Image reconstruction methods, apparatus and equipment

Abstract

The present application discloses an image reconstruction method, device and equipment, which relates to the field of medical image technology. The image domain data without short-axis aliasing is used to form full-sample k-space data, which can be applied to the de-warping of other narrow bands without pre-scanning the image, saving sampling time and obtaining a stable image de-warping effect. The method includes: obtaining a target narrow band with the long axis as the readout direction during the acquisition of the object image sequence; performing virtual expansion processing on the k-space data in the target narrow band in the short-axis direction to obtain the k-space data of the processed target narrow band; obtaining full-sample k-space data applicable to the reconstruction algorithm based on the k-space data of the processed target narrow band; de-warping the k-space data of other narrow bands except the target narrow band using the full-sample k-space data applicable to the reconstruction algorithm, and reconstructing the image in combination with all the de-warped narrow bands.

Description

Technical Field

[0001] The present application relates to the field of medical image technology, and in particular to an image reconstruction method, device and equipment. Background Art

[0002] Magnetic Resonance Imaging (MRI) is one of the main imaging methods in modern medical imaging. Its basic principle is to use the magnetic resonance phenomenon, use radio frequency excitation to excite hydrogen protons in the human body, use gradient fields to encode spatial positions, then use receiving coils to receive signals with position information, and finally reconstruct image information through Fourier transform.

[0003] Considering that magnetic resonance imaging has excellent soft tissue imaging capabilities, it can provide many unique contrasts for clinical diagnosis. However, due to the slow scanning speed of the spatial position encoding process, it is easily affected by motion, especially in the case of abdominal imaging. If the patient cannot cooperate with breath holding or breathing is unstable, the gate control efficiency will be low, and artifacts will be introduced to the image, affecting subsequent diagnosis. PROPELLER imaging is an MRI data acquisition method that can automatically reduce motion artifacts. It has a special acquisition track, which is composed of multiple narrow band (blade) acquisitions passing through the center of the K space. Multiple blades are evenly distributed to cover the entire k space. Here, each blade is equivalent to a low-resolution k space, with the long axis being the readout and the short axis being the phase encoding direction. The length of the short axis is also called the width of the blade. When applying PROPELLER imaging to scanning the human abdomen, since the structure of the human abdomen is long and flat, when performing a cross-sectional scan, the left-right direction is the long axis and the front-to-back direction is the short axis; when performing a sagittal scan, the front-to-back direction is the short axis and the head-to-foot direction is the long axis; the readout direction rotates with the angle of the blade, and when the phase encoding direction rotates to the long axis, it is easy to cause image warping artifacts.

[0004] In order to solve the problem of image warping artifacts, SENSE imaging technology is usually used to virtually expand the scanning field of view, or increase phase-encoded over-acquisition, so that the warping artifact part in the image can be pushed to the edge of the region of interest. However, SENSE imaging technology is limited by the size of the pre-scan range, and the pre-scan is usually performed separately from the imaging, which may produce a series of artifacts due to the misalignment. Increasing the phase-encoded over-acquisition while ensuring the spatial resolution will increase the scanning time length and will not increase the image signal-to-noise ratio, making the acquired image sequence difficult to use in clinical practice. Therefore, the above two methods are difficult to obtain a stable image dewarping effect, and may even bring more artifacts and scanning time costs. Summary of the invention

[0005] In view of this, the present application provides an image reconstruction method, device and equipment, the main purpose of which is to solve the problem in the prior art that it is difficult to obtain a more stable image unwrapping effect by virtually expanding the scanning field of view or increasing phase encoding over-acquisition using Sense imaging technology.

[0006] According to a first aspect of the present application, there is provided an image reconstruction method, the method comprising:

[0007] Obtaining a narrow band of the target with the major axis being the readout direction during the acquisition of the object image sequence;

[0008] Performing virtual expansion processing on the k-space data in the target narrow band in the short axis direction to obtain the processed k-space data of the target narrow band;

[0009] Acquiring full-sample k-space data applicable to a reconstruction algorithm based on the processed target narrow-band k-space data;

[0010] The k-space data of other narrow bands except the target narrow band are dewarped using the full-sample k-space data of the applicable reconstruction algorithm, and image reconstruction is performed by combining all the dewarped narrow bands.

