CN117368817A - Image reconstruction method, magnetic resonance imaging method and computer device - Google Patents

Abstract

This application relates to an image reconstruction method, a magnetic resonance imaging method and a computer device. The method includes: obtaining a K-space calibration data set corresponding to a target part, and the K-space calibration data set is fully sampled in the central area of K-space; obtaining a corresponding K-space calibration data set corresponding to the target part. Multiple under-sampled K-space data sets, each under-sampled K-space data set is the data collected from the target part in one excitation; based on the K-space calibration data set, each under-sampled K-space data set in K-space is The unsampled points in the central area are subjected to the first fitting recovery to obtain multiple intermediate K-space data sets; the unsampled points of the multiple intermediate K-space data sets in the non-central area of the K space are subjected to the second fitting recovery to obtain Multiple target K-space data sets; reconstruct multiple target K-space data sets to obtain magnetic resonance images corresponding to multiple excitations of the target part. In this way, the quality of the magnetic resonance image of the target part generated based on the target K-space data set is better.

Description

Image reconstruction methods, magnetic resonance imaging methods and computer equipment

Technical field

The present application relates to the field of magnetic resonance imaging technology, and in particular to an image reconstruction method, a magnetic resonance imaging method and computer equipment.

Background technique

Parallel magnetic resonance imaging technology uses multi-element coil arrays to simultaneously collect K-space data, allowing K-space to be undersampled to reduce the number of phase encoding steps, thus greatly shortening the magnetic resonance scanning time while maintaining the same spatial resolution of the image. Improve imaging speed.

In related technologies, during the image reconstruction process based on K-space data, the collected small-scale fully sampled calibration lines (calibration lines) of the central area of K-space are used as a restoration benchmark for unsampled data, and then the complete K-space data is synthesized. . In order to speed up the acquisition process or due to constraints such as sequence design, one calibration data is often shared when reconstructing multiple magnetic resonance images.

However, the above method will cause artifacts in the reconstructed image during image reconstruction, and the reconstructed image quality is poor.

Contents of the invention

Based on this, it is necessary to address the above technical problems and provide an image reconstruction method, a magnetic resonance imaging method and a computer device that can improve the quality of reconstructed images in parallel magnetic resonance imaging.

In a first aspect, this application provides an image reconstruction method, which method includes:

Obtain the K-space calibration data set corresponding to the target part. The K-space calibration data set is fully sampled in the central area of K-space;

Obtain multiple under-sampled K-space data sets corresponding to the target part. Each under-sampled K-space data set is the data collected by the target part in one excitation;

Based on the K-space calibration data set, perform first fitting recovery on the unsampled points of each under-sampled K-space data set in the central area of K-space to obtain multiple intermediate K-space data sets;

Perform second fitting recovery on the unsampled points of multiple intermediate K-space data sets in the non-center area of K-space to obtain multiple target K-space data sets;

Reconstruct multiple target K-space data sets and obtain magnetic resonance images corresponding to multiple excitations of the target site.

In one embodiment, based on the K-space calibration data set, the first fitting recovery is performed on the unsampled points in the central area of the K-space of each under-sampled K-space data set, including:

For any undersampled K-space data set, construct a data recovery matrix based on the K-space calibration data set and the under-sampled K-space data set;

Based on the data recovery matrix corresponding to each undersampled K-space data set, a first fitting recovery is performed on the unsampled points of each sampled K-space data set in the central area of K-space.

In one embodiment, a data recovery matrix is constructed based on the K-space calibration data set and the undersampled K-space data set, including:

Based on the K-space calibration data set, the first low-rank matrix is constructed using the preset low-rank matrix construction method;

Based on the undersampled K-space data set, the second low-rank matrix is constructed using the low-rank matrix construction method;

A data recovery matrix is generated based on the first low-rank matrix and the second low-rank matrix.

In one embodiment, the process of constructing a low-rank matrix includes:

Extract a preset number of different first data points from the target data set, and obtain the coordinate information of each first data point; the target data set is a K-space calibration data set or an undersampled K-space data set;

For any first data point, obtain a plurality of second data points whose distance from the first data point is less than a preset length, and obtain a set of data points corresponding to the first data point;

Obtain the signal values of multiple second data points in each data point set;

A low-rank matrix is constructed based on the coordinate information of each first data point and the signal values of multiple second data points in the data point set corresponding to each first data point.

In one embodiment, performing a second fitting recovery on unsampled points of multiple intermediate K-space data sets in non-central areas of K-space includes:

Based on the fitted fully sampled data of each intermediate K-space data set in the central area of K-space, calculate the weight kernel of the unsampled points of each intermediate K-space data set in the non-central area of K-space;

For any intermediate K-space data set, the unsampled points in the non-center area are restored by second fitting according to the weight kernel of the unsampled points.

In a second aspect, this application also provides a magnetic resonance imaging method, which method includes:

Use partial sampling technology to fill the central area of K-space to obtain the K-space calibration data set corresponding to the target part;

Excite the target part multiple times, and collect the undersampled K-space data set corresponding to each excitation;

Based on the K-space calibration data set, perform first fitting recovery on the unsampled points of each under-sampled K-space data set in the central area of K-space to obtain multiple intermediate K-space data sets;

Perform second fitting recovery on the unsampled points of multiple intermediate K-space data sets in the non-center area of K-space to obtain multiple target K-space data sets

Reconstruct multiple target K-space data sets and obtain magnetic resonance images corresponding to multiple excitations of the target site.

In one embodiment, the direction of the diffusion gradient applied in each of the multiple excitations is different.

In one embodiment, the physiological period of the target site corresponding to each of the multiple excitations is different.

In one embodiment, a labeling pulse is applied to the target site for each of multiple excitations, and undersampled K-space data sets are collected at different delay times after the labeling pulse is applied.

