Image reconstruction method and device, computer device and storage medium

Abstract

Embodiments of the present application relate to an image reconstruction method and device, computer equipment and a storage medium. The method comprises obtaining average k-space data of k-space raw data of a target region from the k-space raw data, reconstructing the average k-space data, obtaining reconstructed k-space data of a reconstructed combined image, obtaining difference k-space data between the k-space raw data and the reconstructed k-space data, reconstructing the difference k-space data to obtain a difference image, and determining a reconstructed image of the target region according to the difference image and the reconstructed combined image. The method can improve the quality of the final reconstructed image when the total scanning time is particularly short and the total collected data is particularly small.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular to an image reconstruction method, apparatus, computer device, and storage medium. Background Technology

[0002] Magnetic resonance imaging (MRI) is an imaging technique that reconstructs images by utilizing signals generated when atomic nuclei resonate within a strong magnetic field. This technique can obtain high-contrast, clear images of the interior of samples / tissues without causing damage or ionizing radiation, and has been widely used in various fields, especially in medical diagnostics.

[0003] MRI data acquisition is performed in the Fourier transform space (k-space) of the image. In order to speed up the acquisition process and save time, the k-space is often not fully acquired. However, the partially acquired k-space data needs to be reconstructed before an MRI image can be obtained.

[0004] However, current image reconstruction methods produce images of relatively low quality. Summary of the Invention

[0005] Based on this, and in response to the aforementioned technical problems, an image reconstruction method, apparatus, computer device, and storage medium are provided, which can reconstruct high-quality images.

[0006] In a first aspect, embodiments of this application provide an image reconstruction method, the method comprising:

[0007] Based on the original k-space data of the target location, obtain combined k-space data of the original k-space data;

[0008] Reconstruct the combined k-space data to obtain the reconstructed k-space data of the combined graph;

[0009] Obtain the difference k-space data between the original k-space data and the reconstructed k-space data, reconstruct the difference k-space data, and obtain the difference map;

[0010] Based on the difference map and the combined reconstructed map, the reconstructed image of the target area is determined.

[0011] In one embodiment, the above-described reconstruction of the combined k-space data to obtain the reconstructed k-space data of the reconstructed combined graph includes:

[0012] Obtain the reconstruction parameters, and reconstruct the combined k-space data based on the reconstruction parameters, combined k-space data, and preset reconstruction algorithm to obtain the reconstructed k-space data of the combined graph.

[0013] In one embodiment, the process of obtaining the difference k-space data between the original k-space data and the reconstructed k-space data, and reconstructing the difference k-space data to obtain a difference map includes:

[0014] Subtract the reconstructed k-space data from the original k-space data of each frame of the target area to obtain the difference k-space data of each frame of the target area.

[0015] The k-space data of the difference between each frame of the target area are reconstructed to obtain the difference map of each frame of the target area.

[0016] In one embodiment, determining the reconstructed image of the target region based on the difference map and the reconstructed combined image includes:

[0017] The difference map of each frame of the target area is fused with the reconstructed combined image to obtain the reconstructed image of each frame of the target area.

[0018] In one embodiment, obtaining combined k-space data based on the original k-space data of the target location includes:

[0019] The original data in k-space is combined to obtain combined k-space data.

[0020] In one embodiment, prior to obtaining the combined k-space data of the original k-space data as described above, the method further includes:

[0021] The k-space raw data of each frame of the target area is acquired to obtain the k-space raw data of the target area; the positions of the k-space data acquired in each frame of the target area are different.

[0022] In one embodiment, before obtaining the combined k-space data based on the k-space raw data of the target location, the method further includes:

[0023] Acquire the acquisition time and / or data acquisition volume of the raw k-space data of the target location;

[0024] If the acquisition time is less than the preset time threshold, and / or the data acquisition amount is less than the preset quantity, then the step of obtaining combined k-space data based on the k-space raw data of the target part is executed.

[0025] Secondly, embodiments of this application provide an image reconstruction apparatus, the apparatus comprising:

[0026] The acquisition module is used to acquire combined k-space data based on the original k-space data of the target part.

[0027] The first processing module is used to reconstruct the combined k-space data and obtain the reconstructed k-space data of the combined graph.

[0028] The second processing module is used to obtain the difference k-space data between the original k-space data and the reconstructed k-space data, and to reconstruct the difference k-space data to obtain the difference map.

[0029] The reconstruction module is used to determine the reconstructed image of the target area based on the difference map and the combined reconstructed image.

[0030] Thirdly, embodiments of this application provide 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 of the methods provided in the first aspect of the embodiments described above.

[0031] Fourthly, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of any of the methods provided in the first aspect of the embodiments described above.