[0011] Furthermore, the target narrow band whose long axis is the readout direction during the acquisition process of the object image sequence specifically includes:

[0012] By reading the imaging positioning parameters among various parameters applicable to the medical imaging protocol record scanning, the imaging plane and phase encoding direction are obtained;

[0013] According to the imaging plane and the phase encoding direction, a target narrow band with a long axis as a readout direction during acquisition of an object image sequence is determined.

[0014] Furthermore, the target narrowband data is in k-space, and the k-space data in the target narrowband is virtually expanded in the short-axis direction to obtain the processed k-space data of the target narrowband, specifically including:

[0015] Inverse Fourier transforming the k-space data in the target narrowband into the image domain to obtain image domain data in the target narrowband;

[0016] In the image domain, zero-filling or white noise processing is performed on the image domain data in the target narrow band in the short axis direction to expand the field of view and obtain image domain data of the target narrow band in the image domain;

[0017] The image domain data of the target narrow band in the image domain is transformed into k-space by Fourier transform to obtain the processed k-space data of the target narrow band.

[0018] Furthermore, the step of acquiring full-sample k-space data applicable to a reconstruction algorithm based on the processed target narrow-band k-space data specifically includes:

[0019] According to the size of the k-space data in the target narrow band, extracting the central data from the processed k-space data of the target narrow band;

[0020] The central data is rotated to obtain full-sample k-space data suitable for a reconstruction algorithm.

[0021] Furthermore, the extracting center data from the processed k-space data of the target narrow band according to the size of the k-space data in the target narrow band specifically includes:

[0022] Extracting the width of the k-space center according to the size of the k-space data in the target narrowband in the phase encoding direction;

[0023] Based on the width of the k-space center, center data is intercepted from the processed target narrowband k-space data.

[0024] Furthermore, the rotating the central data to obtain full k-space data applicable to the reconstruction algorithm specifically includes:

[0025] Rotating the center data to form k-space data in the center circle area of ​​k-space;

[0026] The k-space data in the k-space central circle area is intercepted and the k-space data in the inscribed square is determined as the fully sampled k-space data applicable to the reconstruction algorithm.

[0027] Further, the step of unwrapping k-space data of other narrow bands except the target narrow band using the full-sampled k-space data of the applicable reconstruction algorithm, and reconstructing an image by combining all the unwrapped narrow bands, specifically includes:

[0028] Using the fully sampled k-space data of the applicable reconstruction algorithm as a training set, calculating the interpolation weight coefficient corresponding to the downsampling interpolation kernel function;

[0029] Convolution interpolation processing is performed on k-space data of other narrow bands except the target narrow band according to the interpolation weight coefficient, and image reconstruction is performed by combining all the deconvolved narrow bands.

[0030] According to a second aspect of the present application, there is provided an image reconstruction device, the device comprising:

[0031] A first acquisition unit is used to acquire a target narrow band with a long axis as a readout direction during acquisition of an object image sequence;

[0032] A processing unit, used for performing virtual expansion processing on the k-space data in the target narrow band in the short axis direction to obtain the processed k-space data of the target narrow band;

[0033] A second acquisition unit is used to acquire full-sample k-space data applicable to a reconstruction algorithm based on the processed target narrow-band k-space data;

[0034] The reconstruction unit is used to unwrap k-space data of other narrow bands except the target narrow band by using the full-sampled k-space data of the applicable reconstruction algorithm, and to reconstruct an image by combining all the unwrap narrow bands.

[0035] Furthermore, the first acquiring unit includes:

[0036] A reading module, used for obtaining an imaging plane and a phase encoding direction by reading imaging positioning parameters from various parameters applicable to the medical imaging protocol record scanning;

[0037] The determination module is used to determine the target narrow band with the long axis as the readout direction during the acquisition of the object image sequence according to the imaging plane and the phase encoding direction.

[0038] Further, the target narrowband data is in k-space, and the processing unit includes:

[0039] A first transformation module, used for inverse Fourier transforming the k-space data in the target narrowband into the image domain to obtain image domain data in the target narrowband;

[0040] A processing module, configured to perform zero-filling or white noise processing on the image domain data in the target narrow band in the image domain in the short axis direction to expand the field of view and obtain image domain data of the target narrow band in the image domain;

[0041] The second transformation module is used to transform the image domain data of the target narrow band in the image domain into k-space by using Fourier transformation to obtain the processed k-space data of the target narrow band.