In a third aspect, this application also provides an image reconstruction device, which includes:

The calibration data acquisition module is used to obtain the K-space calibration data set corresponding to the target part. The K-space calibration data set is fully sampled in the central area of K-space;

The undersampled data acquisition module is used to obtain multiple undersampled K-space data sets corresponding to the target part. Each undersampled K-space data set is the data collected by the target part in one excitation;

The first data recovery module is used to perform first fitting recovery on the unsampled points of each undersampled K-space data set in the central area of K-space based on the K-space calibration data set, so as to obtain multiple intermediate K-space data. set;

The second data recovery module is used to perform a second fitting recovery on the unsampled points of multiple intermediate K-space data sets in the non-center area of K-space, and obtain multiple target K-space data sets;

The image reconstruction module is used to reconstruct multiple target K-space data sets and obtain magnetic resonance images corresponding to multiple excitations of the target part.

In a fourth aspect, this application also provides a magnetic resonance imaging device, which includes:

The data acquisition module is used to fill the central area of K-space using partial sampling technology and obtain the K-space calibration data set corresponding to the target part;

The scanning module is used to excite the target part multiple times and collect the undersampled K-space data set corresponding to each excitation;

The first data recovery module is used to perform first fitting recovery on the unsampled points of each undersampled K-space data set in the central area of K-space based on the K-space calibration data set, so as to obtain multiple intermediate K-space data. set;

The second data recovery module is used to perform a second fitting recovery on the unsampled points of multiple intermediate K-space data sets in the non-center area of K-space, and obtain multiple target K-space data sets.

The imaging module is used to reconstruct multiple target K-space data sets and obtain magnetic resonance images corresponding to multiple excitations of the target site.

In a fifth aspect, the application also provides a computer device. The computer device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, any one of the method embodiments of the first aspect and the second aspect is implemented. A step of.

In a sixth aspect, the present application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program. When the computer program is executed by a processor, any one of the methods of the first aspect and the second aspect can be implemented. Example steps.

In a seventh aspect, the present application also provides a computer program product. The computer program product includes a computer program. When the computer program is executed by a processor, the steps of any one of the method embodiments of the first and second aspects are implemented.

In the above image reconstruction method, magnetic resonance imaging method and computer equipment, first, a K-space calibration data set corresponding to the target part and multiple undersampled K-space data sets corresponding to the target part are obtained. Among them, the K-space calibration data set is fully sampled in the central area of K-space, and each under-sampled K-space data set is the data collected from the target part in one excitation. Then, based on the K-space calibration data set, the first fitting recovery is performed on the unsampled points in the central area of the K-space of each under-sampled K-space data set to obtain multiple intermediate K-space data sets. Further, a second fitting recovery is performed on the unsampled points of the multiple intermediate K-space data sets in the non-center area of the K-space to obtain multiple target K-space data sets. Finally, multiple target K-space data sets are reconstructed to obtain magnetic resonance images corresponding to multiple excitations of the target part. That is to say, this application does not directly use the K-space calibration data set as the benchmark to fit and restore the unsampled data in the under-sampled K-space data set. Instead, based on the K-space calibration data set, the first fitting recovery is performed on the unsampled points in the center area of each undersampled K-space data set. Through the first fitting recovery, the fully sampled data in the center area can be obtained. Further, using the first fitting recovery to obtain the fully sampled data in the central area as a benchmark, the second fitting recovery is performed on the unsampled data points in the non-central area to obtain the target K-space data set. In this way, using the intermediate K-space data set obtained after the first fitting recovery of the unsampled points in the central area of K-space as the benchmark can ensure that the fully-sampled data of the intermediate K-space data set in the central area of K-space is the same as the under-sampled data. The unsampled points of the K-space data set in the non-center area of K-space belong to the data obtained by the same excitation, and the data matching degree is high. The target K-space data set after the second fitting recovery is also more consistent with the actual data obtained by full sampling. situation, and the quality of the magnetic resonance image of the target part generated based on the target K-space data set is also better.

Description of the drawings

Figure 1 is a schematic flowchart of an image reconstruction method in one embodiment;

Figure 2 is a schematic diagram of a K-space standard data set in one embodiment;

Figure 3 is a schematic diagram of data fitting and restoration of the central area of the undersampled K-space data set in one embodiment;

Figure 4 is a schematic flow chart of the first fitting recovery operation in one embodiment;

Figure 5 is a schematic diagram of constructing a data recovery matrix in one embodiment;

Figure 6 is a schematic flowchart of the second fitting recovery operation in one embodiment;

Figure 7 is a schematic flow chart of a magnetic resonance imaging method in one embodiment;

Figure 8 is a structural block diagram of an image reconstruction device in one embodiment;

Figure 9 is a structural block diagram of a magnetic resonance imaging device in one embodiment;

Figure 10 is an internal structure diagram of a computer device in one embodiment.

Detailed ways

In order to make the purpose, technical solutions and advantages of the present application more clear, the present application will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present application and are not used to limit the present application.

Magnetic resonance imaging (MRI) is a technology that uses the principle of nuclear magnetic resonance to perform tomographic imaging of the human body. It can provide various images of human soft tissues and has rapidly developed into an important application technology in biomedicine. There is no radioactive contamination during the imaging process, the resolution is high, and imaging can be performed at any level. Moreover, unlike various existing imaging technologies, magnetic resonance imaging involves many factors and the resulting image information is large, which has great advantages and application potential in medical diagnosis.

