Image reconstruction method and device, computer device and storage medium

Abstract

The application relates to an image reconstruction method and device, computer equipment and a storage medium. The application comprises the following steps: acquiring ASL control data and ASL mark data of a target region in K space; independently reconstructing the ASL control data of each segment to obtain a first complete image corresponding to the ASL control data of each segment; independently reconstructing the ASL mark data of each segment to obtain a second complete image corresponding to the ASL mark data of each segment; combining the first complete images corresponding to each segment to obtain a control image of the target region after reconstruction; and combining the second complete images corresponding to each segment to obtain a mark image of the target region after reconstruction. The method can reconstruct an image without artifacts, thereby greatly improving the accuracy of the reconstructed image.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, and in particular, to an image reconstruction method and device, a computer device, and a storage medium. BACKGROUND

[0002] Arterial spin labeling (ASL) imaging technology usually requires using a high-speed three-dimensional data acquisition sequence to achieve perfusion imaging covering the whole brain.

[0003] For example, a 3D GRASE (Gradient Recalled And Spin Echo) sequence is one of the most commonly used three-dimensional data acquisition sequences for ASL. Taking the 3D GRASE sequence as an example, a complete three-dimensional K-space data needs to be divided into several GRASE echo chain segments for acquisition, each echo chain acquires a part of K-space data in the phase encoding (PE) direction, and then the K-space data acquired in each segment is spliced along the PE direction to form complete K-space data, and then an image corresponding to the ordinary space is reconstructed according to the complete K-space data.

[0004] However, there is a phase difference between the K-space data acquired in segments in the above method, which causes the reconstructed image to have artifacts, thereby reducing the accuracy of the reconstructed image. SUMMARY

[0005] Therefore, it is necessary to provide an image reconstruction method, device, computer device, and storage medium capable of improving the accuracy of the reconstructed image to solve the above technical problems.

[0006] In a first aspect, an embodiment of the present application provides an image reconstruction method, which comprises:

[0007] acquiring ASL control data and ASL label data of a target region in multiple segments of K-space;

[0008] independently reconstructing the ASL control data of each segment to obtain a first complete image corresponding to the ASL control data of each segment, and independently reconstructing the ASL label data of each segment to obtain a second complete image corresponding to the ASL label data of each segment;

[0009] merging the first complete images corresponding to each segment to obtain a control image of the target region after reconstruction, and merging the second complete images corresponding to each segment to obtain a label image of the target region after reconstruction.

[0010] In one of the embodiments, the method further comprises:

[0011] According to the control image and the labeling image, a perfusion image of the target region is determined.

[0012] In one of the embodiments, the acquiring the ASL control data and the ASL labeling data of the target region in multiple segments in K-space includes:

[0013] According to a preset data acquisition sequence, the ASL control data of the target region in K-space is acquired in multiple segments to obtain multiple ASL control data; and according to the preset data acquisition sequence, the ASL labeling data of the target region in K-space is acquired in multiple segments to obtain multiple ASL labeling data.

[0014] In one of the embodiments, the ASL control data of each segment and the labeling data of each segment present a regular down-sampling distribution in K-space.

[0015] In one of the embodiments, the independently reconstructing the ASL control data of each segment to obtain a first complete image corresponding to the ASL control data of each segment, and independently reconstructing the ASL labeling data of each segment to obtain a second complete image corresponding to the ASL labeling data of each segment includes:

[0016] According to a preset image reconstruction method, the ASL control data of each segment is independently reconstructed to obtain a first complete image; and according to the preset image reconstruction method, the ASL labeling data of each segment is independently reconstructed to obtain a second complete image.

[0017] In one of the embodiments, the merging the first complete images corresponding to the segments to obtain a control image of the target region after reconstruction, and merging the second complete images corresponding to the segments to obtain a labeling image of the target region after reconstruction includes:

[0018] The image composed of the mean values of the pixel values of the same positions in the first complete images is determined as the control image; and the image composed of the mean values of the pixel values of the same positions in the second complete images is determined as the labeling image.

[0019] In one of the embodiments, the determining the perfusion image of the target region after reconstruction according to the control image and the labeling image includes:

[0020] The amplitude of each pixel in the control image is subtracted by the amplitude of the pixel at the corresponding position in the labeling image to obtain the perfusion image.

[0021] In a second aspect, the embodiments of the present application provide a magnetic resonance image reconstruction method, which includes:

[0022] acquire ASL data of the target region in multiple segments in K-space; the multiple segments of ASL data are divided along a phase encoding direction;

[0023] independently reconstruct each segment of ASL data to obtain a complete image corresponding to each segment of ASL data respectively;

[0024] merge the complete images corresponding to each segment to obtain a reconstructed magnetic resonance image of the target region.

[0025] In one embodiment, the ASL data includes ASL control data and ASL labeling data, the complete image includes a first complete image corresponding to the ASL control data and a second complete image corresponding to the ASL labeling data, the first complete image is merged to form a control image, and the second complete image is merged to form a labeling image.

[0026] The method further includes:

[0027] determine a perfusion image of the target region according to the control image and the labeling image.