[0042] Furthermore, the second acquiring unit includes:

[0043] A cutting module, used for cutting central data from the processed k-space data of the target narrow band according to the size of the k-space data in the target narrow band;

[0044] The acquisition module is used to rotate the central data to acquire full-sample k-space data applicable to the reconstruction algorithm.

[0045] Furthermore, the interception module includes:

[0046] An extraction submodule, used for extracting the width of the k-space center according to the size of the k-space data in the target narrowband in the phase encoding direction;

[0047] The interception submodule is used to intercept the center data from the processed target narrowband k-space data based on the width of the k-space center.

[0048] Furthermore, the acquisition module includes:

[0049] A rotation submodule, used for rotating the central data to form k-space data in the central circle area of ​​k-space;

[0050] The determination submodule is used to intercept the k-space data in the inscribed square of the k-space data in the k-space central circle area and determine it as the fully sampled k-space data applicable to the reconstruction algorithm.

[0051] Furthermore, the reconstruction unit comprises:

[0052] A calculation module, used to calculate the interpolation weight coefficient corresponding to the downsampling interpolation kernel function by using the full-sampled k-space data of the applicable reconstruction algorithm as a training set;

[0053] The interpolation module is used to perform convolution interpolation processing on the k-space data of other narrow bands except the target narrow band according to the interpolation weight coefficient, and to reconstruct the image by combining all the deconvolved narrow bands.

[0054] According to a third aspect of the present application, a storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the above-mentioned image reconstruction method is implemented.

[0055] According to a fourth aspect of the present application, an image reconstruction device is provided, comprising a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, wherein the processor implements the above-mentioned image reconstruction method when executing the program.

[0056] By means of the above technical scheme, the present application provides an image reconstruction method, device and equipment, which utilize the characteristics of the abdominal imaging field of view having a long axis and a short axis, obtain a target narrow band with the long axis as the readout direction during the acquisition of the object image sequence, and perform virtual expansion processing on the k-space data in the target narrow band in the short axis direction to obtain the k-space data of the processed target narrow band, and further obtain full-sample k-space data applicable to the reconstruction algorithm based on the processed k-space data of the target narrow band, and use the full-sample k-space data applicable to the reconstruction algorithm to de-warp the k-space data of other narrow bands except the target narrow band, and combine all the de-warped narrow bands for image reconstruction. Compared with the current methods of using SENSE imaging technology to virtually expand the scanning field of view, or increase phase encoding over-acquisition to solve the image warping artifacts, when the phase encoding direction is turned to the long axis, there is no material that can generate signals in the short axis, and this part of the data is very close to the real image domain data. The image domain data without short axis aliasing is used to form full-sample k-space data, which can be applied to other narrow-band dewarping without pre-scanning the image, saving sampling time and obtaining a stable image dewarping effect.

[0057] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0059] Figure 1 A schematic diagram of a process of an image reconstruction method provided in an embodiment of the present application is shown;

[0060] Figure 2 A schematic diagram showing a flow chart of another image reconstruction method provided in an embodiment of the present application is shown;

[0061] Figure 3 A schematic diagram of the structure of an image reconstruction device provided in an embodiment of the present application is shown;

[0062] Figure 4 A schematic structural diagram of another image reconstruction device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0063] The present application will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other without conflict.

[0064] In the related art, in order to solve the problem of image warping artifacts, SENSE imaging technology is usually used to virtually expand the scanning field of view, or increase phase-encoded over-acquisition, so that the warping artifact part in the image can be pushed to the edge of the region of interest. However, SENSE imaging technology is limited by the size of the pre-scan range, and the pre-scan is usually performed separately from the imaging, which may produce a series of artifacts due to the misalignment; increasing the phase-encoded over-acquisition while ensuring the spatial resolution will increase the scanning time length and will not increase the image signal-to-noise ratio, making the acquired image sequence difficult to use in clinical practice. Therefore, the above two methods are difficult to obtain a stable image dewarping effect, and may even bring more artifacts and scanning time costs.

[0065] In order to solve this problem, this embodiment provides an image reconstruction method, such as Figure 1 As shown, the method can be applied to the server of the medical platform, including the following steps:

[0066] 101. During the acquisition process of the object image sequence, a target narrow band whose long axis is the readout direction is obtained.