[0032] This application provides an image reconstruction method, apparatus, computer device, and storage medium. Based on the original k-space data of the target region, the method obtains the average k-space data of the original k-space data, reconstructs the average k-space data, obtains the reconstructed k-space data of the reconstructed composite image, obtains the difference k-space data between the original k-space data and the reconstructed k-space data, reconstructs the difference k-space data to obtain a difference map, and then determines the reconstructed image of the target region based on the difference map and the reconstructed composite image. The k-space base data of the target area is only partially acquired, and the total scanning time and the amount of data acquired are very short. In this case, this application extracts the combined k-space data from the original k-space data of the target area separately for reconstruction. This is equivalent to separating the data with common characteristics and processing them separately, which greatly improves the quality of the combined image reconstructed from this part of the k-space data. After extracting the data with common characteristics, the remaining difference k-space data also greatly reduces the error in the difference image reconstruction process due to the reduction in signal amplitude, and also improves the quality of the reconstructed difference image. Finally, the higher quality difference image and the higher quality combined image are fused together, which further improves the quality of the final reconstructed image. Attached Figure Description

[0033] Figure 1 This is an application environment diagram of an image reconstruction method provided in one embodiment;

[0034] Figure 2 This is a schematic diagram of an image reconstruction process provided in one embodiment;

[0035] Figure 2aThis is a schematic diagram of an image reconstruction algorithm provided in one embodiment;

[0036] Figure 2b This is a schematic diagram of k-space data acquisition provided in one embodiment;

[0037] Figure 3 This is a schematic diagram of an image reconstruction process provided in another embodiment;

[0038] Figure 4 This is a schematic diagram of an image reconstruction process provided in another embodiment;

[0039] Figure 5 This is a schematic diagram of an image reconstruction process provided in another embodiment;

[0040] Figure 6 A flowchart of an image reconstruction process is provided in another embodiment;

[0041] Figure 7 This is a schematic diagram of an image reconstruction process provided in one embodiment;

[0042] Figure 8 This is a schematic diagram of an image reconstruction process provided in one embodiment;

[0043] Figure 9 This is a schematic diagram of an echo-shared data acquisition trajectory provided in one embodiment;

[0044] Figure 10 This is a structural block diagram of an image reconstruction apparatus provided in one embodiment;

[0045] Figure 11 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0047] The image reconstruction method provided in this application can be applied to, for example... Figure 1The application environment shown includes a computer device whose processor provides computational and control capabilities. The computer device's memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The computer device's database contains data relevant to image reconstruction. The computer device's network interface is used for communication with other external devices via a network connection. When the computer program is executed by the processor, it implements an image reconstruction method.

[0048] The technical solution of this application and how it solves the above-mentioned technical problems will be described in detail below with reference to embodiments and 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 the image reconstruction method provided in this application uses a computer device as the execution subject, but the execution subject can also be an image reconstruction apparatus, which can be implemented as part or all of the computer device through software, hardware, or a combination of software and hardware.

[0049] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0050] Typically, each MRI image in dynamic imaging (Magnetic Resonance Imaging, MRI) can be split into the sum of two images, for example, splitting the image into an average image and a difference image. Similarly, k-space data can also be split into average k-space data and difference k-space data. The average value can refer to a time average, which is a linear combination operation. In practical applications, besides linear combination operations, the average value can also be obtained through non-linear combination operations; this application does not limit this. Taking combination as an example, the combined k-space data can be obtained from the original k-space data of the target area. The original k-space data can be split into combined k-space data and difference k-space data. Based on this, this application can obtain the final reconstructed MRI image by splitting the original k-space data into combined k-space data and difference k-space data, and then reconstructing the combined k-space data and difference k-space data separately.

[0051] In one embodiment, Figure 2An image reconstruction method is provided. This embodiment involves: a computer device determining combined k-space data based on the original k-space data of the target region, reconstructing the combined k-space data to obtain the reconstructed k-space data of the reconstructed combined image; then determining the difference k-space data between the original k-space data and the reconstructed k-space data; reconstructing a difference map based on the difference k-space data; and finally determining the reconstructed image of the target region based on the difference map and the reconstructed combined image. The specific process is as follows: Figure 2 As shown, this embodiment includes the following steps:

[0052] S101, Based on the original k-space data of the target part, obtain the combined k-space data of the original k-space data.

[0053] The target region refers to the area where image reconstruction is needed, such as the heart, head, brain, coronary arteries, or any other part of the body; this application does not limit this. The k-space is the dual space of ordinary space under Fourier transform, representing the spatial frequency space in MR (magnetic resonance) images. The data acquired by scanning the target region using an MRI scanner is k-space data, referred to as the raw k-space data of the target region. Here, the raw k-space data of the target region refers to all k-space data of the target region; that is, the k-space data corresponding to each MRI image during the dynamic imaging process of the target region collectively constitutes the raw k-space data of the target region. Furthermore, the k-space data corresponding to each MRI image is only partially sampled. Complete acquisition would take too long and fail to achieve high temporal resolution. Therefore, acquisition efficiency can be improved by acquiring only partial k-space data to reconstruct the MRI image. It is important to emphasize that at this point, only partial k-space data of each MRI image of the target region has been acquired (no single image is fully acquired), but the dynamic MRI image of the target region has not yet been obtained. It should also be noted that in some scenarios, k-space data for each MRI image can be deliberately collected from different k-space locations. For example, the k-space data for the first image can be collected from locations 1, 3, and 7, while the k-space data for the second image can be collected from locations 2, 3, and 9, and so on. The k-space data collection locations for each image are staggered. However, in other scenarios, some acquisition lines can be allowed to overlap. For example, when it is necessary to keep the k-space center of each image fully acquired, these lines (data lines) are not deliberately collected from different k-space locations.