[0028] In a third aspect, an embodiment of the present application provides an image reconstruction device, which includes:

[0029] an acquisition module, configured to acquire ASL control data and ASL labeling data of a target region in multiple segments in K-space;

[0030] a reconstruction module, configured to independently reconstruct each segment of ASL control data to obtain a first complete image corresponding to each segment of ASL control data respectively, and independently reconstruct each segment of ASL labeling data to obtain a second complete image corresponding to each segment of ASL labeling data respectively;

[0031] a merging module, configured to merge the first complete images corresponding to each segment to obtain a reconstructed control image of the target region, and merge the second complete images corresponding to each segment to obtain a reconstructed labeling image of the target region.

[0032] In a third aspect, an embodiment of the present application provides a computer device, including a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the steps of any method provided in the first aspect or the second aspect.

[0033] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the steps of any method provided in the first aspect or the second aspect.

[0034] This application provides an image reconstruction method, apparatus, computer device, and storage medium. It acquires arterial spin labeling (ASL) control data and ASL label data of multiple segments in K-space for a target region. Each ASL control data segment is independently reconstructed to obtain a first complete image of the target region corresponding to each ASL control data segment. Similarly, each ASL label data segment is independently reconstructed to obtain a second complete image of the target region corresponding to each ASL label data segment. The first complete images are then merged to obtain a reconstructed control image of the target region. Finally, the second complete images are merged to obtain a reconstructed label image of the target region. In this method, each segment of K-space data (control data or label data) is reconstructed separately during image reconstruction, thus avoiding phase difference issues between different segments. Each segment produces an image without ghost artifacts, and merging the images from different segments to obtain the reconstructed image of the target region (label image or control image) also results in an image without ghost artifacts, thereby significantly improving the accuracy of the reconstructed image. Attached Figure Description

[0035] Figure 1 An application environment diagram for image reconstruction is provided for one embodiment;

[0036] Figure 1a A schematic diagram illustrating a data splicing method as provided in one embodiment;

[0037] Figure 2 A schematic flowchart of an image reconstruction method provided in one embodiment;

[0038] Figure 2a This is a schematic diagram illustrating the scanning of a target area according to one embodiment;

[0039] Figure 2b A schematic diagram of a pulse sequence provided for one embodiment;

[0040] Figure 3 A schematic diagram of the target region map reconstructed from different segmented data in one embodiment;

[0041] Figure 3a A schematic diagram of two-dimensional K-space acquisition provided for one embodiment;

[0042] Figure 3b A schematic diagram of a three-dimensional K-space acquisition provided for one embodiment;

[0043] Figure 4 A comparison of reconstructed perfusion images provided in one embodiment;

[0044] Figure 5A magnetic resonance image reconstruction method is provided for an embodiment;

[0045] Figure 6 A flowchart of an image reconstruction method is provided for an embodiment;

[0046] Figure 7 A structural block diagram of an image reconstruction device is provided for an embodiment;

[0047] Figure 8 A structural block diagram of an image reconstruction device is provided for another embodiment. DETAILED DESCRIPTION

[0048] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0049] The image reconstruction method provided by the present application can be applied to an application environment as shown in Figure 1 The application environment, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used for related data of image reconstruction. The network interface of the computer device is used for communication with other devices outside through network connection. The computer program is executed by the processor to implement an image reconstruction method.

[0050] Arterial spin labeling (ASL) imaging technology has important application value in magnetic resonance clinical and scientific research perfusion imaging. The most commonly used data acquisition sequence 3D GRASE (Gradient Recalled And Spin Echo, spin echo FSE and gradient echo EPI combined sequence) sequence is a segmented K-space data acquisition, and is reconstructed by K-space data splicing method, as shown in Figure 1aAs shown, it is a schematic diagram of data splicing mode, wherein the points in the columns with the same number represent the K-space data collected by different segments; however, when 3D GRASE segment acquisition is performed, due to system errors or movement of the subject and other factors, the K-space data collected by different GRASE segments has phase difference, thereby causing the image obtained by the traditional splicing reconstruction mode to have obvious Ghost artifacts in the PE direction, further causing the ASL perfusion image obtained by the control image and the marker image to further amplify the artifacts, resulting in that the image reconstruction result is not accurate enough, and affecting the image diagnosis effect. Based on this, the embodiment of the present application provides an image reconstruction method, device, computer equipment and storage medium, which can effectively eliminate the Ghost artifacts in the ASL perfusion image without changing the original sequence acquisition mode, and improve the accuracy of the reconstructed image.

[0051] The technical solutions of the present application and how the technical solutions solve the above technical problems will be described in detail below through embodiments and in combination with the drawings. The following specific embodiments can be combined with each other, and the same or similar concepts or processes can not be described again in some embodiments. It should be noted that the image reconstruction method provided by the present application, Figures 2-6 The execution subject of the image reconstruction method is a computer equipment, wherein the execution subject can also be an image reconstruction device, wherein the device can be realized by software, hardware or a combination of software and hardware to become part or all of the computer equipment.

[0052] To make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all the embodiments.