However, in some clinical application fields, in addition to requiring high image spatial resolution, it is also necessary to reduce imaging time to reduce motion artifacts. For example, in areas such as real-time imaging of the cardiovascular system and functional brain imaging, the influence of factors such as heart movement, respiration, and blood flow must be considered. Magnetic resonance imaging encodes the spatial information of Fourier images through gradient fields. Obtaining a complete image requires continuous gradient spatial encoding. The imaging speed greatly depends on the performance of the gradient system of the magnetic resonance equipment, such as the strength and switching rate of the gradient field. In order to meet the needs of fast imaging, the performance of gradient fields has been greatly enhanced. At the same time, new problems have also arisen. The cost of gradient hardware systems is getting higher and higher. In addition, too high gradient field switching rate can cause neuromuscular electromagnetic stimulation. The improvement of imaging speed that relies on gradient field performance has reached its limit, and more effective methods are expected to be provided clinically to improve imaging speed.

In order to improve the imaging speed, magnetic resonance parallel imaging (Parallel MRI) uses multiple phased array coils to receive induction signals at the same time, which can reduce the number of gradient encodings, thereby greatly shortening the scanning time and increasing the imaging speed.

Specifically, parallel magnetic resonance imaging uses small-scale full sampling calibration data (calibration lines) collected in the central area of K space as a benchmark for unsampled data recovery, and calculates a weighting kernel that can fit the unsampled data. , and in the next step of data synthesis, the weight kernel is applied to the undersampled data to be reconstructed to synthesize complete K-space data.

During this process, accurate full-sample calibration data is critical to imaging quality. If the collected full-sampling calibration data does not match the under-sampled data to be reconstructed, or the quality of the full-sampling calibration data is poor, errors will occur in the calculated weight kernel and the errors will be introduced into the subsequent data synthesis stage, making the reconstruction The image may suffer from image quality degradation such as artifacts or blur.

In some sequence scanning applications, in order to speed up the data collection or due to constraints such as sequence design, it is often necessary to share a full sampling calibration data when reconstructing multiple magnetic resonance images. Due to the movement of the scanned object during the scanning process or the influence of other factors, there will be a deviation between the above-mentioned full-sampled calibration data and the scanned under-sampled data, thus affecting the quality of the reconstructed image.

Based on this, this application provides an image reconstruction method. Based on the full calibration data obtained before scanning, the corresponding target calibration data is constructed for the under-sampled K-space data set collected by scanning the target part of the scanning object each time. Use the target calibration data corresponding to each undersampled K-space data set to restore the unsampled data to obtain complete K-space data, so that the quality of the reconstructed image is better.

The image reconstruction method provided by this application can be applied to an image reconstruction device. The image reconstruction device can be implemented in the form of software and/or hardware. The device can be integrated in a computer device with medical image processing functions, such as magnetic resonance imaging. The imaging equipment in the system, or any computer equipment outside the magnetic resonance system.

Among them, the imaging equipment in the magnetic resonance system is used to fill the magnetic resonance signals into K space, and perform image reconstruction based on the K space data to obtain the target magnetic resonance image. The computer equipment outside the magnetic resonance system can be any terminal or server, where the terminal can include but is not limited to software running in the physical device, such as an application or client installed on the device, and can also include but not Restricted to PCs, laptops, smartphones, tablets and portable wearable devices with the app installed. Servers may include, but are not limited to, at least one independent server, distributed server, cloud server and server cluster.

The technical solutions of the embodiments of the present application and how the technical solutions of the embodiments of the present application solve the above technical problems will be described in detail below through the embodiments and in conjunction with the accompanying drawings. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. It should be noted that, for the image reconstruction method provided by the embodiments of the present application, the execution subject may be a magnetic resonance imaging device, a computer device, or the image reconstruction device provided by the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments.

In one embodiment, as shown in Figure 1, an image reconstruction method is provided. Taking the method as applied to a computer device as an example, the method includes the following steps:

Step 110: Obtain the K-space calibration data set corresponding to the target part. The K-space calibration data set is fully sampled in the central area of K-space.

The target part can be any part of the scanning object to be detected. If the scanning object is a human body, the target part can be the head, chest, abdomen, etc. Of course, the scanning object can also be other living organisms, which is not limited in this embodiment.

It should be noted that the K-space calibration data set can be obtained during the pre-scanning process, during the positioning process, or during the sequence scanning process. This embodiment does not limit the acquisition timing.

Furthermore, for image reconstruction, reducing the scanning time is usually achieved by reducing the number of adopted points. Therefore, when obtaining the K-space calibration data set in the embodiment of the present application, it can be achieved through partial sampling technology, and only for decision-making The central region of the contrasting K-space of the image is fully sampled. For the non-central area of K-space, that is, the peripheral area, it can be under-sampled, as shown in (a) in Figure 2; or it can not be sampled, as shown in (b) in Figure 2.

It should be understood that (b) in Figure 2 is only a data set obtained by undersampling in an interlaced manner in a non-central area of K-space. The undersampling interval can still be multiple lines, and there is no restriction on this.

Step 120: Obtain multiple under-sampled K-space data sets corresponding to the target part. Each under-sampled K-space data set is data collected from the target part in one excitation.

Among them, multiple undersampled K-space data sets are acquired in dynamic or multi-phase scans. For example, in blood oxygen level dependent functional imaging (BOLD fMRI) applications, due to data acquisition time constraints and echo planar imaging (EchoPlanar Imaging, EPI) phase encoding requirements, K-space calibration data sets are pre-collected during parallel imaging , and then collect multiple undersampled K-space data sets.

Step 130: Based on the K-space calibration data set, perform first fitting recovery on the unsampled points of each under-sampled K-space data set in the central area of the K-space, to obtain multiple intermediate K-space data sets.

Among them, compared with the undersampled K-space data set, the intermediate K-space data set restored by the first fitting is equivalent to full sampling in the central region of K-space.

That is, through the fully sampled data of the K-space calibration data set in the central area of K-space, the first fitting recovery is performed on the unsampled points of the under-sampled K-space data set in the central area of K-space, and the unsampled point data values to obtain the complete sampled data of the undersampled K-space data set in the central area of K-space.