[0067] The object is a scanning object of magnetic resonance imaging, usually a part of the human body. In the specific scanning process, a series of radio frequencies and gradients can be used to spatially encode the imaged object and receive signals to acquire an image sequence of the corresponding body part. Here, the PROPELLER imaging method can be used in the acquisition process of the object image sequence. This method uses the data of repeated sampling in the central area of ​​k-space to estimate the motion information of the object during the acquisition process and compensate for it. In the specific acquisition process, multiple narrow bands covering the center of k-space can be set, and each narrow band is composed of L parallel lines, wherein the direction of the parallel lines can be the frequency encoding direction, and the arrangement direction of the parallel lines can be the phase encoding direction. Each narrow band can reconstruct a low-resolution image. Each narrow band rotates continuously in the k-space at a certain angle until all the k-space data is acquired. Since each narrow band rotation process is accompanied by a data acquisition process, the target narrow band with the long axis as the readout direction can be determined according to the acquisition information of the image sequence during the data acquisition process. Here, the acquisition information includes the size of the acquired image, the phase encoding direction of the acquired image, etc. In the k-space, since the narrow bands are evenly distributed, the number of narrow bands required for the image acquisition process and the narrow band position for the start of acquisition can be determined according to the acquisition information, and the target narrow band with the long axis as the readout direction is calculated to acquire.

[0068] It can be understood that, taking the axial image as an example, the long axis is equivalent to the left-right direction of the human body, the short axis is the front-back direction of the human body, and the readout direction refers to the frequency encoding direction of spatial encoding (including frequency and phase encoding), which is usually accompanied by ADC sampling of the magnetic resonance signal. Since the digital sampling frequency can be very high, many times higher than the bandwidth of the received signal, a higher frequency is usually used for sampling, such as 2x frequency sampling, to obtain the collected data. At this time, the scanning field of view needs to be increased synchronously.

[0069] The executor of the embodiment of the present invention may be an image reconstruction device, which may be specifically configured at the service end of a medical platform. By acquiring the target narrow band in the long-axis readout direction during the acquisition process of the object image sequence, taking into account the special prior information of the human body structure, the target narrow band read out in the long-axis direction has twice the over-acquisition in the phase encoding direction. Therefore, when the phase encoding is transferred to the long-axis direction, the non-aliased data on the short axis may be used for dewarping, and the dewarping of other narrow bands can be applied without pre-scanning the image, thereby improving the image dewarping effect.

[0070] 102. Perform virtual expansion processing on the k-space data in the target narrow band in the short-axis direction to obtain processed k-space data of the target narrow band.

[0071] It is understandable that magnetic resonance images are usually sampled at a higher frequency. Compared with the field of view used in the phase encoding direction, a larger scanning field of view is used in the frequency encoding direction, such as a 2-fold scanning field of view. At this time, the image domain data in the target narrow band is field of view*2·field of view. In order to avoid over-acquisition in the phase encoding direction, the prior information of the human body structure is taken into consideration. The k-space data in the target narrow band can be virtually expanded in the image domain to expand the field of view. The image domain data in the target narrow band can be processed into 2·field of view*2·field of view to achieve a 2-fold scanning field of view in the phase encoding direction, saving the acquisition time in the phase encoding direction.

[0072] The virtual expansion processing here can be zero-filling or white noise processing for the data in the image domain, which is not limited here. The specific processing process can be to retain the intermediate data in the image domain, symmetrically fill the data on both sides with zeros, or to fill zeros around the center point of the image domain as the origin, or to add white noise with a mean of zero on the basis of zero filling.

[0073] 103. Acquire full-sample k-space data applicable to a reconstruction algorithm according to the processed target narrow-band k-space data.

[0074] Since the processed target narrow-band k-space data is acquired in the long-axis readout direction, it has a natural over-acquisition characteristic, that is, a 2-fold scanning field of view is used in the frequency encoding direction, while a 1-fold scanning field of view is used in the short-axis readout direction in the phase encoding direction. The acquired k-space data has a non-aliasing characteristic, which is very close to the real oversampled k-space data. This part of the k-space data can be used to create a virtual expansion of the scanning field of view, and then the k-space data in other narrow bands can be unwrapped.

[0075] It can be understood that the k-space data of the target narrowband after processing here is equivalent to the virtually over-sampled k-space data. Taking into account the effect of applying the k-space data to deconvolution, the k-space data collected in the central area of ​​the k-space can be selected for the k-space data of the target narrowband after processing as the fully sampled k-space data suitable for the reconstruction algorithm. The central area in the k-space here can be a square area intercepted by the corresponding width of the target narrowband in the k-space, and an area intercepted in the k-space that is smaller than the corresponding width of the target narrowband can also be selected, which is not limited here.