[0054] For example, if the target site is the heart, then the raw k-space data for the target site refers to the k-space data corresponding to all MRI images during dynamic imaging of the heart (imaging of the heart's beating process). For instance, if the final required dynamic MRI images of the heart are MRI image 1, MRI image 2, MRI image 3, ..., MRI image N, then the raw k-space data for the heart refers to: k-space data 1 (corresponding to MRI image 1) partially acquired at positions 1, 3, and 7; k-space data 2 (corresponding to MRI image 2) partially acquired at positions 2, 3, and 9; k-space data 3 (corresponding to MRI image 3) partially acquired at positions 4, 7, and 10; ..., k-space data N (corresponding to MRI image N) partially acquired at positions 2, 5, and 8.

[0055] Optionally, partial k-space raw data of each frame of the target area needs to be acquired in advance; and the location of the k-space data acquired in each frame of the target area is different. The methods for acquiring the k-space raw data of the target area can be as follows: firstly, the k-space raw data is obtained by scanning the target area of ​​the object to be scanned using an MRI machine, and then stored in a storage device. When the k-space raw data is needed, it can be directly retrieved from the storage device; secondly, it can be k-space raw data downloaded from other devices (e.g., a network platform); thirdly, it can be k-space raw data obtained after scanning the target area of ​​the object to be scanned using an MRI machine. This embodiment does not limit the method of acquiring the k-space raw data, as long as the k-space raw data can be obtained.

[0056] MRI images are obtained using MRI imaging technology based on k-space data. This process, also known as reconstruction, involves using reconstruction algorithms (part of MRI imaging technology) to reconstruct the corresponding MRI image from the acquired k-space data. Decoding the spatial localization encoding information in the k-space data yields image domain data, which in turn produces the MRI image data. Therefore, reconstruction is the process of converting k-space data into image domain data. The relationship between k-space data points and image pixels is that of Fourier transform and inverse Fourier transform; each data point in k-space contains information about the entire image. Generally, when acquiring data in k-space, fully sampled k-space data can be directly subjected to an inverse Fourier transform to obtain the corresponding MRI image.

[0057] Optionally, the reconstruction algorithm includes, but is not limited to, parallel imaging algorithms such as SENSE (Sensitivity Encoding) and GRAPPA (Gene Retized Autoencoding librating Patially Parallel Acquisitions), which can reduce the signal-to-noise ratio loss caused by reducing the number of sampling rows.

[0058] Take the GRAPPA algorithm as an example. The GRAPPA algorithm utilizes the correlation between data points in k-space to calculate uncollected data using acquired data. An uncollected data point in k-space can be obtained by weighted summing of surrounding acquired points from various channels. Specifically, this calculation can be performed using convolution operations, where the convolution kernel acts as the weighting coefficients. Typically, these weighting coefficients are translationally invariant in k-space. For an example, see [link to example]. Figure 2a and Figure 2b As shown, accelerating imaging using the GRAPPA algorithm involves the following steps:

[0059] (1) Obtain the fully sampled portion of the k-space as reference data (also known as the reference line). Please refer to [link to relevant documentation]. Figure 2a , Figure 2a The dashed box in the diagram represents the fully mined area, which can be used as k-space reference data to calculate the weighting coefficients.

[0060] (2) Calculate the weighting coefficients based on the collected k-space reference data:

[0061] Please see Figure 2b , Figure 2b There are three channels (i.e., coils): Coil 1, Coil 2, and Coil 3, each channel including 8 data points. Figure 2b If the black dots in the image represent the input data, then there are a total of 24 data points in the three channels. The output of these 24 data points is a reconstructed data point. The reference data is fully sampled, and each data point has a weight coefficient. The 24 weight coefficients of a convolutional kernel can be calculated by solving a linear equation.

[0062] (3) Reconstruct the uncollected k-space data using the calculated weighting coefficients.

[0063] A total of three convolutional kernels are needed for the three channels to reconstruct the uncollected data points in each of the three channels. Please continue reading... Figure 2b , Figure 2b Point P shown in the diagram is a point in channel Coil2. Based on point P, other data points in channel Coil2 can be reconstructed by this convolution kernel through horizontal movement, that is, by performing convolution operations, thereby obtaining the uncollected k-space data and completing the reconstruction.

[0064] Optionally, obtaining combined k-space data from the original k-space data involves performing a combination operation on the original k-space data to obtain combined k-space data.

[0065] The combination operation can be either linear or nonlinear. For example, time averaging is a type of linear combination operation. Furthermore, while time averaging yields a single k-space data set with an average value, other combination operations can produce different numbers of combined k-space data sets. These "other numbers" can be two, three, etc., and this embodiment does not limit the specific number of combinations.

[0066] The MRI image corresponding to the combined k-space data is called a composite image. A composite image is a map of combined values ​​calculated from the original k-space data; that is, the original k-space data is combined and processed to obtain the combined k-space data. Then, the composite image is obtained by reconstructing the combined k-space data. For example, taking the heart as an example again, after obtaining the original k-space data of the heart, the original k-space data of the heart is combined and processed to obtain the combined k-space data of the heart. Then, the composite image of the heart is obtained by reconstructing the combined k-space data of the heart.

[0067] S102, Reconstruct the combined k-space data to obtain the reconstructed k-space data of the combined graph.