In a possible implementation, the first fitting recovery can take advantage of the low rank property by constructing a low rank matrix and obtaining the zero space matrix to determine each undersampled K-space data The intermediate K-space data set corresponding to the set. This intermediate K-space data set serves as the benchmark for recovering unsampled data points in the under-sampled K-space data set.

It should be noted that the central region position/size of the K-space standard data set and each undersampled K-space data set is the same. In other words, based on the position of the K-space center area of the fully sampled K-space calibration data set, the first fitting recovery is performed on the unsampled points at the same position in each sampled K-space data set.

As an example, see Figure 3. The size of the central area of the K-space standard data set in K-space is 3*3. Then the data range of the first fitting recovery performed on the under-sampled K-space data set is also in the central area of K-space. Unsampled points in the 3*3 range.

Step 140: Perform a second fitting recovery on the unsampled points of the multiple intermediate K-space data sets in the non-center area of the K-space, and obtain multiple target K-space data sets.

It should be understood that the phase encoding lines filling the central area of K-space mainly determine the contrast of the image, while the phase encoding lines in the peripheral area mainly determine the anatomical details of the image. The phase encoding lines on both sides of the zero Fourier line are mirror symmetrical. That is to say, the K-space phase encoding direction and the frequency encoding direction exhibit mirror symmetry characteristics.

Based on this, after recovering all the data of the under-sampled K-space data set in the central area of K-space through the first fitting, the full sampling can be obtained through the second fitting recovery based on the mirror symmetry characteristics of the K-space filling data. case of target K-space dataset.

In a possible implementation, when performing image reconstruction based on the K-space domain, the first fitting recovery can use the Simultaneous Acquisition of Spatial Harmonics (SMASH) reconstruction method or the global automatic calibration partial parallel acquisition (Generalized Auto-calibrating Partially ParallelAcquisition, GRAPPA) reconstruction method, Sensitivity Encoding (Sensitivity Encoding, SENSE) reconstruction method to achieve.

Specifically, the basic idea of the SMASH reconstruction method is to recover the K-space phase-encoded line data lost due to undersampling through a linear combination of the receiving coil sensitivity. GRAPPA no longer fits the data of each array coil to the combined signal, but to the ACS row of a single coil, thereby obtaining a series of linear weights to reconstruct the missing K-space data rows of each array coil.

Step 150: Reconstruct multiple target K-space data sets and obtain magnetic resonance images corresponding to multiple excitations of the target part.

In this step, image reconstruction is performed on each target K-space data set to obtain the corresponding magnetic resonance image. Specifically, image reconstruction is to perform Fourier transform on the target K-space data set to generate a reconstructed image. The reconstructed image of multiple target K-space data sets is the magnetic resonance image obtained after multiple scans of the scanned part.

In the above image reconstruction method, according to the K-space calibration data set, the unsampled points in the central area of each under-sampled K-space data set are first fitted and restored. Through the first fitting and restoration, the full range of the central area can be obtained. Sample data. Further, using the first fitting recovery to obtain the fully sampled data in the central area as a benchmark, the second fitting recovery is performed on the unsampled data points in the non-central area to obtain the target K-space data set. In this way, using the intermediate K-space data set obtained after the first fitting recovery of the unsampled points in the central area of K-space as the benchmark can ensure that the fully-sampled data of the intermediate K-space data set in the central area of K-space is the same as the under-sampled data. The unsampled points of the K-space data set in the non-center area of K-space belong to the data obtained by the same excitation, and the data matching degree is high. The target K-space data set after the second fitting recovery is also more consistent with the actual data obtained by full sampling. situation, and the quality of the magnetic resonance image of the target part generated based on the target K-space data set is also better.

In one embodiment, as shown in Figure 4, in the above step 130, based on the K-space calibration data set, the first fitting recovery is performed on the unsampled points in the central area of the K-space of each under-sampled K-space data set, Includes the following steps:

Step 410: For any undersampled K-space data set, construct a data recovery matrix based on the K-space calibration data set and the under-sampled K-space data set.

In a possible implementation, the implementation process of step 410 may be: based on the K-space calibration data set, using a preset low-rank matrix construction method to construct the first low-rank matrix; based on the under-sampled K-space data set, using a low-rank matrix The rank matrix construction method constructs the second low-rank matrix; based on the first low-rank matrix and the second low-rank matrix, a data recovery matrix is generated.

Among them, the process of constructing the low-rank matrix is: extract a preset number of different first data points from the target data set, and obtain the coordinate information of each first data point; for any first data point, obtain the same data as the first data point For multiple second data points whose point distance is less than the preset length, obtain the data point set corresponding to the first data point; obtain the signal values of multiple second data points in each data point set; according to the coordinate information of each first data point , and the signal values of multiple second data points in the data point set corresponding to each first data point, to construct a low-rank matrix.

That is, the first low-rank matrix and the second low-rank matrix are constructed based on the same low-rank matrix construction method, and the above target data set is a K-space calibration data set or an undersampled K-space data set.

As an example, the construction operation of a low-rank matrix is denoted as P _c (·), and the construction steps are:

(1) Randomly select L different first data points from the K-space center area of the target data set, k represents the coding index of the selected first data point, 1≤k≤L, n _x , n _y represent respectively The abscissa and ordinate of a certain first data point, then the signal values of these first data points can be expressed as

Further, for each first data point selected in the target data set, select the A set of other N _R second data points that are within a radius R. The coordinate index of these N _R second data points is /> express.

Among them, m=1, 2,..., N _R . The corresponding signal value is Usually _L≥NR .