[0076] The reconstruction algorithm here refers to GRAPPA, which is a parallel imaging algorithm that uses the spatial sensitivity information of multi-channel coils to restore a complete image from the downsampled signal, effectively reducing the acquisition time.

[0077] 104. Dewarping k-space data of other narrow bands except the target narrow band using the fully sampled k-space data of the applicable reconstruction algorithm, and performing image reconstruction by combining all the dewarped narrow bands.

[0078] It can be understood that since the full-sampling k-space data comes from the data collected by the imaging itself, it is more stable and will not have artifact problems caused by staggered layers. Using the full-sampling k-space data to de-warp the k-space data of other narrow bands except the target narrow band has a more stable image de-warping effect. The specific de-warping process uses the full-sampling k-space data as a training set to obtain the interpolation weight coefficient from the known data to the target data, and further uses the interpolation weight coefficient to obtain a certain linear relationship between the sampled data points, that is, any data point can be obtained by linear fitting of some adjacent data points, thereby calculating the missing data rows and reconstructing the image through the square sum.

[0079] Related Technology When an imaging object is spatially encoded, it is usually hoped that the spatial state of the object does not change. Since the change in state will cause the phase and amplitude of the encoded signal to change, for an image sequence with a slower acquisition, movement may occur during the acquisition process, affecting the image quality. The image reconstruction method provided in the embodiment of the present application utilizes the characteristics of the abdominal imaging field of view having a long axis and a short axis, obtains a target narrow band with a long axis as the readout direction during the acquisition of the object image sequence, and performs virtual expansion processing on the k-space data in the target narrow band in the short axis direction to obtain the k-space data of the processed target narrow band, and further obtains the full-sampled k-space data applicable to the reconstruction algorithm based on the processed target narrow band k-space data, and uses the full-sampled k-space data applicable to the reconstruction algorithm to unwrap the k-space data of other narrow bands except the target narrow band, and combines all the unwrapped narrow bands for image reconstruction. Compared with the current methods of using SENSE imaging technology to virtually expand the scanning field of view, or increase phase encoding over-acquisition to solve the image warping artifacts, when the phase encoding direction is turned to the long axis, there is no material that can generate signals in the short axis, and this part of the data is very close to the real image domain data. The image domain data without short axis aliasing is used to form full-sample k-space data, which can be applied to other narrow-band dewarping without pre-scanning the image, saving sampling time and obtaining a stable image dewarping effect.

[0080] Further, as a refinement and extension of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, this embodiment provides another image reconstruction method, such as Figure 2 As shown, the method includes:

[0081] 201. The imaging plane and phase encoding direction are obtained by reading the imaging positioning parameters among various parameters applicable to the medical imaging protocol record scanning.

[0082] Typically, imaging software converts the relevant parameters collected by the medical device into documents and stores them in a medical imaging protocol. The medical imaging protocol records various parameters during the scanning process of the medical device, including imaging positioning parameters such as imaging plane, phase encoding direction and frequency encoding direction, scanning field of view, etc.

[0083] 202. Determine, according to the imaging plane and the phase encoding direction, a target narrow band whose long axis is a readout direction during acquisition of an object image sequence.

[0084] Since the phase encoding direction affects the scanning time to a certain extent, the more phase encoding direction is, the longer the scanning time is. Generally, the phase encoding direction is selected in the direction of the short anatomical diameter to save scanning time. In the actual k-space data acquisition process, different phase encoding directions will be selected for scanning different parts and different tomographic planes. In order to minimize the artifacts in the scanning process, the anterior-posterior phase encoding direction is usually selected as the long-axis readout direction, and in the process of acquiring k-space data, the long axis is selected as the target narrow band of the readout direction.

[0085] It can be understood that in the process of scanning images using medical equipment, the phase encoding direction is also a continuous acquisition process. Let the phase encoding in the long-axis readout direction be i, and each narrow band corresponds to a phase encoding. The matrix size of the k-space data acquired each time is scanMatrix, and the width is BladeWidth. Here, the matrix size is defined by the user and can be read from the medical imaging protocol.

[0086] 203. Perform virtual expansion processing on the k-space data in the target narrow band in the short-axis direction to obtain processed k-space data of the target narrow band.