[0068] When combining raw k-space data of a target region, the averaging time corresponds to the acquisition time of that raw k-space data. As mentioned earlier, the acquisition of raw k-space data for the target region is partial; the less k-space data acquired, the shorter the acquisition time. Acquisition efficiency is higher. Therefore, because the raw k-space data for the target region is partially acquired, its acquisition time is relatively short, resulting in an insufficient averaging time. This means the motion of the target region cannot be completely averaged out, and motion artifacts may remain, leading to low quality of the final composite image. For example, the heart undergoes continuous motion. Since the raw k-space data for the heart is specifically acquired from different locations, combining the raw k-space data for the heart may result in different k-space locations containing different motion information, or corresponding to different movements and different positions of cardiac contraction and relaxation. In particular, the acquisition time of the raw k-space data of the heart is relatively short. Combining the raw k-space data of the heart with the average of the various states of heart motion results in poor consistency of the combined k-space data of the heart. There are many inconsistencies between different locations. This inconsistency leads to low quality of the MRI image corresponding to the combined k-space data of the heart (i.e., the combined image corresponding to the combined k-space data).

[0069] Therefore, the MRI images corresponding to the combined k-space data calculated directly from the original k-space data have low quality. Based on this, this step reconstructs the acquired combined k-space data. Reconstruction is a process that forces the data to improve its self-consistency; improving data self-consistency means improving data consistency. Therefore, the quality of the combined image corresponding to the reconstructed combined k-space data is higher than that of the combined image corresponding to the unreconstructed combined k-space data. Furthermore, all stationary parts (signals that do not change over time) in the k-space data appear in the combined data. Therefore, the combined data in the k-space data contains significantly more difference data containing motion than motion. Thus, the combined data can be called data with commonalities. In this way, processing the commonalities in the partially acquired (shorter acquisition time) original k-space data separately can effectively improve the quality of the combined image corresponding to the combined k-space data.

[0070] Optionally, one way to reconstruct the combined k-space data and obtain the reconstructed k-space data of the combined graph includes: obtaining reconstruction parameters, and reconstructing the combined k-space data according to the reconstruction parameters, the combined k-space data and a preset reconstruction algorithm to obtain the reconstructed k-space data of the combined graph.

[0071] Here, reconstruction parameters refer to the parameters used in the reconstruction algorithm. For example, reconstruction parameters may be a coil sensitivity function. Optionally, reconstruction parameters can be obtained from a reference image.

[0072] The reference image can be obtained either by reconstructing k-space data or by combining historically acquired data. In the method of reconstructing by acquiring k-space data, the acquired k-space data must be fully captured, or the k-space center must be fully captured. Generally, using only the reference image with the fully captured k-space center is sufficient to obtain the parameters needed for reconstruction. Full image capture refers to capturing the entire k-space, while full k-space center capture means capturing only a portion of the central k-space region, with the surrounding k-space regions only partially captured (when reconstructing the reference image, the surrounding data is discarded; for example, it can be filled with zeros or not used at all, using only the data captured at the k-space center). The size of the centrally captured region is not limited; for example, it can be more than 20 lines. In the method of obtaining the reference image by combining historically acquired data, the image corresponding to the combined k-space data can be used as the reference image; that is, the combined image can be used as the reference image to estimate the reconstruction parameters in the reconstruction algorithm.

[0073] After estimating the reconstruction parameters using the reference image, these parameters are used as known parameters in the reconstruction algorithm. Since the combined k-space data has already been obtained in the preceding steps, substituting the reconstruction parameters and the combined k-space data into the preset algorithm yields the reconstructed composite image. The k-space data corresponding to this reconstructed composite image is the reconstructed k-space data of the reconstructed composite image.

[0074] It should be emphasized that the combined k-space data mentioned above can include multiple combined k-space data. That is, the combined k-space data includes at least one sub-combined k-space data. For example, it can include one sub-combined k-space data, or it can include two sub-combined k-space data, etc.

[0075] Based on this, the combined graph mentioned in the embodiments of this application refers to the sub-reconstructed graph and sub-reconstructed k-space data obtained after reconstructing the at least one sub-combined k-space data respectively. Then, a linear combination operation is performed on these sub-reconstructed graphs, for example, by calculating an average value, which is the combined graph mentioned in the embodiments of this application. Similarly, a linear combination operation is performed on these sub-reconstructed k-space data, for example, by calculating an average value, which is the reconstructed k-space data of the combined graph mentioned in the embodiments of this application.

[0076] S103, obtain the difference k-space data between the original k-space data and the reconstructed k-space data, reconstruct the difference k-space data, and obtain the difference map.

[0077] Image reconstruction is a statistical estimation process, and it is impossible to completely restore the image; there will always be errors. Besides noise in the original data, another source of error is inaccuracy during the reconstruction process. This error is highly dependent on the signal amplitude in the k-space data; that is, the less k-space data used for reconstruction, the smaller the signal amplitude, and thus the smaller the error in the final reconstructed image.

[0078] In this way, the combined k-space data is extracted from the original k-space data and reconstructed separately, making all the time-invariant parts of the remaining difference k-space data very small (the time-invariant signal has been carried away by the combined k-space data). Therefore, the amplitude of the signal in the difference k-space data is very low, and the error will be greatly reduced when reconstructing the difference k-space data.