(2) Arrange according to the following formula (1) to form a low-rank C matrix:

In the formula, k=1, 2,..., L, m=1, 2,..., N _R , Represents the abscissa coordinate of the m-th point nearest neighbor of the k-th point,/> Represents the ordinate of the m-th point that is the nearest neighbor of the k-th point. Then the size of the C(k,m) matrix is L× _NR .

Referring to Figure 5, the process of generating the data recovery matrix in step 410 can be: from the partial data k _acs of the K-space calibration data set in the central area of K-space, construct the first low-rank matrix P _c ( k _acs ); at the same time, the data k _i of the corresponding position is taken from the undersampled K-space data set Q _i , and the second low-rank matrix P _c (k _i ) is constructed through the above-mentioned low-rank matrix construction method; and then according to the first low-rank matrix matrix and the second low-rank matrix, generating the data recovery matrix [P _c (k _acs )P _c (k _i )].

Among them, the under-sampled K-space data set Q _i is any one of multiple under-sampled K-space data sets.

Step 420: Based on the data recovery matrix corresponding to each undersampled K-space data set, perform first fitting recovery on the unsampled points of each sampled K-space data set in the central area of K-space.

In a possible implementation, the implementation process of step 420 can be: finding the zero space matrix N of the data recovery matrix, and N ^H N=I, which means that the collected K-space calibration data set is placed at the center of K-space The fully sampled data of the region and the data corresponding to the Qi _-th under-sampled K-space data are combined to estimate the data of the unsampled points in the central area of K-space for each under-sampled K-space data set.

Specifically, the problem that needs to be solved is:

Among them, η represents the proportion of full-sampled data in the K-space calibration data set. The larger η means, the fully-sampled data in the center area of K-space of the under-sampled K-space data set recovered by fitting is different from the full-sampled data in the K-space calibration data set. The more similar the data is.

In this embodiment, through the K-space calibration data set, the first fitting recovery is performed on the unsampled points of each under-sampled K-space data set in the central area of the K-space, so as to obtain multiple intermediate K-space data sets. In this way, the problem of mismatch between the K-space calibration data set and the under-sampled K-space data set can be avoided. By determining the intermediate K-space data set corresponding to each under-sampled K-space data set when the central region of K-space is fully sampled, it is possible to Alleviating possible artifacts or degradation of reconstructed image quality.

Based on the above method embodiments, through the K-space calibration data set, the first fitting recovery is performed on the unsampled points of each under-sampled K-space data set in the central area of the K-space, and multiple intermediate K-space data sets are obtained. Among them, the intermediate K-space data set is obtained by supplementing the data of unsampled points in the central area of K-space with the corresponding under-sampled K-space data set, achieving full sampling of the intermediate K-space data set in the central area of K-space.

Furthermore, for each intermediate K-space data set, based on the complete data in the central area of K-space, the unsampled points in the non-central area of K-space can be fitted and restored.

In one embodiment, as shown in Figure 6, performing a second fitting recovery on the unsampled points of the multiple intermediate K-space data sets in the non-center area of K-space in the above step 140 includes the following steps:

Step 610: Based on the fitted fully sampled data of each intermediate K-space data set in the central area of K-space, calculate the weight kernel of the unsampled points of each intermediate K-space data set in the non-central area of K-space.

Among them, for the unsampled points in the non-central area of K-space, the weight of each sampling point in the central area can be calculated based on the full sampling data of the central area when restoring the unsampled points in the non-central area. In other words, it is necessary to calculate the contribution value of each fully sampled data in the central area of K-space to the recovery of unsampled points in the non-central area.

As an example, the weight kernel can be calculated using the GRAPPA reconstruction method, which will not be described again here.

Step 620: For any intermediate K-space data set, perform a second fitting recovery on the unsampled points in the non-center area according to the weight kernel of the unsampled points.

That is to say, the weight kernel calculated based on the fitted fully sampled data of each intermediate K-space data set in the central area of K-space is applied to the unsampled points in the non-central area, and the complete K-space can be obtained through the GRAPPA reconstruction method. data, which improves the data quality of the target K-space data set obtained after the second fitting recovery, and better suppresses possible artifacts in the image.

Based on the same inventive concept, this application also provides a magnetic resonance imaging method, as shown in Figure 7. This method is explained by taking the application of this method to a computer device as an example. The computer device can be specifically one of the magnetic resonance imaging systems. or multiple devices, including the following steps:

Step 710: Use partial sampling technology to fill the central area of K-space to obtain the K-space calibration data set corresponding to the target part.

That is, before using the set number of excitations to scan the target site multiple times, the K-space calibration data set is obtained during the pre-scan or positioning process.

It should be understood that each magnetic resonance signal collected during the pre-scan process contains full-layer information. Therefore, the magnetic resonance signal needs to be spatially positioned and encoded, that is, frequency encoding and phase encoding. The magnetic resonance signals collected by the receiving coil in the magnetic resonance scanning equipment are actually radio waves with spatially encoded information. They are analog signals rather than digital information. They need to be converted into digital information through analog-to-digital conversion (ADC), and the latter is filled in. to K space, and obtain the K space digital lattice.

In this step, obtaining the K-space calibration data set through under-sampling means fully sampling the central area of K-space and under-sampling the non-central area.

Step 720: Excite the target part multiple times, and collect an undersampled K-space data set corresponding to each excitation.

The number of excitations is determined by a preset number of sequences, and this embodiment does not limit the number of excitations. Each excitation results in an undersampled K-space data set that can be used to generate an MRI image.

Step 730: Based on the K-space calibration data set, perform first fitting recovery on the unsampled points of each under-sampled K-space data set in the central area of the K-space to obtain multiple intermediate K-space data sets.

Step 740: Perform a second fitting recovery on the unsampled points of the multiple intermediate K-space data sets in the non-center area of the K-space, and obtain multiple target K-space data sets.