[0053] In one embodiment, Figure 2 A computer equipment method is provided, and the embodiment relates to the specific process that the computer equipment independently reconstructs the complete image of the target region in each segment according to the ASL control data and ASL marker data of the target region in each segment of K-space, and then determines the marker image of the target region after reconstruction according to the complete image of the target region in each segment, as shown in Figure 2 The method comprises the following steps:

[0054] S101, acquiring ASL control data and ASL marker data of a target region in multiple segments of K-space.

[0055] The target region refers to the area that needs to be reconstructed, such as the head, brain, or myocardium. K-space, also known as Fourier space, is the filling space of the original digital data of the MR signal containing spatial positioning coding information. This space can represent the spatial frequency in the MR image. Performing a Fourier transform on the K-space data decodes the spatial positioning coding information in the original digital data, decomposing it into magnetic resonance signals of different frequencies, phases, and amplitudes. Multiple segments refer to the segmentation of the acquisition sequence. For example, a 3D gradient-and-spin-echo (GRASE) sequence is used to acquire K-space data of the target region, and the K-space is segmented according to the echo chain number of the corresponding echo signal. In this embodiment, the echo chain represents the number of echoes generated and acquired after one radio frequency pulse excitation; echoes generated by one excitation are divided into one segment.

[0056] Taking the target area as the brain as an example, ASL control data refers to the brain control image acquired through ASL imaging, while ASL labeling data refers to the brain labeling image acquired through ASL imaging. The brain labeling image labels the blood flowing into the brain, and the control image also labels the blood flowing into the brain, but the effect is the same as no labeling.

[0057] like Figure 2a As shown in Figure 2a, this is a schematic diagram illustrating the scanning of a target region according to an embodiment of this application. Along the time axis, it includes an ASL control pulse application module, a 3D GRASE sequence application module (control data acquisition module), an ASL marker pulse application module, and a 3D GRASE sequence application module (marker data acquisition module). In this embodiment, the first type of sequence formed by the ASL control pulse application module and the 3D GRASE sequence application module, and the second type of sequence formed by the ASL marker pulse application module and the 3D GRASE sequence application module, can contain multiple (i.e., two or more types) sequences applied alternately.

[0058] One embodiment of this application includes a first type sequence or a second type sequence comprising an ASL preparation pulse sequence and a GRASE pulse sequence applied after a set delay time. When the ASL preparation pulse sequence uses an ASL control pulse, the scan sequence is a first type sequence; when the ASL preparation pulse sequence uses an ASL marker pulse, the scan sequence is a second type sequence, wherein there is a phase difference between the ASL control pulse and the ASL marker pulse.

[0059] like Figure 2b As shown, Figure 2b This is a schematic diagram of a pulse sequence; where RF represents the radio frequency pulse emitted by the radio frequency coil; the gradient coils respectively form a gradient G selected along the plane. z Direction and phase encoding gradient Gy direction and frequency encoding gradient G x direction. The ASL preparation pulse sequence includes a number of small flip angle radio frequency pulses and a slice selection gradient G z direction. The GRASE pulse sequence includes a 90° excitation pulse and a number of 180° aggregation pulses applied subsequently, each radio frequency pulse is applied at a corresponding timing position with a slice selection gradient G z direction. The ASL preparation pulse sequence includes a number of small flip angle radio frequency pulses and a slice selection gradient G z direction and phase encoding gradient G y direction. The encoding gradient 203, 204 is applied, and after the application of the encoding gradient 203, 204, a frequency encoding gradient G x direction. At the moment of the polarity change of the continuous inversion gradient 207, a phase encoding gradient G y direction. The ASL preparation pulse sequence includes a number of small flip angle radio frequency pulses and a slice selection gradient G y direction. In some scenarios, to ensure the center priority acquisition of K-space data, a frequency encoding gradient G x direction. After the application of the continuous inversion gradient 207 is completed, a slice selection gradient G z direction and phase encoding gradient G y direction. The re-aggregation gradient 208, 209 is applied to reduce the influence of the echo signal of the previous acquisition on the signal acquisition of the next 180° aggregation pulse.

[0060] Optionally, when the computer device acquires the ASL control data and the ASL labeling data of the target region in multiple segments of K-space, the ASL control data of the target region in K-space can be acquired under multiple segments according to a preset data acquisition sequence to obtain multiple ASL control data; the ASL labeling data of the target region in K-space is acquired under multiple segments according to a preset data acquisition sequence to obtain multiple ASL labeling data. Optionally, the ASL control data of each segment and the labeling data of each segment present a regular down-sampling distribution in K-space.

[0061] The preset data acquisition sequence can be a 3D GRASE sequence or other sequences, which are not limited in the embodiment. When the 3D GRASE sequence is used to segmentally acquire the K-space data of the target region, the echo chain of the 3D GRASE is divided into several segments according to the data of the complete three-dimensional K-space of the target region, each segment acquires part of the K-space data in the phase encoding (PE) direction, and then the data acquired by each segment is spliced along the PE direction to form the complete K-space data. Based on the segmental acquisition process, the ASL control data of the target region is acquired by segmentally acquiring the ASL control data after the target region is labeled, and the ASL control data of the target region is acquired by segmentally acquiring the ASL control data without labeling the target region, so that the ASL control data of the target region in multiple segments and the ASL labeled data of the target region in multiple segments are obtained.