[0079] Furthermore, each k-space data point in the original k-space data is only partially collected. The reconstructed k-space data obtained in step S102 is the reconstructed k-space data obtained by splitting the combined k-space data from the original k-space data. Reconstruction is a process that forces the data to improve consistency. Therefore, reconstructed k-space data can be seen as: combined k-space data split from the original k-space data composed of fully collected k-space data points. If the original k-space data was collected over a long period and a large amount of data was collected, then when the combined k-space data was split from the original k-space data, a large amount of data would be close to the state average. In this case, the quality of the combined k-space data would be high (high signal-to-noise ratio, no artifacts). Such a combined k-space data does not need to be reconstructed; the combined k-space data can be directly subjected to an inverse Fourier transform to obtain the combined k-space data.

[0080] In other words, by reconstructing the combined k-space data extracted from the partially acquired raw k-space data through the above S102 step, the extracted combined k-space data is closer to the fully acquired k-space data. This allows the reconstructed k-space data, which is subtracted from the raw k-space data, to carry away more signals that do not change over time, further reducing the amplitude of the signals in the difference k-space data and greatly reducing the error of the reconstructed difference map.

[0081] Therefore, in this embodiment, the reconstructed k-space data is subtracted from the original k-space data in step S101 to obtain the difference k-space data. The reconstructed k-space data is then extracted and processed further so that only the difference map needs to be reconstructed. All parts of the difference map that do not change over time become very small, resulting in a low signal amplitude and thus a small error after reconstruction. Therefore, a smaller signal will only cause a smaller error, resulting in lower image artifacts, a higher signal-to-noise ratio, and ultimately, a difference map with higher image quality.

[0082] S104. Based on the difference map and the reconstructed combined map, determine the reconstructed image of the target area.

[0083] After the above steps S102 and S103, a high-quality composite image and a high-quality difference image can be obtained. Based on the high-quality composite image and the high-quality difference image, the image quality of the reconstructed image of the target area will be greatly improved.

[0084] Optionally, one method for determining the reconstructed image of a target region based on the difference map and the reconstructed combined image is to fuse the difference map of each frame of the target region with the reconstructed combined image to obtain the reconstructed image of each frame of the target region.

[0085] The acquired k-space raw data includes k-space data collected from multiple parts. Therefore, the reconstructed difference map and the reconstructed composite map in the above steps also include multiple frames of images (one k-space data point in the original k-space data corresponds to one frame of image). Therefore, when determining the reconstructed image of the target area based on the difference map and the reconstructed composite map, fusion is also performed on a single frame basis.

[0086] Specifically, the difference map and the reconstructed combined image are in one-to-one correspondence. For a pair of difference maps and reconstructed combined images, pixels at the same location are fused, for example, by weighted summation or by fusion using a preset fitting function, to obtain the fused pixel value at the corresponding location. Thus, the image composed of the fused pixel values ​​at each location is the reconstructed image after fusing the difference map and the reconstructed combined image for that pair. After fusion, multiple fused reconstructed images can be obtained, forming a dynamic MRI image (a series of MRI images) of the target area.

[0087] In this embodiment, based on the original k-space data of the target region, the average k-space data of the original k-space data is obtained. The average k-space data is then reconstructed to obtain the reconstructed k-space data of the combined image. The difference k-space data between the original k-space data and the reconstructed k-space data is also obtained. This difference k-space data is then reconstructed to obtain a difference map. Finally, based on the difference map and the reconstructed combined image, the reconstructed image of the target region is determined. Since the k-space data of the target region is only partially acquired due to a very short total scanning time and limited data, this application extracts the combined k-space data from the original k-space data of the target region for reconstruction. This is equivalent to separating and processing the common data segments, significantly improving the quality of the reconstructed combined image. Furthermore, after extracting the common data segments, the remaining difference k-space data also experiences a reduction in signal amplitude, greatly reducing errors in the difference map reconstruction process and improving the quality of the reconstructed difference map. Finally, the higher-quality difference map and the higher-quality combined image are merged, further improving the quality of the final reconstructed image.

[0088] Based on the above embodiments, the reconstruction process of the difference map is described below using an example. Figure 3 As shown, in one embodiment, the above-described S103 includes the following steps:

[0089] S201, subtract the reconstructed k-space data from the original k-space data of each frame of the target area to obtain the difference k-space data of each frame of the target area.

[0090] The raw k-space data mentioned earlier includes multiple parts of acquired k-space data. Therefore, the difference k-space data also includes multiple k-space data, each corresponding to an MRI image. This process is explained using a single frame image as a unit; however, this single frame image is only for distinguishing different k-space data and does not actually exist. It is important to emphasize that the k-space data resulting from combining the k-space data mentioned earlier can be multiple. Each of the multiple sub-combinations of k-space data can correspond to a reconstructed k-space data. A linear combination operation can be performed on the reconstructed k-space data from each of these multiple sub-combinations, for example, calculating an average value. This average value is called the reconstructed k-space data, and it represents the average of the reconstructed k-space data from each of the multiple sub-combinations.

[0091] Then, by subtracting the reconstructed k-space data from the original k-space data of each frame of the target area, the difference k-space data of each frame of the target area can be obtained.