Step 750: Reconstruct multiple target K-space data sets and obtain magnetic resonance images corresponding to multiple excitations of the target part.

Based on the above magnetic resonance imaging method, multiple excitations of the target site can include any of the following settings:

(1) The direction of the diffusion gradient applied to each excitation in multiple excitations is different.

This setting can be used for magnetic resonance imaging equipment to perform diffusion tensor imaging (DiffusionTensor Imaging, DTI) and diffusion-weighted imaging (Diffusion-Weighted Imaging, DWI) of the target site.

(2) The physiological period of the target part corresponding to each of the multiple excitations is different.

(3) Apply a labeling pulse to the target site for each of multiple excitations, and collect undersampled K-space data sets at different delay times after the labeling pulse is applied.

This setting can be used by magnetic resonance imaging equipment to perform Arterial Spin Labeling (ASL) on the target site.

It should be noted that the implementation principles and technical effects of the image reconstruction step in the magnetic resonance imaging method provided by this embodiment are similar to the previous image reconstruction method embodiments. For specific limitations and explanations, please refer to the previous method embodiments. Here No longer.

It should be understood that although the steps in the flowcharts involved in the above-mentioned embodiments are shown in sequence as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated in this article, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least some of the steps in the flowcharts involved in the above embodiments may include multiple steps or stages. These steps or stages are not necessarily executed at the same time, but may be completed at different times. The execution order of these steps or stages is not necessarily sequential, but may be performed in turn or alternately with other steps or at least part of the steps or stages in other steps.

Based on the same inventive concept, embodiments of the present application also provide an image reconstruction device for implementing the above-mentioned image reconstruction method. The solution to the problem provided by this device is similar to the solution described in the above method. Therefore, for the specific limitations in one or more image reconstruction device embodiments provided below, please refer to the above limitations on the image reconstruction method. I won’t go into details here.

In one embodiment, as shown in Figure 8, an image reconstruction device is provided. The device 800 includes: a calibration data acquisition module 810, an undersampling data acquisition module 820, a first data recovery module 830, and a second data recovery module. 840 and image reconstruction module 850, wherein:

The calibration data acquisition module 810 is used to acquire the K-space calibration data set corresponding to the target part. The K-space calibration data set is fully sampled in the central area of K-space;

The under-sampled data acquisition module 820 is used to acquire multiple under-sampled K-space data sets corresponding to the target part. Each under-sampled K-space data set is the data collected by the target part in one excitation;

The first data recovery module 830 is configured to perform first fitting recovery on the unsampled points of each undersampled K-space data set in the central area of the K-space based on the K-space calibration data set, so as to obtain multiple intermediate K-spaces. data set;

The second data recovery module 840 is used to perform a second fitting recovery on the unsampled points of multiple intermediate K-space data sets in the non-center area of K-space, and obtain multiple target K-space data sets;

The image reconstruction module 850 is used to reconstruct multiple target K-space data sets and obtain magnetic resonance images corresponding to multiple excitations of the target part.

In one embodiment, the first data recovery module 830 includes:

A matrix construction unit used to construct a data recovery matrix for any undersampled K-space data set based on the K-space calibration data set and the under-sampled K-space data set;

The first recovery unit is used to perform first fitting recovery on the unsampled points of each sampled K-space data set in the central area of K-space based on the data recovery matrix corresponding to each under-sampled K-space data set.

In one embodiment, the matrix building unit includes:

The first construction subunit is used to construct the first low-rank matrix based on the K-space calibration data set using a preset low-rank matrix construction method;

The second construction subunit is used to construct a second low-rank matrix based on the under-sampled K-space data set using a low-rank matrix construction method;

The matrix construction subunit is used to generate a data recovery matrix based on the first low-rank matrix and the second low-rank matrix.

In one embodiment, the process of constructing a low-rank matrix includes:

Extract a preset number of different first data points from the target data set, and obtain the coordinate information of each first data point; the target data set is a K-space calibration data set or an undersampled K-space data set;

For any first data point, obtain a plurality of second data points whose distance from the first data point is less than a preset length, and obtain a set of data points corresponding to the first data point;

Obtain the signal values of multiple second data points in each data point set;

A low-rank matrix is constructed based on the coordinate information of each first data point and the signal values of multiple second data points in the data point set corresponding to each first data point.

In one embodiment, the second data recovery module 840 includes:

The weight calculation unit is used to calculate the weight kernel of the unsampled points of each intermediate K-space data set in the non-center area of K-space based on the fitted fully sampled data of each intermediate K-space data set in the central area of K-space;

The second recovery unit is used for performing a second fitting recovery on the unsampled points in the non-center area according to the weight kernel of the unsampled points for any intermediate K-space data set.

Each module in the above-mentioned image reconstruction device 800 can be implemented in whole or in part by software, hardware, and combinations thereof. Each of the above modules may be embedded in or independent of the processor of the computer device in the form of hardware, or may be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

Based on the same inventive concept, embodiments of the present application also provide a magnetic resonance imaging device for implementing the above-mentioned magnetic resonance imaging method. The solution to the problem provided by this device is similar to the solution described in the above method. Therefore, for specific limitations in one or more embodiments of the magnetic resonance imaging device provided below, please refer to the above description of the magnetic resonance imaging method. Limitations will not be repeated here.

In one embodiment, as shown in Figure 9, a magnetic resonance imaging device is provided. The device 900 includes: a data acquisition module 910, a scanning module 920, a first data recovery module 930, a second data recovery module 940 and an imaging device. Module 950, which:

The data acquisition module 910 is used to fill the central area of K-space using partial sampling technology and obtain the K-space calibration data set corresponding to the target part;

The scanning module 920 is used to excite the target part multiple times and collect the undersampled K-space data set corresponding to each excitation;

The first data recovery module 930 is configured to perform first fitting recovery on the unsampled points of each undersampled K-space data set in the central area of the K-space based on the K-space calibration data set, so as to obtain multiple intermediate K-spaces. data set;

The second data recovery module 940 is used to perform a second fitting recovery on the unsampled points of multiple intermediate K-space data sets in the non-center area of K-space, and obtain multiple target K-space data sets.