[0062] In addition, when the K-space data of the target region is acquired by using the acquisition sequence, the K-space data is acquired by downsampling, so that the ASL control data of each segment and the ASL labeled data of each segment acquired by the computer device present a regular downsampling distribution in the K-space.

[0063] S102, independently reconstructing each segment of the ASL control data to obtain a first complete image corresponding to each segment of the ASL control data; independently reconstructing each segment of the ASL labeled data to obtain a second complete image corresponding to each segment of the ASL labeled data.

[0064] Based on the above-mentioned multiple segments of the ASL control data and the multiple segments of the ASL labeled data, a complete image of the target region is independently established for each segment of the ASL control data, and each complete image of the target region obtained is called a first complete image; similarly, a complete image of the target region is also independently established for each segment of the ASL labeled data, and each complete image of the target region obtained is called a second complete image. For example, in the above-mentioned S101 step, there are three segments of the ASL control data: A1, A2, A3; there are three segments of the ASL labeled data: B1, B2, B3; then a first complete image of the target region is reconstructed according to A1, a first complete image of the target region is reconstructed according to A2, and a first complete image of the target region is reconstructed according to A3, so that the three segments of the ASL control data obtain three first complete images of the target region; the three segments of the ASL labeled data also obtain three second complete images of the target region, and the process is not described again in the embodiment.

[0065] Optionally, the ASL control data of each segment can be independently reconstructed according to a preset image reconstruction method to obtain each first complete image; the ASL marker data of each segment can be independently reconstructed according to a preset image reconstruction method to obtain each second complete image.

[0066] The preset image reconstruction methods include, but are not limited to, GRAPPA (generalized auto-calibrating partially parallel acquisitions), which estimates unacquired points by linearly weighting adjacent acquired points, with the weights estimated using the fully acquired auto-calibration signal region; or iterative reconstruction methods, such as SPIRIT (Iterative self-consistent parallel imaging reconstruction from arbitrary k-space), which is another more general parallel imaging method based on GRAPPA and can be applied to arbitrary sampling trajectories. Its principle is that all frequency domain points are assumed to be equal to the weighted average of all points in their surrounding coils, i.e., a convolution kernel can be found. Convolving the fully acquired frequency domain data through this kernel should still yield the same image. This embodiment does not limit the preset image reconstruction method.

[0067] like Figure 3 The diagram shows the reconstructed complete images of the target region from different data segments (segments A1, A2, and A3). Hollow dots represent data points not collected in that segment, while solid dots represent data points collected in that segment. After obtaining segmented K-space data through 3D GRASE sequence acquisition, the segmented K-space data are not directly stitched together for reconstruction. Instead, each segment of K-space data is reconstructed separately, thus avoiding the phase difference problem between different segments. This results in images without ghost artifacts for each segment.

[0068] like Figure 3a The diagram shown is a schematic representation of a two-dimensional K-space acquisition in one embodiment of this application. In the diagram, Kx represents the frequency encoding direction / readout direction, and Ky represents the phase encoding direction. Furthermore, the diagram contains three different data lines, each corresponding to... Figure 2b The echo signal filling data after the first 180° focusing pulse, the echo signal filling data after the second 180° focusing pulse, and the echo signal filling data after the third 180° focusing pulse. In this embodiment, the K-space is divided into three segments along the Ky direction, and the spacing between adjacent data lines in each segment is based on... Figure 2bThe area of the sharp waveform gradient field 206 in the K-space determines.

[0069] As shown in Figure 3b Fig. 1 shows a schematic diagram of the acquisition of a three-dimensional K-space in an embodiment of the present application. In the figure, Gx represents a gradient applied along the x direction, Gy represents a gradient applied along the y direction, and Gz represents a gradient applied along the z direction. In this embodiment, Gx corresponds to the Kx direction (frequency encoding direction / readout direction) in the K-space; Gy corresponds to the Ky direction (phase encoding direction) in the K-space; and Gz is a phase encoding gradient in the slice direction. In this embodiment, the three-dimensional K-space includes a plurality of two-dimensional K-spaces corresponding to the slices respectively, and each two-dimensional K-space is divided into three segments along the Ky direction. Figure 3a Figure 3a As shown in

[0070] S103, merging the first complete images corresponding to each segment to obtain a control image of the target region after reconstruction; and merging the second complete images corresponding to each segment to obtain a labeling image of the target region after reconstruction.

[0071] After obtaining the first complete images of each segment reconstruction of the ASL control image in the above process, the first complete images of each segment reconstruction are merged to obtain the control image of the target region after reconstruction. For example, there are three segments of ASL control data: A1, A2, and A3. A first complete image T1 of a target region is reconstructed according to A1, a first complete image T2 of a target region is reconstructed according to A2, and a first complete image T3 of a target region is reconstructed according to A3. Then, T1, T2, and T3 are merged to obtain the control image of the target region after reconstruction.

[0072] Similarly, after obtaining the second complete images of each segment reconstruction of the ASL labeling image, the second complete images of each segment reconstruction are merged to obtain the labeling image of the target region after reconstruction. For example, there are three segments of ASL control data: B1, B2, and B3. A first complete image R1 of a target region is reconstructed according to B1, a first complete image R2 of a target region is reconstructed according to B2, and a first complete image R3 of a target region is reconstructed according to B3. Then, R1, R2, and R3 are merged to obtain the labeling image of the target region after reconstruction.

