi-layer reference image when generating the multi-layer reference image. The same coil sensitivity distribution map is then used to perform dealiasing and reconstruct the first compressed aliased image to obtain a magnetic resonance image.

[0118] It should be noted that obtaining the coil sensitivity distribution map when generating a multi-layer reference image and reconstructing the first compressed aliased image based on the coil sensitivity distribution map (de-aliasing the first compressed aliased image has been described above) are existing technologies and will not be described in detail.

[0119] See also Figure 6 , Figure 6 This is a flowchart of a magnetic resonance image reconstruction method provided in an embodiment of the present application. Detailed description is as follows:

[0120] S601 , obtaining a first aliased image of a target object, where the first aliased image is reconstructed by scanning the target object using a multi-layer simultaneous excitation pulse sequence.

[0121] The multi-layer simultaneous excitation pulse sequence has been described above as a single-shot EPI imaging sequence or a multi-shot EPI imaging sequence, and further explanation thereof is omitted. Furthermore, the method for obtaining the first aliased image can be referred to the example described in step S301 above, and further explanation thereof is omitted.

[0122] S602: Acquire a second aliased image, where the resolution of the second aliased image is lower than that of the first aliased image, and the first aliased image and the second aliased image have the same aliasing mode.

[0123] The same aliasing mode has been explained in the above S302 and will not be described again.

[0124] It should be noted that the method for obtaining the second aliased image is to first execute the above-mentioned step S301 to obtain multiple first single-layer images of the target object, and then execute the above-mentioned step S302 to perform translation and aliasing processing on the first single-layer images in sequence to obtain the second aliased image.

[0125] In another embodiment, the terminal device may also directly scan the target object based on a preset reference sequence to obtain reconstruction, wherein the reference sequence may also be the single-shot EPI imaging sequence or the multi-shot EPI imaging sequence described above.

[0126] In this embodiment, the method of using a reference sequence to scan the target object to obtain the first aliased image is similar to the method of using a multi-layer simultaneous excitation pulse sequence to excite the target object multiple times to obtain the first aliased image in the above embodiment, and will not be described again.

[0127] S603: Extract a channel compression matrix from the second aliased image.

[0128] S604 : compress the first aliased image using a channel compression matrix to obtain a first compressed aliased image, and compress the second aliased image to obtain a second compressed aliased image.

[0129] S605 : Obtain a multi-layer reference image of the target object according to the second compressed aliased image.

[0130] S606 : Perform dealiasing and reconstruction on the first compressed aliased image based on the multi-layer reference image to obtain a multi-layer magnetic resonance image of the target object.

[0131] The above steps S603 to S606 have been explained in the above embodiments and will not be described again.

[0132] Based on the above generation of the first compressed aliased image and the second compressed aliased image, it can be seen that in this embodiment, by using the same channel compression matrix to compress the first and second compressed aliased images simultaneously, the coil sensitivities of the first and second compressed aliased images can be made consistent, and the amount of data required for processing is reduced. Furthermore, because the first and second compressed aliased images have the same aliasing pattern, the multi-layer reference image obtained based on the second compressed aliased image is more accurate and effective. By performing dealiasing and reconstruction on the first compressed aliased image based on the more effective reference image, the resulting multi-layer magnetic resonance image is also more accurate and effective.

[0133] See also Figure 7 , Figure 7 Schematic diagram of the structure of a magnetic resonance image processing device provided in an embodiment of the present application. The magnetic resonance image processing device in this embodiment includes modules for executing Figure 3 Each step in the corresponding embodiment. Please refer to Figure 3 as well as Figure 3 For the convenience of explanation, only the parts related to this embodiment are shown. Figure 7 The magnetic resonance image processing apparatus 700 may include: a first acquisition module 710, a first processing module 720, a first extraction module 730, and a first compression module 740, wherein:

[0134] The first acquisition module 710 is configured to acquire multiple layers of first aliased images of the target object and acquire multiple first single-layer images of the target object.

[0135] The first processing module 720 is configured to perform translation and aliasing processing on the first single-layer image in sequence to obtain a second aliased image; the first aliased image and the second aliased image have the same aliasing mode.

[0136] The first extraction module 730 is configured to extract a channel compression matrix from the second aliased image.

[0137] The first compression module 740 is configured to compress the first aliased image using a channel compression matrix to obtain a first compressed aliased image, and to compress the second aliased image to obtain a second compressed aliased image.

[0138] In one embodiment, the first obtaining module 710 is further configured to:

[0139] A multi-layer simultaneous excitation pulse sequence is used to excite the target object multiple times to obtain k-space data of the multiple excitations; the phase difference between two adjacent rows of the k-space data is a preset phase difference; the k-space data is reconstructed to obtain a first aliased image.

[0140] In one embodiment, the first obtaining module 710 is further configured to:

[0141] Acquire multiple first single-layer images that meet preset conditions; the preset conditions include one or more of the following conditions: the distance between two adjacent data points in the k-space data corresponding to each first single-layer image is consistent with the distance between two adjacent data points in the k-space data corresponding to the first aliased image; the image resolution of each first single-layer image is less than the image resolution of the first aliased image; and the k-space data corresponding to each first single-layer image has not been downsampled.

