Magnetic resonance imaging reconstruction method and device, medical imaging equipment and storage medium

Abstract

The invention relates to the technical field of medical images, and discloses a magnetic resonance imaging reconstruction method and device, medical imaging equipment and a storage medium. The method comprises the steps that diffusion weighted imaging DWI data and an estimation image sequence are obtained, the diffusion weighted imaging DWI data are obtained through multiple imaging sequences, and the estimation image sequence is related to each excitation of the multiple imaging sequences; a phase weight is then calculated for each estimated image based on the signal strength of the sequence of estimated images. The larger the signal intensity is, the higher the corresponding weight is, the phase weight of each estimation image in the estimation image sequence is obtained, the weight image sequence is obtained, it is ensured that each position has signals, the influence of tissue information on the result is reduced, and therefore signal loss is effectively avoided. And finally, reconstruction is carried out based on diffusion weighted imaging DWI data, the estimated image sequence and the weighted image sequence, artifacts are reduced, the problem of signal loss in the reconstruction process is optimized, the overestimation phenomenon of an ADC is reduced, and a target reconstruction image is obtained.

Description

Technical Field

[0001] The present invention relates to the field of medical imaging technology, and in particular to a magnetic resonance imaging reconstruction method, device, medical imaging equipment and storage medium. Background Art

[0002] Diffusion MRI (DWI) is a noninvasive technique used to detect the diffusion of water molecules in living tissues. It is particularly valuable in the detection of acute brain diseases such as cerebral infarction.

[0003] In the related art, the navigation echo technology cannot effectively judge the quality of the navigation image. If the navigation echo data of a certain excitation is wrong, the final image quality will also be affected. Therefore, it is necessary to propose a new magnetic resonance imaging reconstruction method. Summary of the invention

[0004] The embodiments of this specification are intended to solve at least one of the technical problems in the related art to a certain extent. To this end, the embodiments of this specification propose a magnetic resonance imaging reconstruction method, device, medical imaging equipment and storage medium.

[0005] The present specification provides a magnetic resonance imaging reconstruction method, the method comprising:

[0006] Acquiring diffusion weighted imaging (DWI) data and an estimated image sequence, wherein the diffusion weighted imaging (DWI) data is acquired through multiple imaging sequences, and the estimated image sequence is associated with each excitation of the multiple imaging sequences;

[0007] Based on the signal strength of the estimated image sequence, acquiring the phase weight of each estimated image in the estimated image sequence to obtain a weighted image sequence;

[0008] Reconstruction is performed based on the diffusion weighted imaging DWI data, the estimated image sequence and the weighted image sequence to obtain a target reconstructed image.

[0009] In one embodiment, the reconstructing based on the diffusion weighted imaging DWI data, the estimated image sequence and the weighted image sequence to obtain a target reconstructed image includes:

[0010] Obtaining a navigation image sequence based on the phase of the estimated image sequence and the weight image sequence;

[0011] Reconstruction is performed based on the diffusion weighted imaging DWI data and the navigation image sequence to obtain a target reconstructed image.

[0012] In one embodiment, obtaining the estimated image sequence includes:

[0013] Based on the diffusion weighted imaging DWI data, obtaining DWI sub-data corresponding to each excitation;

[0014] Image reconstruction is performed on the DWI sub-data to obtain the estimated image sequence.

[0015] In one embodiment, obtaining the estimated image sequence includes:

[0016] In the process of acquiring diffusion weighted imaging DWI data, low-resolution acquisition is performed after each excitation to obtain low-resolution DWI data;

[0017] Image reconstruction is performed on the low-resolution DWI data to obtain the estimated image sequence.

[0018] In one embodiment, based on the signal strength information of the estimated image sequence, the phase weight of each estimated image in the estimated image sequence is obtained to obtain a weighted image sequence, including:

[0019] Extracting the signal strength of each estimated image in the estimated image sequence to obtain an intermediate weighted image sequence;

[0020] determining a normalization factor based on the signal intensity of the intermediate weighted image sequence;

[0021] The intermediate weight image sequence and the normalization factor are divided to obtain the weight image sequence.

[0022] In one embodiment, extracting the signal strength of each estimated image in the estimated image sequence to obtain an intermediate weighted image sequence includes:

[0023] Acquire an amplitude image of each estimated image in the estimated image sequence;

[0024] The amplitude image is low-pass filtered to obtain an intermediate weight image sequence.