[0073] Optionally, an image formed by the mean values of the pixel values at the same positions in each first complete image can be determined as the control image; and an image formed by the mean values of the pixel values at the same positions in each second complete image can be determined as the labeling image.

[0074] ​In the merging of the first complete images of each segment reconstruction, the mean value of the pixel values of the same position in each first complete image constitutes an image, which is the control image of the target region after reconstruction; for example, T1, T2 and T3 are complete images of the target region reconstructed according to different segment ASL control data, so the number and arrangement of the pixels in T1, T2 and T3 are the same, and the mean value of the pixels of the same position in T1, T2 and T3 is obtained, and an image composed of the mean value is obtained, which is the control image of the target region after reconstruction. Similarly, the second complete images of each segment of the labeling data are merged by using this method to obtain the labeling image of the target region after reconstruction.

[0075] The image reconstruction method provided in the embodiment is to obtain the arterial spin labeling (ASL) control data and ASL labeling data of the target region in multiple segments of K space, independently reconstruct each ASL control data to obtain the first complete image of the target region corresponding to each ASL control data, and independently reconstruct each ASL labeling data to obtain the second complete image of the target region corresponding to each ASL labeling data, then merge each first complete image to obtain the control image of the target region after reconstruction, and merge each second complete image to obtain the labeling image of the target region after reconstruction. In the method, each segment of K space data (control data or labeling data) is independently reconstructed in the image reconstruction process, thereby avoiding the phase difference problem between different segments of data. Each segment of data can generate an image without ghost artifacts, and the images generated by different segments are also ghost artifact-free images after being merged to obtain the image (labeling image or control image) of the target region after reconstruction, thereby greatly improving the accuracy of the reconstructed image.

[0076] Generally, the main function of ASL is to perform magnetic resonance perfusion imaging, so after obtaining the control image of the target region after reconstruction and the labeling image of the target region after reconstruction, the perfusion image of the target region can be further obtained. The perfusion refers to the process of flowing through the blood vessels or pipeline system and being distributed in the capillary bed of the target organ or tissue; for example, when the target region is the brain, ASL imaging collects two groups of data to generate two groups of images: one is a label image, and the other is a control image. The brain perfusion image is obtained by subtracting the label image from the control image. Then in an embodiment, the method further includes: determining the perfusion image of the target region after reconstruction according to the control image and the labeling image. Optionally, the amplitude of each pixel in the control image is subtracted from the amplitude of the pixel at the corresponding position in the labeling image to obtain the perfusion image.

[0077] The control image corresponding to the target region after reconstruction in the above embodiment is a magnetic resonance image of the target region without marking, and the marked image corresponding to the target region after reconstruction is a magnetic resonance image of the target region after marking. The difference between the amplitude of each pixel in the control image and the amplitude of each pixel in the marked image is the perfusion image of the target region.

[0078] Please refer to Figure 4 The effect of the eyeball ASL perfusion image reconstructed by the traditional reconstruction method and the image reconstruction method provided in the present application is shown in FIG. 1. In FIG. 1, a is the perfusion image obtained by the traditional reconstruction method, and the arrow in a indicates that there is a significant eyeball Ghost artifact in the position; b is the perfusion image obtained by the image reconstruction method provided in the present application, and the arrow in b indicates that the artifact is effectively eliminated in the position. It can be seen that the image reconstruction method provided in the embodiment of the present application can effectively eliminate the Ghost artifact in the ASL perfusion image, and improve the accuracy of the reconstructed image.

[0079] In addition, the embodiment of the present application further provides a magnetic resonance image reconstruction method, as shown in Figure 5 The method comprises the following steps:

[0080] In S201, the ASL data of the target region in multiple segments of K space is obtained, and the ASL data in multiple segments is divided along the phase encoding direction.

[0081] In S202, the ASL data in each segment is independently reconstructed to obtain a complete image corresponding to the ASL control data in each segment.

[0082] In S203, the complete images corresponding to each segment are combined to obtain a magnetic resonance image of the target region after reconstruction.

[0083] Optionally, the ASL data comprises ASL control data and ASL marking data, and the complete image comprises a first complete image corresponding to the ASL control data and a second complete image corresponding to the ASL marking data. The first complete image forms a control image after combination, and the second complete image forms a marked image after combination. The method further comprises: determining a perfusion image of the target region according to the control image and the marked image. Optionally, the first complete image corresponding to each segment can be combined to generate a control image by using a weighted manner, and the second complete image corresponding to each segment can be combined to generate a marked image by using a weighted manner.

[0084] In this embodiment, the ASL data of the target region in multiple segments in K space includes ASL control data and ASL marker data. The segmented ASL data is divided along the phase encoding direction. The complete images reconstructed independently are the first complete image corresponding to the ASL control data and the second complete image corresponding to the ASL marker data. Accordingly, the first complete image is merged to form the control image, and the second complete image is merged to form the marker image.

[0085] For detailed procedures of each step in this embodiment, please refer to the description of the foregoing embodiments, which will not be repeated here.