[0087] Here, the target narrowband data is in k-space. Considering that the k-space data acquired in the long-axis readout direction uses a 2x scanning field of view in the frequency encoding direction and a 1x scanning field of view in the phase encoding direction, the k-space data acquired at this time has a non-aliasing characteristic. It is possible to virtually create a 2x scanning field of view in the phase encoding direction to obtain the over-acquired k-space data, and then unwrap the other narrowband k-space data to improve the image reconstruction effect. Specifically, the k-space data in the target narrowband can be inverse Fourier transformed to the image domain to obtain the image domain data in the target narrowband. Then, the image domain data in the target narrowband can be filled with zeros or white noise in the short-axis direction in the image domain to expand the field of view and obtain the image domain data of the target narrowband in the image domain. The image domain data of the target narrowband in the image domain can then be transformed to k-space using Fourier transform to obtain the processed k-space data of the target narrowband.

[0088] It should be noted that, in the process of expanding the field of view by filling zeros or performing white noise processing in the short axis direction on the image domain data in the target narrow band in the image domain, the specific multiple of expanding the field of view needs to ensure that the scanning field of view used in the long axis readout direction and the scanning field of view used in the phase encoding direction are the same. That is to say, for a 2-fold scanning field of view in the long axis readout direction and a 1-fold scanning field of view in the phase encoding direction, the scanning field of view in the phase encoding direction can be virtually expanded to 2 times the scanning field of view. Of course, the scanning field of view used in the long axis direction and the scanning field of view used in the phase encoding direction can also be virtually expanded synchronously. Similarly, for a 2-fold scanning field of view in the long axis readout direction and a 1-fold scanning field of view in the phase encoding direction, the scanning field of view in the long axis readout direction can be virtually expanded from 2 times to 3 times, while the scanning field of view in the phase encoding direction can be virtually expanded to 3 times the scanning field of view.

[0089] Specifically, during the k-space data acquisition process, when the i-th narrow band is acquired, that is, when the long axis is acquired as the readout direction, the matrix size of the acquired k-space data is 2scanMatrix*BladeWidth, and the acquired k-space data is inverse Fourier transformed into the image domain. After 2x zero-filling or white noise processing is performed in the image domain to the scanning field of view, the matrix size of the image domain data is 2scanMatrix*2BladeWidth, and then Fourier transform is used to acquire the processed k-space data.

[0090] 204. According to the size of the k-space data in the target narrow band, extract central data from the processed k-space data of the target narrow band.

[0091] Since the data in the processed target narrowband is in k-space, in the automatic calibration reconstruction, the automatic calibration signal is collected at the Nyquist rate, and the central data is generally intercepted from the middle position area of ​​the k-space. Specifically, the width of the k-space center can be extracted according to the size of the middle k-space data of the target narrowband in the phase encoding direction. The center width can be the width of the target narrowband, and the center data is intercepted from the k-space data of the processed target narrowband based on the width of the k-space center.

[0092] Here, k-space data with a virtual over-acquisition k-space center size of 2BladeWidth*2BladeWidth can be intercepted as the center data. In order to ensure the unwrapping effect of the k-space data, k-space data with a size smaller than 2BladeWidth*2BladeWidth can also be intercepted as the center data.

[0093] 205. Rotate the central data to obtain full-sample k-space data applicable to a reconstruction algorithm.

[0094] It can be understood that since the image acquisition trajectory covers the entire k-space, the stability of the k-space data in the edge area of ​​the central data is relatively poor. In order to select k-space data with a better dewarping effect, the central data can be rotated to form k-space data in the central circle area of ​​the k-space, and the k-space data in the inscribed square of the k-space data in the central circle area of ​​the k-space can be intercepted to determine the full-sampling k-space data suitable for the reconstruction algorithm.

[0095] Here, after the central data with a size of 2BladeWidth*2BladeWidth is cut out, it can be rotated. The rotation angle corresponds to the remaining k-space data in the narrow band that needs to be unwrapped, and after the rotation, the k-space data with a size of floor(2BladeWidth / sqr(2)) is cut out as the full-sample k-space data.

[0096] 206. Using the fully sampled k-space data of the applicable reconstruction algorithm as a training set, calculate the interpolation weight coefficient corresponding to the downsampling interpolation kernel function.