[0025] In one embodiment, for each pixel position in the intermediate weight image sequence, determining a normalization factor based on signal intensity information of the intermediate weight image sequence includes:

[0026] The maximum value of the signal intensity of the intermediate weighted image sequence is determined as a normalization factor; or,

[0027] The average value of the images with the largest signal intensities in the intermediate weighted image sequence is determined as the normalization factor.

[0028] The present specification provides a magnetic resonance imaging reconstruction device, the device comprising:

[0029] An information acquisition module, used for acquiring diffusion weighted imaging DWI data and an estimated image sequence, wherein the diffusion weighted imaging DWI data is acquired through multiple imaging sequences, and the estimated image sequence is related to each excitation of the multiple imaging sequences;

[0030] An image processing module, used for acquiring a phase weight of each estimated image in the estimated image sequence based on the signal strength of the estimated image sequence, so as to obtain a weighted image sequence;

[0031] An image reconstruction module is used to reconstruct based on the diffusion weighted imaging DWI data, the estimated image sequence and the weighted image sequence to obtain a target reconstructed image.

[0032] An embodiment of the present specification provides a medical imaging device, which includes: a memory, and one or more processors communicatively connected to the memory; the memory stores instructions executable by the one or more processors, and the instructions are executed by the one or more processors to enable the one or more processors to implement the steps of the method described in any of the above embodiments.

[0033] The embodiments of this specification provide a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the method described in any of the above embodiments are implemented.

[0034] An embodiment of the present specification provides a computer program product, wherein the computer program product includes instructions, and when the instructions are executed by a processor of a computer device, the computer device is enabled to perform the steps of the method described in any one of the above embodiments.

[0035] In the above-mentioned implementation manner of the specification, diffusion weighted imaging DWI data and an estimated image sequence are first obtained, wherein the diffusion weighted imaging DWI data is obtained through multiple imaging sequences, and the estimated image sequence is related to each excitation of the multiple imaging sequences. Then, based on the signal intensity of the estimated image sequence, a phase weight is calculated for each estimated image. The greater the signal intensity, the higher the corresponding weight. The phase weight of each estimated image in the estimated image sequence is obtained to obtain a weighted image sequence, which ensures that there is a signal at each position and reduces the influence of tissue information on the results, thereby effectively avoiding signal loss. Finally, reconstruction is performed based on the diffusion weighted imaging DWI data, the estimated image sequence and the weighted image sequence to reduce the occurrence of artifacts, optimize the problem of signal loss during the reconstruction process, and reduce the overestimation of ADC (apparent diffusion coefficient), thereby improving the robustness of diffusion imaging and obtaining a target reconstructed image. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1aA schematic diagram of a magnetic resonance imaging reconstruction method provided in an embodiment of this specification;

[0037] Figure 1b A schematic diagram of the artifact removal result provided in an embodiment of this specification;

[0038] Figure 1c A schematic diagram of the artifact removal result provided in an embodiment of this specification;

[0039] Figure 1d A schematic diagram of improving the problem of blackening of the left liver lobe provided in an embodiment of this specification;

[0040] Figure 1e A schematic diagram of a corresponding apparent diffusion coefficient image provided for an embodiment of this specification;

[0041] Figure 2 A schematic diagram of a process for obtaining a target reconstructed image provided in an embodiment of this specification;

[0042] Figure 3a A schematic diagram of a single shot acquisition trajectory provided for an embodiment of this specification;

[0043] Figure 3b A schematic diagram of a multi-shot acquisition trajectory provided for an embodiment of this specification;

[0044] Figure 4 A schematic diagram of a process for obtaining an estimated image sequence provided in an embodiment of this specification;

[0045] Figure 5 A schematic diagram of a process for obtaining an estimated image sequence provided in an embodiment of this specification;

[0046] Figure 6 A schematic diagram of a process for obtaining a weighted image sequence provided in an embodiment of this specification;

[0047] Figure 7 A schematic diagram of a process for obtaining an intermediate weighted image sequence provided in an embodiment of this specification;

[0048] Figure 8 A schematic diagram of a magnetic resonance imaging reconstruction device provided in an embodiment of the present specification. DETAILED DESCRIPTION

[0049] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0050] Diffusion resonance imaging (DWI) is a noninvasive technique used to detect the diffusion of water molecules in living tissue. DWI is of great value in the detection of acute brain diseases such as cerebral infarction. However, the traditional single-shot echo planar imaging (EPI) sequence has many limitations in DWI applications, such as sensitivity to motion, easy to produce motion artifacts, low imaging resolution and image deformation.