[0086] like Figure 6 As shown, in one embodiment, an embodiment of an image reconstruction method is also provided, which includes:

[0087] S1, using 3D GRASE sequence segmentation to collect cranial ASL control data and ASL marker data;

[0088] S2, the data obtained from each segment presents a regular downsampling distribution in the K space;

[0089] S3. Using the GRAPPA method or iterative reconstruction method, the K-space data of each segment is reconstructed independently to obtain the corresponding cranial image of each segment of ASL control data and ASL labeled data.

[0090] S4, merge the brain images of each segment of the obtained ASL control data to obtain the final ASL control image;

[0091] S5, merge the brain images of each segment of the obtained ASL-labeled data to obtain the final ASL-labeled image;

[0092] S6, subtract the amplitude of the ASL labeled image from the amplitude of the ASL control image to obtain the ASL perfusion image of the brain.

[0093] The implementation principle and technical effect of each step in the image reconstruction method provided in this embodiment are similar to those in the previous image reconstruction method embodiments, and will not be repeated here. Figure 6 The implementation methods of each step in the embodiment are only examples and are not limited to any particular implementation method. The order of each step can be adjusted in actual applications, as long as the purpose of each step can be achieved.

[0094] It should be understood that, although Figures 2-6The steps in the flowchart are shown in sequence according to the arrows, but the steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, the steps are not strictly limited in sequence, and the steps can be executed in other sequences. Moreover, Figures 2-6 At least part of the steps in the flowchart can include multiple sub-steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence of the sub-steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least part of other steps or sub-steps or stages of other steps.

[0095] In one embodiment, as shown in Figure 7 An image reconstruction apparatus is provided, comprising: an acquisition module 10, a reconstruction module 11 and a merging module 12, wherein,

[0096] The acquisition module 10 is configured to acquire ASL control data and ASL marker data of a target region in K-space in multiple segments;

[0097] The reconstruction module 11 is configured to independently reconstruct the ASL control data of each segment to obtain a first complete image corresponding to the ASL control data of each segment; and independently reconstruct the ASL marker data of each segment to obtain a second complete image corresponding to the ASL marker data of each segment.

[0098] The merging module 12 is configured to merge the first complete images corresponding to each segment to obtain a reconstructed control image of the target region; and merge the second complete images corresponding to each segment to obtain a reconstructed marker image of the target region.

[0099] In one embodiment, as shown in Figure 8 The apparatus further comprises a determination module 13 configured to determine a perfusion image of the target region according to the control image and the marker image.

[0100] In one embodiment, the acquisition module 10 is specifically configured to acquire ASL control data of a target region in K-space in multiple segments according to a preset data acquisition sequence, to obtain multiple ASL control data; and acquire ASL marker data of the target region in K-space in multiple segments according to the preset data acquisition sequence, to obtain multiple ASL marker data.

[0101] In one embodiment, the ASL control data of each segment and the marker data of each segment exhibit regular down-sampling distribution in K-space.

[0102] In one embodiment, the reconstruction module 11 is specifically configured to independently reconstruct each of the ASL control data according to a preset image reconstruction method to obtain each first complete image; and independently reconstruct each of the ASL labeling data according to the preset image reconstruction method to obtain each second complete image.

[0103] In one embodiment, the merging module 12 is specifically configured to determine an image composed of mean values of pixel values at the same positions in each of the first complete images as a control image; and determine an image composed of mean values of pixel values at the same positions in each of the second complete images as a labeling image.

[0104] In one embodiment, the determining module 13 is specifically configured to subtract the amplitude of a pixel at a corresponding position in the labeling image from the amplitude of each pixel in the control image to obtain a perfusion image.

[0105] Embodiments of the present application provide a magnetic resonance image reconstruction device, which comprises:

[0106] An ASL data acquisition module is configured to acquire a plurality of segmented ASL data of a target region in K-space; the plurality of segmented ASL data is divided along a phase encoding direction;

[0107] An ASL data reconstruction module is configured to independently reconstruct each of the segmented ASL data to obtain a complete image corresponding to each of the segmented ASL control data;

[0108] A magnetic resonance image determination module is configured to merge each of the corresponding complete images of the segmented ASL data to obtain a reconstructed magnetic resonance image of the target region.

[0109] In one embodiment, the ASL data comprises ASL control data and ASL labeling data, the complete image comprises a first complete image corresponding to the ASL control data and a second complete image corresponding to the ASL labeling data, the first complete image is merged to form a control image, and the second complete image is merged to form a labeling image;

[0110] The device further comprises a perfusion image determination module configured to determine a perfusion image of the target region according to the control image and the labeling image.

[0111] For specific limitations of the image reconstruction device and the magnetic resonance image reconstruction device, refer to the limitations of the image reconstruction method and the magnetic resonance image reconstruction method in the foregoing, which will not be repeated here. Each module in the image reconstruction device and the magnetic resonance image reconstruction device can be realized by software, hardware, or a combination thereof, in whole or in part. Each module can be embedded in or independent of a processor in a computer device in hardware form, or can be stored in a memory in a computer device in software form, so as to be called and executed by a processor to perform operations corresponding to each module.