[0092] The k-space raw data consists of N layers, each of which is a two-dimensional dynamic movie, meaning each layer contains multiple frames of images, and each frame corresponds to a capture time. So, suppose that the k-space data captured for each frame of a certain layer of the k-space raw data includes: k-space raw data 1, k-space raw data 2, k-space raw data 3, ..., k-space raw data N.

[0093] Therefore, the combined k-space data extracted from the original k-space data is: combined k-space data; the reconstructed k-space data is: reconstructed k-space data.

[0094] The original k-space data of each frame of the target area is then subtracted from the reconstructed k-space data to obtain the difference k-space data 1, ... and so on, until the original k-space data N is subtracted from the reconstructed k-space data to obtain the difference k-space data N.

[0095] S202, reconstruct the k-space data of the difference between each frame of the target area to obtain the difference map of each frame of the target area.

[0096] After obtaining the k-space difference data of each frame of the target area, the k-space difference data of each frame of the target area is reconstructed to obtain the difference map of each frame of the target area, that is, the MRI image corresponding to the k-space difference data of each frame of the image. For example, reconstruction can be performed by a reconstruction algorithm, and the embodiments of this application do not limit the algorithm used.

[0097] In this embodiment, the k-space data of each frame image is used as a unit for reconstruction, so that each difference k-space data corresponds to a high-quality difference map.

[0098] For cases with long acquisition times and a large amount of k-space data acquired (e.g., full acquisition), performing an inverse Fourier transform on the original k-space data can yield a reconstructed image of high quality. However, for cases with short acquisition times and a small amount of k-space data acquired, the image reconstruction method provided in this application can also obtain a reconstructed image of high quality. Therefore, before executing the embodiments of this application, the acquisition time or the amount of k-space data acquired can be determined to select a suitable image reconstruction method. In one embodiment, such as... Figure 4 As shown, the method also includes:

[0099] S301, acquire the acquisition time and / or data acquisition amount of the raw k-space data of the target part.

[0100] S302, if the acquisition time is less than the preset time threshold, and / or the data acquisition amount is less than the preset quantity, then the step of obtaining the combined k-space data based on the k-space raw data of the target part is executed.

[0101] The acquisition time can be determined by the start and end scan times recorded by the magnetic resonance imaging (MRI) device; the amount of data acquired can also be determined by the total amount of data acquired recorded by the MRI device, or by receiving the acquisition time and / or the amount of data acquired by the user. This application embodiment does not limit this.

[0102] A time threshold and a preset quantity are pre-defined. The system determines whether the acquisition time is less than the preset time threshold and whether the data acquisition quantity is less than the preset quantity. If either condition is met, it is determined that image reconstruction can be performed according to the image reconstruction method provided in this application embodiment. This involves starting the step of obtaining combined k-space data based on the original k-space data of the target region.

[0103] Alternatively, if the total amount of data is so small that the average number of times each data point is sampled is less than a certain value, then the data can be considered too little. For example, if the average number of times each data point is sampled is less than two, then the total amount of data collected is considered small.

[0104] For example, if we acquire 20 heart images per cardiac cycle, and the data is fully acquired (a large amount of data is acquired), then each data point is acquired 20 times (more acquisition times mean longer acquisition time). However, if the speed is increased fourfold, the acquisition time is shortened, and each data point is acquired an average of five times. Similarly, if the speed is increased tenfold, the acquisition time is even shorter, and each data point is acquired an average of two times. Thus, the number of acquisitions can be used as a criterion for judging shorter acquisition time and less data acquired, making it simpler and faster to determine the acquisition time and amount of raw K-space data.

[0105] The image reconstruction method provided in this application determines whether to perform image reconstruction according to the image reconstruction method provided in this application by judging the acquisition time and / or data acquisition amount of the k-space original data of the target part. This avoids the situation where if this condition is not met, such as a long acquisition time or a large amount of k-space data is acquired (e.g., full acquisition), an inverse Fourier transform can be directly performed, thereby making the reconstruction method of the k-space original data of the target part more flexible.

[0106] In addition, such as Figure 5 As shown, in one embodiment, this application also provides an image reconstruction method, which includes:

[0107] S401, acquire partial k-space raw data of each frame of the target area to obtain the k-space raw data of the target area.

[0108] S402, acquire the acquisition time and / or data acquisition amount of the raw k-space data of the target part.

[0109] S403, if the acquisition time is less than the preset time threshold, and / or the data acquisition amount is less than the preset quantity, then based on the k-space original data of the target part, obtain the combined k-space data of the k-space original data.

[0110] S404: Obtain reconstruction parameters, and reconstruct the combined k-space data according to the reconstruction parameters, combined k-space data and preset reconstruction algorithm to obtain the reconstructed k-space data of the combined graph.

[0111] S405, subtract the reconstructed k-space data from the original k-space data of each frame of the target area to obtain the difference k-space data of each frame of the target area.

[0112] S406, Reconstruct the k-space data of the difference between each frame of the target area to obtain the difference map of each frame of the target area.

[0113] S407, the difference map of each frame of the target area and the reconstructed combined map are fused to obtain the reconstructed image of each frame of the target area.

[0114] The image reconstruction method provided in this embodiment is similar in principle and technical effect to the above-described method embodiments, and will not be repeated here.