[0051] In the related art, the multi-shot method is used to solve the above problem. The multi-shot method can provide higher imaging resolution and smaller deformation, but it also faces the problem of motion-induced phase disturbance caused by strong diffusion gradients. Since the phase disturbances generated by different excitations are inconsistent, multi-shot diffusion-weighted imaging cannot be directly reconstructed using conventional Fourier transform.

[0052] In summary, methods that correct the phase error between multiple excitations are used to reconstruct high-quality DWI images. For example, the navigator echo technique is used to correct the phase change, that is, it is assumed that the navigator echo shares the same phase change with the previous image signal at each excitation. However, the method using the navigator echo technique affects the accuracy due to the mismatch between the navigator echo image and the target image.

[0053] In the related art, reconstruction methods without navigation echoes, such as MUSE, avoid the mismatch problem. MUSE uses two-step parallel imaging reconstruction, and the first step relies on the SENSE reconstruction method. However, when the number of excitations increases, the first SENSE reconstruction will be inaccurate, thus affecting the accuracy of phase estimation.

[0054] In the related technology, the MUSSELS method based on the low-rank relationship in k-space has also achieved good reconstruction effect without phase estimation.

[0055] If the phase estimation is not accurate enough, artifacts will be generated, and this problem is particularly prominent in abdominal imaging. The abdominal area is easily affected by involuntary movements of the heart and breathing, so after the diffusion gradient is applied, the corresponding phase disturbance will be greater. In addition, there is another common problem with abdominal diffusion (including single excitation and multiple excitation), that is, the signal of the left lobe of the liver is lost. Because this area is close to the heart, the beating of the heart will cause the movement of the liver area, resulting in signal loss of phase.

[0056] To improve this phenomenon, multiple averaging methods can be used. In some averages, the liver is relatively still, which helps to alleviate the problem of signal dephasing, thereby improving the signal loss in other acquisition processes. Since this problem has little impact on low b-values, it will lead to an overestimation of the quantitative parameter ADC. Therefore, a weighted averaging method can be used to give greater weight to voxels with higher signals to effectively suppress signal loss.

[0057] In the related art, for multi-excitation abdominal diffusion imaging, the weight (spatial variation) of each excitation is estimated by navigating the echo amplitude image, where the weight reflects the accuracy of phase estimation and the degree of phase loss. This weighting is integrated into multi-excitation reconstruction to improve image quality.

[0058] In the related art, navigation echo technology cannot effectively judge the quality of navigation images. If there is an error in the navigation echo data in a certain excitation, the final image quality will also be affected. For abdominal multi-excitation reconstruction, the inconsistency between the navigation image and the target image is more significant. Therefore, for abdominal multi-excitation reconstruction, self-estimated phase or k-space low-rank methods may be more suitable. In addition, in single-shot imaging commonly used in clinical practice, weighted averaging often relies on amplitude images, which will lead to noise amplification.

[0059] Based on this, an embodiment of the present specification provides a magnetic resonance imaging reconstruction method. First, diffusion weighted imaging DWI data and an estimated image sequence are obtained, wherein the diffusion weighted imaging DWI data is obtained through multiple imaging sequences, and the estimated image sequence is related to each excitation of the multiple imaging sequences. Then, based on the signal intensity of the estimated image sequence, a phase weight is calculated for each estimated image. The greater the signal intensity, the higher the corresponding weight. The phase weight of each estimated image in the estimated image sequence is obtained to obtain a weighted image sequence, which ensures that there is a signal at each position and reduces the influence of tissue information on the results, thereby effectively avoiding signal loss. Finally, reconstruction is performed based on the diffusion weighted imaging DWI data, the estimated image sequence and the weighted image sequence to reduce the occurrence of artifacts, optimize the problem of signal loss during the reconstruction process, and reduce the overestimation of ADC (apparent diffusion coefficient), thereby improving the robustness of diffusion imaging and obtaining a target reconstructed image.

[0060] For the abdomen, firstly, the MRI device is used to collect m imaging sequences of the abdomen to obtain diffusion weighted imaging DWI data. After each acquisition of m imaging sequences of the abdomen, low-resolution acquisition of m imaging sequences of the abdomen is performed to obtain low-resolution DWI data. Then, based on the low-resolution DWI data, an estimated image sequence is reconstructed, and the estimated image sequence is modulo processed to obtain the amplitude image of each estimated image in the estimated image sequence. Then, the amplitude image is low-pass filtered to obtain an intermediate weight image sequence.