[0112] In one embodiment, a computer device is provided, which can be a terminal, and an internal structure diagram thereof can be as shown in the above Figure 1 The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is configured to communicate with external terminals through a network connection. The computer program is executed by the processor to implement an image reconstruction method. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0113] Those skilled in the art can understand that the structure shown in the above Figure 1 is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the diagram, or combine certain components, or have a different arrangement of components.

[0114] In one embodiment, a computer device is provided, which includes a memory and a processor, and the memory stores a computer program. The processor implements the following steps when executing the computer program:

[0115] obtain ASL control data and ASL label data of a plurality of segments of a target region in K-space;

[0116] independently reconstruct the ASL control data of each segment to obtain a first complete image corresponding to the ASL control data of each segment, and independently reconstruct the ASL label data of each segment to obtain a second complete image corresponding to the ASL label data of each segment;

[0117] merge the first complete images corresponding to each segment to obtain a control image of the target region after reconstruction, and merge the second complete images corresponding to each segment to obtain a label image of the target region after reconstruction.

[0118] In one embodiment, the processor implements the following steps when executing the computer program:

[0119] determine a perfusion image of the target region according to the control image and the label image.

[0120] In one embodiment, the processor implements the following steps when executing the computer program:

[0121] According to a preset data acquisition sequence, ASL control data of the target region in K space is acquired under multiple segments to obtain multiple ASL control data; and according to the preset data acquisition sequence, ASL mark data of the target region in K space is acquired under multiple segments to obtain multiple ASL mark data.

[0122] In one embodiment, the ASL control data of each segment and the mark data of each segment are regularly down-sampled in K space.

[0123] In one embodiment, the processor implements the following steps when executing the computer program:

[0124] According to a preset image reconstruction method, the ASL control data of each segment is independently reconstructed to obtain a first complete image; and according to the image reconstruction method, the ASL mark data of each segment is independently reconstructed to obtain a second complete image.

[0125] In one embodiment, the processor implements the following steps when executing the computer program:

[0126] An image composed of the mean values of the pixel values of the same positions in the first complete images is determined as a control image; and an image composed of the mean values of the pixel values of the same positions in the second complete images is determined as a mark image.

[0127] In one embodiment, the processor implements the following steps when executing the computer program:

[0128] The amplitude of each pixel in the control image is subtracted by the amplitude of the pixel at the corresponding position in the mark image to obtain a perfusion image.

[0129] In one embodiment, the processor implements the following steps when executing the computer program:

[0130] Obtaining ASL data of the target region in K space under multiple segments; the ASL data of the multiple segments is divided along the phase encoding direction;

[0131] Independently reconstructing the ASL data of each segment to obtain a complete image corresponding to the ASL control data of each segment;

[0132] Merging the corresponding complete images of each segment to obtain a reconstructed magnetic resonance image of the target region.

[0133] The computer device provided in the above embodiment has similar implementation principles and technical effects to the above method embodiments, and thus will not be described here.

[0134] In one embodiment, a computer readable storage medium is provided, having stored thereon a computer program which, when executed by a processor, implements the following steps:

[0135] obtaining ASL control data and ASL labeling data of the target region in K-space in a plurality of segments;

[0136] independently reconstructing the ASL control data of each segment to obtain a first complete image corresponding to the ASL control data of each segment respectively; and independently reconstructing the ASL labeling data of each segment to obtain a second complete image corresponding to the ASL labeling data of each segment respectively;

[0137] merging the first complete images corresponding to each segment to obtain a control image of the target region after reconstruction; and merging the second complete images corresponding to each segment to obtain a labeling image of the target region after reconstruction.

[0138] In one embodiment, the computer program, when executed by the processor, implements the following steps:

[0139] determining a perfusion image of the target region according to the control image and the labeling image.

[0140] In one embodiment, the computer program, when executed by the processor, implements the following steps:

[0141] acquiring the ASL control data of the target region in K-space in a plurality of segments according to a preset data acquisition sequence, to obtain a plurality of ASL control data; and acquiring the ASL labeling data of the target region in K-space in a plurality of segments according to the data acquisition sequence, to obtain a plurality of ASL labeling data.

[0142] In one embodiment, the ASL control data of each segment and the labeling data of each segment are regularly down-sampled in K-space.

[0143] In one embodiment, the computer program, when executed by the processor, implements the following steps:

[0144] independently reconstructing the ASL control data of each segment according to a preset image reconstruction method to obtain a first complete image; and independently reconstructing the ASL labeling data of each segment according to the preset image reconstruction method to obtain a second complete image.

[0145] In one embodiment, the computer program, when executed by the processor, implements the following steps:

[0146] determining an image composed of the mean values of the pixel values at the same positions in the first complete images as the control image; and determining an image composed of the mean values of the pixel values at the same positions in the second complete images as the labeling image.

[0147] In one embodiment, the computer program, which when executed by the processor, implements the following steps:

[0148] Subtracting the amplitude of each pixel in the control image from the amplitude of the pixel at the corresponding position in the marker image to obtain a perfusion image.