[0142] In one embodiment, the magnetic resonance image processing apparatus 700 further includes:

[0143] The fourth processing module is configured to perform a dealiasing process on the first compressed aliasing image according to the second compressed aliasing image to generate a multi-slice magnetic resonance image of the target object.

[0144] In one embodiment, the fourth processing module is further configured to:

[0145] A multi-layer reference image of the target object is obtained according to the second compressed aliased image; and the first compressed aliased image is dealiased and reconstructed based on the multi-layer reference image to obtain a multi-layer magnetic resonance image of the target object.

[0146] In one embodiment, the fourth processing module is further configured to:

[0147] Based on the multi-layer reference image, a receiving coil sensitivity distribution map is obtained; based on the receiving coil sensitivity distribution map, the first compressed aliased image is dealiased and reconstructed to obtain a multi-layer magnetic resonance image of the target object, wherein the resolution of the magnetic resonance image is higher than that of the first single-layer image or the reference image.

[0148] When it is understood that Figure 7 In the structural diagram of the magnetic resonance image processing device shown, each module is used to perform Figure 3 The steps in the corresponding embodiment, and for Figure 3 Each step in the corresponding embodiment has been explained in detail in the above embodiment. Figure 3 as well as Figure 3 The relevant descriptions in the corresponding embodiments will not be repeated here.

[0149] See also Figure 8 , Figure 8 Schematic diagram of the structure of a magnetic resonance image reconstruction device provided in an embodiment of the present application. The magnetic resonance image reconstruction device in this embodiment includes modules for executing Figure 6 Each step in the corresponding embodiment. Please refer to Figure 6 as well as Figure 6 For the convenience of explanation, only the parts related to this embodiment are shown. Figure 8 The magnetic resonance image reconstruction apparatus 800 may include: a second acquisition module 810, a third acquisition module 820, a second extraction module 830, a second compression module 840, a second processing module 850, and a third processing module 860, wherein:

[0150] The second acquisition module 810 is configured to acquire a first aliased image of the target object. The first aliased image is reconstructed by scanning the target object using a multi-layer simultaneous excitation pulse sequence.

[0151] The third acquisition module 820 is configured to acquire a second aliased image. The second aliased image has a lower resolution than the first aliased image, and the first aliased image and the second aliased image have the same aliasing pattern.

[0152] The second extraction module 830 is configured to extract a channel compression matrix from the second aliased image.

[0153] The second compression module 840 is configured to compress the first aliased image using a channel compression matrix to obtain a first compressed aliased image, and to compress the second aliased image to obtain a second compressed aliased image.

[0154] The second processing module 850 is configured to obtain a multi-layer reference image of the target object according to the second compressed aliased image.

[0155] The third processing module 860 is configured to perform dealiasing and reconstruction on the first compressed aliased image based on the multi-layer reference image to obtain a multi-layer magnetic resonance image of the target object.

[0156] In one embodiment, the second aliased image is reconstructed by scanning the target object using a reference sequence.

[0157] In one embodiment, the third obtaining module 820 is further configured to:

[0158] A plurality of first single-layer images of the target object are acquired; and the first single-layer images are sequentially translated and aliased to obtain second aliased images.

[0159] When it is understood that Figure 8 In the structural diagram of the magnetic resonance image processing device shown, each module is used to perform Figure 6The steps in the corresponding embodiment, and for Figure 6 Each step in the corresponding embodiment has been explained in detail in the above embodiment. Figure 6 as well as Figure 6 The relevant descriptions in the corresponding embodiments will not be repeated here.

[0160] Figure 9 This is a schematic diagram of the structure of a terminal device provided by an embodiment of the present application. Figure 9 As shown, the terminal device 900 of this embodiment includes: a processor 910, a memory 920, and a computer program 930 stored in the memory 920 and executable by the processor 910, such as a program for the magnetic resonance image processing method. When the processor 910 executes the computer program 930, the steps in each embodiment of the above-mentioned magnetic resonance image processing method and each magnetic resonance image reconstruction method are implemented, such as Figure 3 S301 to S304 shown and Figure 6 Alternatively, the processor 910 executes the computer program 930 to implement the above Figure 7 and Figure 8 The functions of each module in the corresponding embodiment are, for example, Figure 7 and Figure 8 For details on the functions of each module shown, please refer to Figure 7 and Figure 8 Related description in the corresponding embodiment.

[0161] Exemplarily, the computer program 930 may be divided into one or more modules, one or more of which are stored in the memory 920 and executed by the processor 910 to implement the magnetic resonance image processing method and the magnetic resonance image reconstruction method provided in the embodiments of the present application. One or more modules may be a series of computer program instruction segments capable of performing specific functions, and the instruction segments are used to describe the execution process of the computer program 930 in the terminal device 900. For example, the computer program 930 may implement the magnetic resonance image processing method and the magnetic resonance image reconstruction method provided in the embodiments of the present application.