[0061] When processing the intermediate weighted image sequence, each pixel is analyzed and the pixel value of the corresponding position in each image in the intermediate weighted image sequence is taken. In order to ensure that the extracted value is representative, the maximum value in each pixel is selected as the normalization factor. Next, for the values ​​of all pixels in the intermediate weighted image sequence, they are divided by their corresponding normalization factors to obtain the weighted image sequence.

[0062] Let the phase image corresponding to the estimated image sequence be Phi, let the weighted image sequence be w, and combine them to get the navigation image sequence, which is w×exp(Phi*i). Among them, i is the imaginary part. The diffusion weighted imaging DWI data and the navigation image sequence are brought into the SENSE framework for joint solution to obtain the target reconstructed image.

[0063] This specification provides a method for magnetic resonance imaging reconstruction. Figure 1a , the magnetic resonance imaging reconstruction method may include the following steps:

[0064] S110 , acquiring diffusion weighted imaging DWI data and estimating image sequences.

[0065] The diffusion weighted imaging DWI data are acquired through multiple imaging sequences, and the estimated image sequence is related to each excitation of the multiple imaging sequences.

[0066] Specifically, in MRI scanning, the parameters of diffusion-weighted imaging are set, such as diffusion-sensitive direction, b value (diffusion coefficient) and echo time (TE). During data acquisition, the use of specific DWI sequences, such as fast echo planar imaging (EPI), can significantly increase the scanning speed and reduce the influence of motion artifacts. In order to improve the signal-to-noise ratio of the image, multiple imaging sequences can be used for data acquisition to obtain diffusion-weighted imaging DWI data. However, in diffusion-weighted imaging, the phase disturbance of the target tissue will lead to a decrease in image quality. Based on the relevant information of each excitation of multiple imaging sequences, a variety of implementation methods can be used to repair and optimize it to obtain the corresponding estimated image sequence. The estimated image sequence not only reflects the diffusion characteristics of water molecules, but also provides important diagnostic information for clinicians to help identify and evaluate cerebral infarction, tumors and other lesions. Among them, the choice of echo time (TE) has a direct impact on the contrast and signal-to-noise ratio of the image. Generally speaking, a shorter TE can increase the intensity of the signal and is suitable for the detection of acute lesions, while a longer TE may provide better contrast and is suitable for the observation of chronic lesions.

[0067] S120 . Based on the signal strength of the estimated image sequence, obtain the phase weight of each estimated image in the estimated image sequence to obtain a weighted image sequence.

[0068] S130 , reconstructing based on the diffusion weighted imaging DWI data, the estimated image sequence and the weighted image sequence to obtain a target reconstructed image.

[0069] Specifically, first, the signal intensity of the estimated image sequence is analyzed to calculate the intensity distribution of each estimated image in the spatial domain. Based on the intensity distribution of each estimated image in the spatial domain, the phase information of each estimated image is calculated, and the corresponding phase weight is assigned to each estimated image according to the phase information, and finally a weighted image sequence is generated. Next, the diffusion weighted imaging DWI data, the estimated image sequence and the weighted image sequence are combined for reconstruction to improve the signal-to-noise ratio and resolution of the image and obtain the target reconstructed image.

[0070] Please note that Figure 1b , compared to Figure 1b In the left image 102, Figure 1b The target reconstructed image 104 on the right shows a more significant artifact removal effect. Figure 1c , compared to Figure 1c In the middle left image 106, Figure 1c The target reconstructed image 108 on the right shows a more significant artifact removal effect. Figure 1d , compared to Figure 1d In the left image 110, Figure 1d The target reconstructed image 112 on the right side of the middle improves the problem of the darkening of the left liver lobe. Figure 1e , compared to Figure 1e In the middle left image 114, Figure 1e The target reconstructed image 116 on the right side of the center is closer to the real apparent diffusion coefficient.