[0115] In addition, such as Figure 6 As shown, in one embodiment, this application also provides an image reconstruction method, which includes:

[0116] S501: Acquire partial k-space raw data of each frame of the target area to obtain the k-space raw data of the target area.

[0117] S502, acquire the acquisition time and / or data acquisition amount of the raw k-space data of the target part.

[0118] S503, if the acquisition time is less than the preset time threshold, and / or the data acquisition amount is less than the preset quantity, then based on the k-space original data of the target part, obtain a combination map of part of the k-space original data.

[0119] S504: Obtain reconstruction parameters, and reconstruct the composite graph of part of the original k-space data according to the reconstruction parameters and the preset reconstruction algorithm to obtain the reconstructed k-space data of the composite graph.

[0120] S505, subtract the reconstructed k-space data of the combined image from the original k-space data of each frame of the target area to obtain the difference k-space data of each frame of the target area.

[0121] S506, Reconstruct the k-space data of the difference between each frame of the target area to obtain the difference map of each frame of the target area.

[0122] S507, the difference map of each frame of the target area and the reconstructed combined map are fused to obtain the reconstructed image of each frame of the target area.

[0123] The image reconstruction method provided in this embodiment is similar in principle and technical effect to the above-described method embodiments, and will not be repeated here.

[0124] The following two figures compare and contrast existing image reconstruction methods with the image reconstruction methods provided in the embodiments of this application.

[0125] like Figure 7As shown, the existing image reconstruction method first acquires the original k-space data of the MRI image, then combines the original k-space data to obtain combined k-space data, and finally reconstructs the MRI image corresponding to the original k-space data.

[0126] and Figure 8 The image reconstruction method provided in this application embodiment first acquires the original k-space data of an MRI image, then combines the original k-space data to obtain combined k-space data, and then reconstructs the combined k-space data separately to obtain reconstructed k-space data or a reconstructed combined image; simultaneously, the reconstructed k-space data is subtracted from the original k-space data to obtain difference k-space data, and then the difference k-space data is reconstructed separately to obtain a difference image; finally, the reconstructed combined image and the reconstructed difference image are fused to obtain the final MRI image.

[0127] Regarding the reconstruction of partially collected k-space data, a comparison of the two methods reveals... Figure 7 The method used in this study involves directly reconstructing the image based on partially acquired k-space data. This results in very poor quality reconstructed MRI images. Figure 8 The reconstruction method involves reconstructing the combined k-space data and the difference k-space data separately, and then fusing them to obtain the reconstructed MRI image. This results in high-quality combined and difference images being obtained separately, leading to a high-quality MRI image.

[0128] Additionally, in this embodiment, echo-sharing can be incorporated, where lines in the k-space near the maximum and minimum ky values ​​in each image are shared by two adjacent images. Echo-sharing is a technique that reuses a k-space line between two adjacent images, i.e., reusing a portion of data, or in other words, when two images have a temporal overlap, the data collected in the overlapping portion is used by both images. Figure 9 To collect a dynamic image of the trajectory, for example, the initial sampling trajectory is collected from top to bottom in k-space, then from bottom to top, and so on, until the data collection is complete.

[0129] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some of the steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the steps or stages of other steps.

[0130] In one embodiment, such as Figure 10 As shown, an image reconstruction apparatus is provided, comprising: an acquisition module 10, a first processing module 11, a second processing module 12, and a reconstruction module 13, wherein:

[0131] The acquisition module 10 is used to acquire combined k-space data based on the original k-space data of the target part.

[0132] The first processing module 11 is used to reconstruct the combined k-space data and obtain the reconstructed k-space data of the combined graph.

[0133] The second processing module 12 is used to obtain the difference k-space data between the original k-space data and the reconstructed k-space data, and to reconstruct the difference k-space data to obtain the difference map.

[0134] The reconstruction module 13 is used to determine the reconstructed image of the target area based on the difference map and the combined reconstructed image.

[0135] In one embodiment, the first processing module is further configured to acquire reconstruction parameters, and reconstruct the combined k-space data according to the reconstruction parameters, the combined k-space data and the preset reconstruction algorithm to obtain the reconstructed k-space data of the reconstructed combined graph.

[0136] In one embodiment, the second processing module 12 includes:

[0137] The data acquisition unit is used to subtract the reconstructed k-space data from the original k-space data of each frame image of the target part to obtain the difference k-space data of each frame image of the target part.

[0138] The image acquisition unit is used to reconstruct the k-space data of the difference between each frame of the target area to obtain the difference map of each frame of the target area.

[0139] In one embodiment, the reconstruction module 13 is further configured to fuse the difference map of each frame of the target region with the reconstructed combined map to obtain the reconstructed image of each frame of the target region.

[0140] In one embodiment, the acquisition module 10 is further configured to combine the original k-space data to obtain combined k-space data.

[0141] In one embodiment, the device further includes: a data acquisition module, used to acquire partial k-space raw data of each frame image of the target region to obtain the k-space raw data of the target region; the positions of the k-space data acquired from each frame image of the target region are different.

[0142] In one embodiment, the device further includes: a condition detection module, used to acquire the acquisition time and / or data acquisition amount of the k-space raw data of the target part; if the acquisition time is less than a preset time threshold, and / or the data acquisition amount is less than a preset quantity, then the step of acquiring combined k-space data based on the k-space raw data of the target part is executed.