The imaging module 950 is used to reconstruct multiple target K-space data sets and obtain magnetic resonance images corresponding to multiple excitations of the target site.

In one embodiment, the direction of the diffusion gradient applied in each of the multiple excitations is different.

In one embodiment, the physiological period of the target site corresponding to each of the multiple excitations is different.

In one embodiment, a labeling pulse is applied to the target site for each of multiple excitations, and undersampled K-space data sets are collected at different delay times after the labeling pulse is applied.

Each module in the above-mentioned magnetic resonance imaging apparatus 900 can be implemented in whole or in part by software, hardware, and combinations thereof. Each of the above modules may be embedded in or independent of the processor of the computer device in the form of hardware, or may be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

In one embodiment, a computer device is provided. The computer device may be an image reconstruction device, a magnetic resonance imaging device, or other terminal for generating magnetic resonance images. Its internal structure diagram may be as follows: As shown in Figure 10. The computer device includes a processor, memory, communication interface, display screen and input device connected through a system bus. Wherein, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes non-volatile storage media and internal memory. The non-volatile storage medium stores operating systems and computer programs. This internal memory provides an environment for the execution of operating systems and computer programs in non-volatile storage media. The communication interface of the computer device is used for wired or wireless communication with external terminals. The wireless mode can be implemented through WIFI, operator network, NFC (Near Field Communication) or other technologies. When the computer program is executed by the processor, it implements the image reconstruction method and/or the magnetic resonance imaging method provided by this application. The display screen of the computer device may be a liquid crystal display or an electronic ink display. The input device of the computer device may be a touch layer covered on the display screen, or may be a button, trackball or touch pad provided on the computer device shell. , it can also be an external keyboard, trackpad or mouse, etc.

Those skilled in the art can understand that the structure shown in Figure 10 is only a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer equipment to which the solution of the present application is applied. Specific computer equipment can May include more or fewer parts than shown, or combine certain parts, or have a different arrangement of parts.

In one embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory. When the processor executes the computer program, it implements the following steps:

Obtain the K-space calibration data set corresponding to the target part. The K-space calibration data set is fully sampled in the central area of K-space;

Obtain multiple under-sampled K-space data sets corresponding to the target part. Each under-sampled K-space data set is the data collected by the target part in one excitation;

Based on the K-space calibration data set, perform first fitting recovery on the unsampled points of each under-sampled K-space data set in the central area of K-space to obtain multiple intermediate K-space data sets;

Perform second fitting recovery on the unsampled points of multiple intermediate K-space data sets in the non-center area of K-space to obtain multiple target K-space data sets;

Reconstruct multiple target K-space data sets and obtain magnetic resonance images corresponding to multiple excitations of the target site.

In another embodiment, the processor also performs the following steps when executing the computer program:

Use partial sampling technology to fill the central area of K-space to obtain the K-space calibration data set corresponding to the target part;

Excite the target part multiple times, and collect the undersampled K-space data set corresponding to each excitation;

Based on the K-space calibration data set, perform first fitting recovery on the unsampled points of each under-sampled K-space data set in the central area of K-space to obtain multiple intermediate K-space data sets;

Perform second fitting recovery on the unsampled points of multiple intermediate K-space data sets in the non-center area of K-space to obtain multiple target K-space data sets

Reconstruct multiple target K-space data sets and obtain magnetic resonance images corresponding to multiple excitations of the target site.

The implementation principles and technical effects of the computer device provided by the above embodiment are similar to those of the above method embodiment, and will not be described again here.

In one embodiment, a computer-readable storage medium is provided with a computer program stored thereon. When the computer program is executed by a processor, the following steps are implemented:

Obtain the K-space calibration data set corresponding to the target part. The K-space calibration data set is fully sampled in the central area of K-space;

Obtain multiple under-sampled K-space data sets corresponding to the target part. Each under-sampled K-space data set is the data collected by the target part in one excitation;

Based on the K-space calibration data set, perform first fitting recovery on the unsampled points of each under-sampled K-space data set in the central area of K-space to obtain multiple intermediate K-space data sets;

Perform second fitting recovery on the unsampled points of multiple intermediate K-space data sets in the non-center area of K-space to obtain multiple target K-space data sets;

Reconstruct multiple target K-space data sets and obtain magnetic resonance images corresponding to multiple excitations of the target site.

In another embodiment, the processor also performs the following steps when executing the computer program:

Use partial sampling technology to fill the central area of K-space to obtain the K-space calibration data set corresponding to the target part;

Excite the target part multiple times, and collect the undersampled K-space data set corresponding to each excitation;

Based on the K-space calibration data set, perform first fitting recovery on the unsampled points of each under-sampled K-space data set in the central area of K-space to obtain multiple intermediate K-space data sets;

Perform second fitting recovery on the unsampled points of multiple intermediate K-space data sets in the non-center area of K-space to obtain multiple target K-space data sets

Reconstruct multiple target K-space data sets and obtain magnetic resonance images corresponding to multiple excitations of the target site.

The implementation principles and technical effects of the computer-readable storage medium provided by the above embodiments are similar to those of the above method embodiments, and will not be described again here.