[0071] In the above implementation, diffusion weighted imaging DWI data and an estimated image sequence are first obtained, wherein the diffusion weighted imaging DWI data is obtained through multiple imaging sequences, and the estimated image sequence is related to each excitation of the multiple imaging sequences. Then, based on the signal intensity of the estimated image sequence, a phase weight is calculated for each estimated image. The greater the signal intensity, the higher the corresponding weight. The phase weight of each estimated image in the estimated image sequence is obtained to obtain a weighted image sequence, which ensures that there is a signal at each position and reduces the influence of tissue information on the results, thereby effectively avoiding signal loss. Finally, reconstruction is performed based on the diffusion weighted imaging DWI data, the estimated image sequence and the weighted image sequence to reduce the occurrence of artifacts, optimize the problem of signal loss during the reconstruction process, and reduce the overestimation of ADC (apparent diffusion coefficient), thereby improving the robustness of diffusion imaging and obtaining a target reconstructed image.

[0072] In some embodiments, see Figure 2, reconstructing based on diffusion weighted imaging DWI data, estimated image sequence and weighted image sequence to obtain a target reconstructed image may include the following steps:

[0073] S210 , obtaining a navigation image sequence based on the phase of the estimation image sequence and the weight image sequence.

[0074] S220 , reconstructing based on diffusion weighted imaging DWI data and navigation image sequence to obtain a target reconstructed image.

[0075] Specifically, the navigation images of different excitations are filtered and compared together. Ideally, the mode map distributions of these images should be very close. If there are significant differences in the mode map distributions of these images, it indicates that some navigation images may have signal loss or low signal-to-noise ratio. Therefore, by performing mode processing on the estimated image sequence, the phase information corresponding to each estimated image is extracted to obtain the phase image sequence corresponding to the estimated image sequence. Then, the phase image sequence corresponding to the estimated image sequence is combined with the weight image sequence to highlight important features, suppress noise, highlight the navigation images with higher signals and more comprehensive, reduce the weights of other navigation images in reconstruction, and obtain the navigation image sequence, thereby avoiding signal loss. Finally, the diffusion weighted imaging DWI data and the navigation image sequence are used as input data for the reconstruction method of the accelerated imaging technology to achieve image reconstruction and obtain the target reconstructed image.

[0076] For example, let the phase image sequence Phi and the weight image sequence w be combined to obtain the navigation image sequence, that is, w×exp(Phi*i). Where i is the imaginary part. The diffusion weighted imaging DWI data and the navigation image sequence are brought into the SENSE framework for joint solution to obtain the target reconstructed image.

[0077] In the above implementation, a navigation image sequence is obtained based on the phase of the estimated image sequence and the weighted image sequence, and reconstruction is performed based on the diffusion weighted imaging DWI data and the navigation image sequence to reduce the occurrence of artifacts, optimize the problem of signal loss during the reconstruction process, and reduce the overestimation of ADC (apparent diffusion coefficient), thereby improving the robustness of diffusion imaging and obtaining a target reconstructed image.

[0078] In some embodiments, acquiring diffusion weighted imaging DWI data may include: acquiring m imaging sequences on the target tissue to obtain diffusion weighted imaging DWI data.

[0079] Wherein, n excitations are performed each time an imaging sequence is acquired, and m and n are positive integers.

[0080] Specifically, determine the values ​​of m and n, and select the appropriate number of imaging times m and excitation times n according to specific clinical needs and equipment performance. In order to improve the signal-to-noise ratio of the image, the acquisition can be repeated multiple times, which is called multiple averaging, that is, repeating several single excitation or multi-excitation acquisitions, which helps to reduce errors through averaging processing and improve the reliability and robustness of the data. Perform m imaging sequences and perform n excitations each time. During the excitation process, record the DWI data of each excitation. The data acquired multiple times is used as diffusion-weighted imaging DWI data.

[0081] In some embodiments, EPI acquisition trajectories are generally used, which can be divided into single-shot and multi-shot modes. Figure 3a , Figure 3a A single shot acquisition trace is shown. Figure 3b , Figure 3b A multi-shot acquisition trajectory is shown.

[0082] In the above implementation, m imaging sequences are collected for the target tissue to obtain diffusion weighted imaging DWI data, which provides a data basis for subsequent determination of the target reconstructed image.

[0083] In some embodiments, see Figure 4 , obtaining the estimated image sequence may include the following steps:

[0084] S410 , based on the diffusion weighted imaging DWI data, obtaining DWI sub-data corresponding to each excitation.

[0085] S420 , reconstructing the DWI sub-data to obtain an estimated image sequence.