[0143] Specific limitations regarding the image reconstruction apparatus can be found in the limitations of the image reconstruction method described above, and will not be repeated here. Each module in the aforementioned image reconstruction apparatus can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0144] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 11 As shown, the computer device includes a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements an image reconstruction method. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the computer device casing, or an external keyboard, touchpad, or mouse.

[0145] Those skilled in the art will understand that Figure 11 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0146] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0147] Based on the original k-space data of the target location, obtain combined k-space data of the original k-space data;

[0148] Reconstruct the combined k-space data to obtain the reconstructed k-space data of the combined graph;

[0149] Obtain the difference k-space data between the original k-space data and the reconstructed k-space data, reconstruct the difference k-space data, and obtain the difference map;

[0150] Based on the difference map and the combined reconstructed map, the reconstructed image of the target area is determined.

[0151] The computer device provided in the above embodiments has a similar implementation principle and technical effect to the above method embodiments, and will not be described again here.

[0152] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0153] Based on the original k-space data of the target location, obtain combined k-space data of the original k-space data;

[0154] Reconstruct the combined k-space data to obtain the reconstructed k-space data of the combined graph;

[0155] Obtain the difference k-space data between the original k-space data and the reconstructed k-space data, reconstruct the difference k-space data, and obtain the difference map;

[0156] Based on the difference map and the combined reconstructed map, the reconstructed image of the target area is determined.

[0157] The computer-readable storage medium provided in the above embodiments has a similar implementation principle and technical effect to the above method embodiments, and will not be described again here.

[0158] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0159] The technical features of the above embodiments can be combined in any way. For the sake of brevity, 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, they should be considered to be within the scope of this specification.

[0160] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the embodiments of this application, and these all fall within the protection scope of the embodiments of this application. Therefore, the protection scope of the patent for the embodiments of this application should be determined by the appended claims.

Claims

1. An image reconstruction method, characterized in that, The method includes: If the acquisition time of the raw k-space data of the target part is less than a preset time threshold and / or the amount of data acquired is less than a preset number, a time-averaged combination operation is performed on the raw k-space data to obtain combined k-space data of the raw k-space data; the raw k-space data of the target part includes the k-space data corresponding to each MRI image during the dynamic imaging process of the target part, and the position of the k-space data acquired in each frame of the image is different. The combined k-space data is reconstructed to obtain the reconstructed k-space data of the combined graph; the combined k-space data consists of common data in the original k-space data. Obtain the difference k-space data between the original k-space data and the reconstructed k-space data, and reconstruct the difference k-space data to obtain the difference map; Based on the difference map and the reconstructed combined map, the reconstructed image of the target region is determined.

2. The method according to claim 1, characterized in that, The step of reconstructing the combined k-space data to obtain the reconstructed k-space data of the combined graph includes: Obtain reconstruction parameters, and reconstruct the combined k-space data according to the reconstruction parameters, the combined k-space data and the preset reconstruction algorithm to obtain the reconstructed k-space data of the reconstructed combined graph.

3. The method according to claim 2, characterized in that, The acquisition of reconstruction parameters includes: The k-space data of the target area is acquired using a full acquisition method, and the k-space data is reconstructed to obtain a reference image; The reconstruction parameters are obtained from the reference image.

4. The method according to claim 2, characterized in that, The acquisition of reconstruction parameters includes: The k-space data of the target area is acquired by full sampling of the k-space center, and the k-space data is reconstructed to obtain a reference image; The reconstruction parameters are obtained from the reference image.

5. The method according to any one of claims 1-4, characterized in that, The step of obtaining the difference k-space data between the original k-space data and the reconstructed k-space data, and reconstructing the difference k-space data to obtain a difference map includes: Subtract the reconstructed k-space data from the original k-space data of each frame of the target region to obtain the difference k-space data of each frame of the target region. The difference k-space data of each frame image of the target region is reconstructed to obtain the difference map of each frame image of the target region.

6. The method according to claim 5, characterized in that, Determining the reconstructed image of the target region based on the difference map and the reconstructed combined image includes: The difference map of each frame of the target area and the reconstructed combined image are fused together to obtain the reconstructed image of each frame of the target area.

7. The method according to any one of claims 1-4, characterized in that, The method further includes: The k-space raw data of each frame of the target area is acquired to obtain the k-space raw data of the target area; the positions of the k-space data acquired in each frame of the target area are different.

8. An image reconstruction apparatus, characterized in that, The device includes: The acquisition module is used to perform a time-averaged combination operation on the raw k-space data when the acquisition time of the raw k-space data of the target part is less than a preset time threshold and / or the amount of data acquired is less than a preset number, to obtain combined k-space data of the raw k-space data; the raw k-space data of the target part includes the k-space data corresponding to each MRI image during the dynamic imaging process of the target part, and the position of the k-space data acquired in each frame of the image is different. The first processing module is used to reconstruct the combined k-space data and obtain the reconstructed k-space data of the reconstructed combined graph; the combined k-space data is the data with common characteristics in the original k-space data. The second processing module is used to obtain the difference k-space data between the original k-space data and the reconstructed k-space data, and to reconstruct the difference k-space data to obtain a difference map. The reconstruction module is used to determine the reconstructed image of the target region based on the difference map and the reconstructed combined map.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images