In one embodiment, a computer program product is provided, comprising a computer program that when executed by a processor implements the following steps:

Obtain the K-space calibration data set corresponding to the target part. The K-space calibration data set is fully sampled in the central area of K-space;

Obtain multiple under-sampled K-space data sets corresponding to the target part. Each under-sampled K-space data set is the data collected by the target part in one excitation;

Based on the K-space calibration data set, perform first fitting recovery on the unsampled points of each under-sampled K-space data set in the central area of K-space to obtain multiple intermediate K-space data sets;

Perform second fitting recovery on the unsampled points of multiple intermediate K-space data sets in the non-center area of K-space to obtain multiple target K-space data sets;

Reconstruct multiple target K-space data sets and obtain magnetic resonance images corresponding to multiple excitations of the target site.

In another embodiment, the processor also performs the following steps when executing the computer program:

Use partial sampling technology to fill the central area of K-space to obtain the K-space calibration data set corresponding to the target part;

Excite the target part multiple times, and collect the undersampled K-space data set corresponding to each excitation;

Based on the K-space calibration data set, perform first fitting recovery on the unsampled points of each under-sampled K-space data set in the central area of K-space to obtain multiple intermediate K-space data sets;

Perform second fitting recovery on the unsampled points of multiple intermediate K-space data sets in the non-center area of K-space to obtain multiple target K-space data sets

Reconstruct multiple target K-space data sets and obtain magnetic resonance images corresponding to multiple excitations of the target site.

The implementation principles and technical effects of the computer program product provided by the above embodiments are similar to those of the above method embodiments, and will not be described again here.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage. In the media, when executed, the computer program may include the processes of the above method embodiments. Any reference to memory, storage, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory may include random access memory (Random Access Memory, RAM) or external cache memory. By way of illustration but not limitation, RAM can be in various forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM).

The technical features of the above embodiments can be combined in any way. To simplify the description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, all possible combinations should be used. It is considered to be within the scope of this manual.

The above-described embodiments only express several implementation modes of the present application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that, for those of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all fall within the protection scope of the present application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims (10)

1. An image reconstruction method, characterized in that the method includes: Obtain a K-space calibration data set corresponding to the target part, and the K-space calibration data set is fully sampled in the central area of K-space; Obtain multiple under-sampled K-space data sets corresponding to the target part, each under-sampled K-space data set being data collected by the target part in one excitation; Based on the K-space calibration data set, perform first fitting recovery on the unsampled points of each under-sampled K-space data set in the central area of K-space to obtain multiple intermediate K-space data sets; Perform a second fitting recovery on the unsampled points of the multiple intermediate K-space data sets in the non-center area of K-space to obtain multiple target K-space data sets; The multiple target K-space data sets are reconstructed to obtain magnetic resonance images corresponding to multiple excitations of the target part. 2. The method according to claim 1, characterized in that, based on the K-space calibration data set, the first simulation is performed on the unsampled points of each under-sampled K-space data set in the central area of the K-space. Comprehensive recovery, including: For any undersampled K-space data set, construct a data recovery matrix according to the K-space calibration data set and the under-sampled K-space data set; Based on the data recovery matrix corresponding to each of the under-sampled K-space data sets, a first fitting recovery is performed on the unsampled points in the central area of the K-space of each of the sampled K-space data sets. 3. The method according to claim 2, characterized in that, constructing a data recovery matrix according to the K-space calibration data set and the undersampled K-space data set includes: Based on the K-space calibration data set, construct a first low-rank matrix using a preset low-rank matrix construction method; Based on the undersampled K-space data set, use the low-rank matrix construction method to construct a second low-rank matrix; The data recovery matrix is generated according to the first low-rank matrix and the second low-rank matrix. 4. The method according to claim 3, characterized in that the process of constructing a low-rank matrix includes: Extract a preset number of different first data points from the target data set, and obtain the coordinate information of each first data point; the target data set is the K-space calibration data set or the under-sampled K-space data set; For any of the first data points, obtain a plurality of second data points whose distance from the first data point is less than a preset length, and obtain a set of data points corresponding to the first data points; Obtain signal values of a plurality of second data points in each of the data point sets; A low-rank matrix is constructed based on the coordinate information of each first data point and the signal values of a plurality of second data points in the data point set corresponding to each first data point. 5. The method according to any one of claims 1 to 4, characterized in that the second fitting recovery is performed on unsampled points of the plurality of intermediate K-space data sets in non-central areas of K-space. ,include: Calculate the weight kernel of the unsampled points of each intermediate K-space data set in the non-center area of K-space based on the fitted fully sampled data of each intermediate K-space data set in the central area of K-space; For any intermediate K-space data set, perform a second fitting recovery on the unsampled points in the non-center area according to the weight kernel of the unsampled points. 6. A magnetic resonance imaging method, characterized in that the method includes: Use partial sampling technology to fill the central area of K-space to obtain the K-space calibration data set corresponding to the target part; Excite the target part multiple times, and collect an undersampled K-space data set corresponding to each excitation; Based on the K-space calibration data set, perform first fitting recovery on the unsampled points of each under-sampled K-space data set in the central area of K-space to obtain multiple intermediate K-space data sets; Perform a second fitting recovery on the unsampled points of the multiple intermediate K-space data sets in the non-center area of K-space to obtain multiple target K-space data sets. The multiple target K-space data sets are reconstructed to obtain magnetic resonance images corresponding to multiple excitations of the target part. 7. The method according to claim 6, characterized in that the direction of the diffusion gradient applied in each of the multiple excitations is different. 8. The method according to claim 6, wherein the physiological period of the target part corresponding to each of the multiple excitations is different. 9. The method according to claim 6, characterized in that each of the multiple excitations applies a labeling pulse to the target site, and the undersampling is collected at different delay times after the labeling pulse is applied. K-space data set. 10. A computer device, comprising a memory and a processor, the memory stores a computer program, characterized in that when the processor executes the computer program, the method of any one of claims 1 to 9 is implemented. step.