[0086] Specifically, the way of implementing phase perturbation will be different according to the different characteristics of the target tissue. These perturbation methods not only affect the imaging effect, but also determine the type of data required. For example, some tissues may require higher spatial resolution. Therefore, after determining the target tissue, it is necessary to clarify the DWI data required. Based on diffusion-weighted imaging DWI data, the DWI sub-data corresponding to each excitation are obtained through excitation pulses in different directions and changing gradient magnetic fields. Then, the DWI sub-data are reconstructed using an appropriate image reconstruction algorithm to obtain an estimated image sequence.

[0087] Exemplarily, based on two independent acquisitions, two excitations are performed during each acquisition process to obtain diffusion weighted imaging DWI data, and for each layer, the diffusion weighted imaging DWI data includes 4 DWI sub-data. The 4 DWI sub-data are image reconstructed to obtain four clinical images.

[0088] Combine clinical images reconstructed from different excitations and acquisitions at the same level together. For example, the size of a single clinical image is 128 × 128. The number of excitations is 2, the number of acquisitions is 2, and the matrix of the stacked clinical image sequence is 128 × 128 × 4.

[0089] In the above implementation, based on the diffusion weighted imaging DWI data, the DWI sub-data corresponding to each excitation are acquired, and image reconstruction is performed on the DWI sub-data to obtain an estimated image sequence.

[0090] The application scenarios are very wide, not only can multiple excitations be achieved, but also single excitation can be flexibly selected. The most suitable phase estimation method can be selected according to specific needs to optimize the imaging process.

[0091] In some embodiments, see Figure 5 , obtaining the estimated image sequence may include the following steps:

[0092] S510 . In the process of acquiring diffusion weighted imaging DWI data, after each excitation, low-resolution acquisition is performed to obtain low-resolution DWI data.

[0093] S520 , reconstructing the low-resolution DWI data to obtain an estimated image sequence.

[0094] Specifically, for target tissues with smaller phase disturbances, the navigation echo method is used to estimate the phase to improve the quality of image reconstruction. The navigation echo uses a known reference signal to correct the phase, thereby reducing image distortion caused by phase disturbances and ensuring the accuracy and clarity of the final reconstructed image. In the process of acquiring diffusion-weighted imaging DWI data, after each excitation, a low-resolution acquisition of the imaging sequence is performed on the target tissue. After completing multiple low-resolution acquisitions of the imaging sequence of the target tissue, low-resolution DWI data is obtained. The low-resolution DWI data is image-reconstructed into a low-resolution image sequence. In the next stage, images from different shots of the same layer with different averages (i.e., the same layer of the scanned target tissue) are combined to obtain an estimated image sequence.

[0095] In the above implementation, in the process of acquiring diffusion weighted imaging DWI data, low-resolution acquisition is performed after each excitation to obtain low-resolution DWI data, and image reconstruction is performed on the low-resolution DWI data to obtain an estimated image sequence, providing a data basis for subsequent determination of the target reconstructed image.

[0096] In some embodiments, see Figure 6 , based on the signal strength information of the estimated image sequence, obtaining the phase weight of each estimated image in the estimated image sequence to obtain the weighted image sequence may include the following steps;

[0097] S610 , extracting the signal strength of each estimated image in the estimated image sequence to obtain an intermediate weighted image sequence.

[0098] S620 , determining a normalization factor based on the signal intensity of the intermediate weight image sequence.

[0099] S630 , taking the quotient of the intermediate weight image sequence and the normalization factor to obtain a weight image sequence.

[0100] Specifically, first, the signal intensity is extracted from each image in the estimated image sequence, and these signal intensity values ​​will be used as the basis for calculation to form an intermediate weighted image sequence. Normalization processing can reduce the intensity difference between different images, thereby providing a basis for subsequent analysis. In order not to affect the numerical range of the final clinical image reconstruction, the intermediate weighted image sequence needs to be normalized. It is necessary to calculate a normalization factor by statistically analyzing the signal intensity data in the intermediate weighted image sequence. The normalization factor is used as a standard to make relative adjustments to the intensity of all pixels in the sequence. After the normalization factor is determined, it is applied to the intermediate weighted image sequence, that is, by taking the quotient of each image in the intermediate weighted image sequence and the normalization factor, it is adjusted to a standard intensity range to obtain a weighted image sequence.

[0101] In the above implementation, the signal strength of each estimated image in the estimated image sequence is extracted to obtain an intermediate weighted image sequence. Based on the signal strength of the intermediate weighted image sequence, a normalization factor is determined. The intermediate weighted image sequence and the normalization factor are divided to obtain a weighted image sequence to improve the image quality and provide a data basis for the subsequent process.