[0162] The terminal device 900 may include, but is not limited to, a processor 910 and a memory 920. Those skilled in the art will appreciate that Figure 9 It is merely an example of the terminal device 900 and does not constitute a limitation of the terminal device 900. The terminal device may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the terminal device may also include input and output devices, network access devices, buses, etc.

[0163] The processor 910 may be a central processing unit, or other general-purpose processor, a digital signal processor, an application-specific integrated circuit, an off-the-shelf programmable gate array or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0164] The memory 920 may be an internal storage unit of the terminal device 900, such as a hard disk or memory of the terminal device 900. The memory 920 may also be an external storage device of the terminal device 900, such as a plug-in hard disk, a smart memory card, a flash memory card, etc. equipped on the terminal device 900. Furthermore, the memory 920 may include both an internal storage unit of the terminal device 900 and an external storage device.

[0165] An embodiment of the present application provides a computer-readable storage medium storing a computer program. The computer program is used by a processor to execute the magnetic resonance image processing method and / or magnetic resonance image reconstruction method in each of the above embodiments.

[0166] An embodiment of the present application provides a computer program product. When the computer program product is run on a terminal device, the terminal device executes the magnetic resonance image processing method and / or the magnetic resonance image reconstruction method in the above-mentioned embodiments.

[0167] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A magnetic resonance image processing method, characterized in that: The method comprises: Acquire multiple layers of first aliased images of a target object, and acquire multiple first single-layer images of the target object; performing translation and aliasing processing on the first single-layer image in sequence to obtain a second aliased image; wherein the first aliased image and the second aliased image have the same aliasing mode; extracting a channel compression matrix from the second aliased image; The first aliased image is compressed using the channel compression matrix to obtain a first compressed aliased image, and the second aliased image is compressed to obtain a second compressed aliased image.

2. The method according to claim 1, characterized in that The step of obtaining a multi-layer first aliased image of the target object includes: The target object is excited multiple times using a multi-layer simultaneous excitation pulse sequence to obtain k-space data of the multiple excitations; the phase difference between two adjacent rows of data in the k-space data is a preset phase difference; The k-space data is reconstructed to obtain the first aliased image.

3. The method according to claim 2, characterized in that The multi-layer simultaneous excitation pulse sequence is a single-shot excitation EPI imaging sequence or a multi-shot excitation EPI imaging sequence.

4. The method according to claim 1, wherein The acquiring of a plurality of first single-layer images of the target object comprises: Acquire a plurality of first single-layer images that meet preset conditions; the preset conditions include one or more of the following conditions: The distance between two adjacent data points in the k-space data corresponding to each of the first single-layer images is consistent with the distance between two adjacent data points in the k-space data corresponding to the first aliased image; The image resolution of each of the first single-layer images is smaller than the image resolution of the first aliased image; The k-space data corresponding to each of the first single-layer images is not downsampled.

5. The method according to any one of claims 1 to 4, characterized in that After compressing the first aliased image using the channel compression matrix to obtain a first compressed aliased image, and compressing the second aliased image to obtain a second compressed aliased image, the method further includes: Performing a dealissing process on the first compressed aliasing image according to the second compressed aliasing image to generate a multi-slice magnetic resonance image of the target object.

6. The method according to claim 5, characterized in that The performing a de-aliasing process on the first compressed aliasing image according to the second compressed aliasing image to generate a multi-slice magnetic resonance image of the target object includes: Obtain a multi-layer reference image of the target object according to the second compressed aliased image; The first compressed aliased image is dealiased and reconstructed based on the multi-layer reference image to obtain a multi-layer magnetic resonance image of the target object.

7. The method according to claim 6, characterized in that The step of performing dealiasing and reconstruction on the first compressed aliased image based on the multi-layer reference image to obtain a multi-layer magnetic resonance image of the target object includes: acquiring a receiving coil sensitivity distribution map based on the multi-layer reference image; The first compressed aliased image is dealiased and reconstructed based on the receiving coil sensitivity distribution map to obtain a multi-layer magnetic resonance image of the target object, where the resolution of the magnetic resonance image is higher than that of the first single-layer image or the reference image.

8. A magnetic resonance image reconstruction method, characterized in that: The method comprises: Acquiring a first aliased image of the target object, where the first aliased image is reconstructed by scanning the target object using a multi-layer simultaneous excitation pulse sequence; Acquire a second aliased image, where the second aliased image has a lower resolution than the first aliased image, and the first aliased image and the second aliased image have the same aliasing mode; extracting a channel compression matrix from the second aliased image; compressing the first aliased image using the channel compression matrix to obtain a first compressed aliased image, and compressing the second aliased image to obtain a second compressed aliased image; Obtain a multi-layer reference image of the target object according to the second compressed aliased image; The first compressed aliased image is dealiased and reconstructed based on the multi-layer reference image to obtain a multi-layer magnetic resonance image of the target object.

9. The method according to claim 8, characterized in that The second aliased image is reconstructed and obtained by scanning the target object using a reference sequence.

10. The method according to claim 8, characterized in that The acquiring of the second aliased image comprises: Acquiring a plurality of first single-layer images of the target object; The first single-layer image is sequentially subjected to translation and aliasing processing to obtain a second aliased image.

Images