[0149] In one embodiment, the computer program, which when executed by the processor, implements the following steps:

[0150] Obtaining ASL data of the target region in multiple segments of K-space; the ASL data in multiple segments is divided along the phase encoding direction;

[0151] Independently reconstructing the ASL data in each segment to obtain a complete image corresponding to the ASL control data in each segment;

[0152] Merging the complete images corresponding to each segment to obtain a reconstructed magnetic resonance image of the target region.

[0153] The computer readable storage medium provided in the above embodiment has similar implementation principles and technical effects to the method embodiments described above, and thus will not be described here.

[0154] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by a computer program instructing related hardware, and the computer program can be stored in a non-volatile computer readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiments. Any reference to memory, storage, database or other medium used in the embodiments provided by the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0155] Any combination of the technical features in the above embodiments can be made, and for the sake of brevity, not all possible combinations are described above, however, as long as the combination of the technical features does not exist in contradiction, it shall be considered within the scope of the present disclosure.

[0156] The above embodiments only express several implementation manners of the present application, and the description is relatively specific and detailed, but it shall not be understood as a limitation on the patent scope of the present application. It shall be pointed out that, for ordinary skilled persons in the art, several modifications and improvements can be made without departing from the concept of the present application, and these shall be within the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the appended claims.

Claims

1. A method of image reconstruction, characterized by, The method comprises: acquiring ASL control data and ASL marking data of a target region in multiple segments of K space, the ASL control data of each segment and the ASL marking data of each segment being distributed in K space in a reduced sampling manner along a phase encoding direction; independently reconstructing the ASL control data of each segment to obtain multiple first complete images corresponding to the ASL control data of each segment respectively, and independently reconstructing the ASL marking data of each segment to obtain multiple second complete images corresponding to the ASL marking data of each segment respectively; determining an image composed of mean values of pixel values of the same positions in the multiple first complete images corresponding to each segment as a control image of the target region after reconstruction, and determining an image composed of mean values of pixel values of the same positions in the multiple second complete images corresponding to each segment as a marking image of the target region after reconstruction.

2. The method of claim 1, wherein, The method further comprises: determining a perfusion image of the target region according to the control image and the marking image.

3. The method according to claim 1 or 2, characterized in that, The acquiring of the ASL control data and the ASL marking data of the target region in multiple segments of K space comprises: acquiring ASL control data of the target region in multiple segments of K space according to a preset data acquisition sequence to obtain multiple ASL control data, and acquiring ASL marking data of the target region in multiple segments of K space according to the preset data acquisition sequence to obtain multiple ASL marking data.

4. The method according to claim 1 or 2, characterized in that, The independently reconstructing of the ASL control data of each segment to obtain the first complete image corresponding to the ASL control data of each segment respectively comprises: independently reconstructing the ASL control data of each segment according to a preset image reconstruction method to obtain the first complete image. The independently reconstructing of the ASL marking data of each segment to obtain the second complete image corresponding to the ASL marking data of each segment respectively comprises: independently reconstructing the ASL marking data of each segment according to the preset image reconstruction method to obtain the second complete image.

5. The method of claim 1, wherein, The combining of the first complete images corresponding to each segment to obtain the control image of the target region after reconstruction, and the combining of the second complete images corresponding to each segment to obtain the marking image of the target region after reconstruction comprise: determining an image composed of mean values of pixel values of the same positions in the first complete images as the control image, and determining an image composed of mean values of pixel values of the same positions in the second complete images as the marking image.

6. A method of magnetic resonance image reconstruction, characterized by, The method comprises: acquiring ASL data of a target region in multiple segments of K space, the ASL data of the multiple segments being distributed in a reduced sampling manner along a phase encoding direction; independently reconstructing the ASL data of each segment to obtain multiple complete images corresponding to the ASL data of each segment respectively; determining an image composed of mean values of pixel values of the same positions in the multiple complete images corresponding to each segment as a magnetic resonance image of the target region after reconstruction.

7. The method of claim 6, wherein, The ASL data includes ASL control data and ASL labeling data, the complete images include a first complete image corresponding to the ASL control data and a second complete image corresponding to the ASL labeling data, the first complete images are merged to form a control image, and the second complete images are merged to form a labeling image. The method further includes determining a perfusion image of the target region according to the control image and the labeling image.

8. An image reconstruction apparatus, characterized by comprising: The apparatus includes: An acquisition module configured to acquire ASL control data and ASL labeling data of a target region in multiple segments of K-space, the ASL control data of each segment and the ASL labeling data of each segment being distributed in a phase encoding direction in K-space in a manner of down-sampling; A reconstruction module configured to independently reconstruct the ASL control data of each segment to obtain multiple first complete images corresponding to the ASL control data of each segment respectively, and independently reconstruct the ASL labeling data of each segment to obtain multiple second complete images corresponding to the ASL labeling data of each segment respectively; A merging module configured to determine an image formed by averaging pixel values of the same positions in the multiple first complete images corresponding to each segment as a control image of the target region after reconstruction, and determine an image formed by averaging pixel values of the same positions in the multiple second complete images corresponding to each segment as a labeling image of the target region after reconstruction. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-8 when the computer program is executed by the processor. The processor executes the computer program to implement the steps of the method in any one of claims 1 to 7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method in any one of claims 1 to 7.

Images