[0102] In some embodiments, see Figure 7 , extracting the signal intensity of each estimated image in the estimated image sequence to obtain an intermediate weighted image sequence may include the following steps:

[0103] S710: Obtain an amplitude image of each estimated image in the estimated image sequence.

[0104] S720 , low-pass filter the amplitude image to obtain an intermediate weight image sequence.

[0105] Specifically, amplitude extraction can find important features in the image, such as edges, textures, and signal intensity. Signal intensity can also reflect the degree of voxel dephasing. Amplitude extraction is achieved by performing modulo processing on the estimated image sequence, and the amplitude image of each estimated image in the estimated image sequence is obtained. The amplitude image is low-pass filtered to retain only the signal intensity information of the amplitude image, remove other detail information such as edges and textures, and obtain an intermediate weight image containing only signal intensity information, thereby avoiding the influence of edges and textures on the normalization factor determined by the subsequent intermediate weight image sequence. In addition, image sequences are usually affected by noise, which may affect the image quality and lead to poor feature extraction. Low-pass filtering can also effectively reduce high-frequency noise and retain low-frequency information, thereby improving the clarity of the image. Therefore, by low-pass filtering the amplitude image of each estimated image in the estimated image sequence, an intermediate weight image sequence is obtained.

[0106] In the above implementation, the amplitude image of each estimated image in the estimated image sequence is obtained, and the amplitude image is low-pass filtered to obtain an intermediate weighted image sequence, thereby effectively reducing noise and enhancing image features.

[0107] In some embodiments, for each pixel position in the intermediate weight image sequence, a normalization factor is determined based on signal intensity information of the intermediate weight image sequence, including: determining the maximum value of the signal intensity of the intermediate weight image sequence as the normalization factor.

[0108] Specifically, for each image in the intermediate weighted image sequence, the signal intensity value of each pixel is first extracted. Next, the maximum signal intensity is found by analyzing the signal intensity values ​​of all pixels in the entire intermediate weighted image sequence. The maximum signal intensity obtained in the intermediate weighted image sequence is determined as the normalization factor.

[0109] Exemplarily, the generation process of the intermediate weight image sequence is based on two independent acquisitions, and two excitations are performed during each acquisition process, so the intermediate weight image sequence contains four images. When processing these images, each pixel is analyzed and the pixel value of the corresponding position in the four images is taken. In order to ensure that the extracted value is representative, the maximum value of each pixel among the four values ​​is selected as the normalization factor. Effectively offset the accidental fluctuations caused by different acquisition conditions or noise, ensuring that the value that best represents the signal intensity at the position is selected.

[0110] In the above implementation, the maximum value of the signal intensity of the intermediate weighted image sequence is determined as the normalization factor to improve the quality of the image and provide a data basis for the subsequent process.

[0111] In some embodiments, for each pixel position in the intermediate weighted image sequence, a normalization factor is determined based on the signal strength information of the intermediate weighted image sequence, including: determining the average of the largest signal strengths in the intermediate weighted image sequence as the normalization factor.

[0112] Specifically, the signal intensity of each image in the intermediate weighted image sequence is first analyzed, that is, the intensity values ​​of each pixel position in all images are compared. For the intermediate weighted image sequence, several pixels with the highest signal intensity in the sequence are selected. The signal intensity values ​​of these pixels are added and the average value is obtained to obtain the normalization factor. The normalization factor represents the overall brightness level of the strongest signal area in the intermediate weighted image sequence.

[0113] Exemplarily, the generation process of the intermediate weight image sequence is based on two independent acquisitions, and two excitations are performed in each acquisition process, so the intermediate weight image sequence contains four images. The average of the first two maximum values ​​of the corresponding position of each pixel in the four images can be taken as the normalization factor.

[0114] In the above implementation, the average value of the largest signal intensities in the intermediate weighted image sequence is determined as the normalization factor to improve the image quality and provide a data basis for the subsequent process.

[0115] This specification provides a magnetic resonance imaging reconstruction device 800, see Figure 8 The magnetic resonance imaging reconstruction device 800 includes: an information acquisition module 810 , an image processing module 820 , and an image reconstruction module 830 .

[0116] An information acquisition module 810 is used to acquire diffusion weighted imaging DWI data and an estimated image sequence, wherein the diffusion weighted imaging DWI data is acquired through multiple imaging sequences, and the estimated image sequence is related to each excitation of the multiple imaging sequences;

[0117] An image processing module 820 is used to obtain a phase weight of each estimated image in the estimated image sequence based on the signal strength of the estimated image sequence to obtain a weighted image sequence;

[0118] The image reconstruction module 830 is used to reconstruct based on the diffusion weighted imaging DWI data, the estimated image sequence and the weighted image sequence to obtain a target reconstructed image.

[0119] For a detailed description of the magnetic resonance imaging reconstruction apparatus, reference may be made to the above description of the magnetic resonance imaging reconstruction method, which will not be repeated here.

[0120] In some embodiments, a medical imaging device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the method steps in the above embodiments when executing the computer program.

[0121] An embodiment of the present specification provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the method in any of the above embodiments are implemented.

[0122] One embodiment of the present specification provides a computer program product, which includes instructions. When the instructions are executed by a processor of a computer device, the computer device can perform the steps of the method of any of the above embodiments.

[0123] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection portion with one or more wirings (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or otherwise processing in a suitable manner if necessary, and then stored in a computer memory.

Claims

1. A magnetic resonance imaging reconstruction method, characterized in that: The method comprises: Acquiring diffusion weighted imaging (DWI) data and an estimated image sequence, wherein the diffusion weighted imaging (DWI) data is acquired through multiple imaging sequences, and the estimated image sequence is associated with each excitation of the multiple imaging sequences; Based on the signal strength of the estimated image sequence, acquiring the phase weight of each estimated image in the estimated image sequence to obtain a weighted image sequence; Reconstruction is performed based on the diffusion weighted imaging DWI data, the estimated image sequence and the weighted image sequence to obtain a target reconstructed image.

2. The method according to claim 1, characterized in that The step of reconstructing based on the diffusion weighted imaging DWI data, the estimated image sequence and the weighted image sequence to obtain a target reconstructed image includes: Obtaining a navigation image sequence based on the phase of the estimated image sequence and the weight image sequence; Reconstruction is performed based on the diffusion weighted imaging DWI data and the navigation image sequence to obtain a target reconstructed image.

3. The method according to claim 1, characterized in that The obtaining of the estimated image sequence comprises: Based on the diffusion weighted imaging DWI data, obtaining DWI sub-data corresponding to each excitation; Image reconstruction is performed on the DWI sub-data to obtain the estimated image sequence.

4. The method according to claim 1, characterized in that: The acquisition of the estimated image sequence includes: in the process of acquiring diffusion weighted imaging DWI data, after each excitation, performing low-resolution acquisition to obtain low-resolution DWI data; Image reconstruction is performed on the low-resolution DWI data to obtain the estimated image sequence.

5. The method according to any one of claims 1 to 4, characterized in that: Based on the signal strength information of the estimated image sequence, acquiring the phase weight of each estimated image in the estimated image sequence to obtain a weighted image sequence, including: Extracting the signal strength of each estimated image in the estimated image sequence to obtain an intermediate weighted image sequence; determining a normalization factor based on the signal intensity of the intermediate weighted image sequence; The intermediate weight image sequence and the normalization factor are divided to obtain the weight image sequence.

6. The method according to claim 5, characterized in that The step of extracting the signal strength of each estimated image in the estimated image sequence to obtain an intermediate weighted image sequence comprises: Acquire an amplitude image of each estimated image in the estimated image sequence; The amplitude image is low-pass filtered to obtain an intermediate weight image sequence.

7. The method according to claim 5, characterized in that For each pixel position in the intermediate weight image sequence, determining a normalization factor based on signal intensity information of the intermediate weight image sequence includes: The maximum value of the signal intensity of the intermediate weighted image sequence is determined as a normalization factor; or, The average value of the images with the largest signal intensities in the intermediate weighted image sequence is determined as the normalization factor.

8. A magnetic resonance imaging reconstruction device, characterized in that: The device comprises: An information acquisition module, used for acquiring diffusion weighted imaging DWI data and an estimated image sequence, wherein the diffusion weighted imaging DWI data is acquired through multiple imaging sequences, and the estimated image sequence is related to each excitation of the multiple imaging sequences; An image processing module, used for acquiring a phase weight of each estimated image in the estimated image sequence based on the signal strength of the estimated image sequence, so as to obtain a weighted image sequence; An image reconstruction module is used to reconstruct based on the diffusion weighted imaging DWI data, the estimated image sequence and the weighted image sequence to obtain a target reconstructed image.

9. A medical imaging device, comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

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

Images