[0097] It should be noted that the relative position relationship between the uncollected data at different positions in the k-space and the adjacent fully collected data is variable. Theoretically, for the process of sampling the uncollected data at each position, a set of corresponding weighting coefficients is required according to the positions of the surrounding fully collected data. When obtaining this set of weighting coefficients, it is necessary to construct a sufficient number of fitting data combinations with the same relative position relationship to establish an overdetermined set of linear equations. The coefficients of this set of linear equations constitute the interpolation weight coefficients corresponding to the downsampling interpolation kernel function.

[0098] 207. Perform convolution interpolation processing on k-space data of other narrow bands except the target narrow band according to the interpolation weight coefficient, and perform image reconstruction by combining all the deconvolved narrow bands.

[0099] It can be understood that the full-sampled k-space data of the reconstruction algorithm applied here is used as the full-sampled data to obtain the automatic calibration signal, and the k-space data of other narrowbands except the target narrowband is used as the under-sampled data. The weight coefficient of the under-sampled data is estimated using the full-sampled data to obtain the interpolation weight coefficient of the under-sampled data, and the interpolation weight coefficient of the under-sampled data is further used to reconstruct the k-space data of other narrowbands, and the reconstructed multi-channel images are merged to obtain the final image with the dewarping effect.

[0100] Further, as Figure 1-Figure 2 The specific implementation of the method, the embodiment of the present application provides an image reconstruction device, such as Figure 3 As shown, the device includes: a first acquisition unit 31, a processing unit 32, a second acquisition unit 33, and a reconstruction unit 34.

[0101] A first acquisition unit 31 is used to acquire a target narrow band with a long axis in a readout direction during acquisition of an object image sequence;

[0102] A processing unit 32 is used to perform virtual expansion processing on the k-space data in the target narrow band in the short axis direction to obtain the processed k-space data of the target narrow band;

[0103] A second acquisition unit 33 is used to acquire full-sample k-space data applicable to a reconstruction algorithm based on the processed target narrow-band k-space data;

[0104] The reconstruction unit 34 is used to de-warp the k-space data of other narrow bands except the target narrow band by using the full-sampled k-space data of the applicable reconstruction algorithm, and to reconstruct the image by combining all the de-warped narrow bands.

[0105] The image reconstruction device provided by the embodiment of the present invention utilizes the characteristics of the abdominal imaging field of view having a long axis and a short axis, obtains the target narrow band with the long axis as the readout direction during the acquisition of the object image sequence, and performs virtual expansion processing on the k-space data in the target narrow band in the short axis direction to obtain the k-space data of the processed target narrow band, and further obtains the full-sampled k-space data applicable to the reconstruction algorithm based on the k-space data of the processed target narrow band, and uses the full-sampled k-space data applicable to the reconstruction algorithm to de-warp the k-space data of other narrow bands except the target narrow band, and combines all the de-warped narrow bands to reconstruct the image. Compared with the existing method of using SENSE imaging technology to virtually expand the scanning field of view or increase the over-acquisition of phase encoding to solve the warping artifact of the image, since when the phase encoding direction is turned to the long axis, there is no material that can generate signals in the short axis, and this part of the data is very close to the real image domain data, and the full-sampled k-space data is formed by using the image domain data without aliasing of the short axis, which can be applied to the de-warping of other narrow bands without pre-scanning the image, saving sampling time and obtaining a stable image de-warping effect.

[0106] In specific application scenarios, such as Figure 4 As shown, the first acquisition unit 31 includes:

[0107] The reading module 311 may be used to obtain the imaging plane and the phase encoding direction by reading the imaging positioning parameters among various parameters applicable to the medical imaging protocol record scanning;

[0108] The determination module 312 may be configured to determine, according to the imaging plane and the phase encoding direction, a target narrow band with a long axis as a readout direction during acquisition of an object image sequence.

[0109] In specific application scenarios, such as Figure 4As shown, the target narrowband data is in k-space, and the processing unit 32 includes:

[0110] The first transformation module 321 may be used to perform an inverse Fourier transformation of the k-space data in the target narrowband into the image domain to obtain image domain data in the target narrowband;

[0111] The processing module 322 may be used to perform zero-filling or white noise processing on the image domain data in the target narrow band in the image domain in the short axis direction to expand the field of view and obtain the image domain data of the target narrow band in the image domain;

[0112] The second transformation module 323 may be used to transform the image domain data of the target narrow band in the image domain into k-space by using Fourier transformation to obtain processed k-space data of the target narrow band.

[0113] In specific application scenarios, such as Figure 4 As shown, the second acquisition unit 33 includes:

[0114] The interception module 331 may be used to intercept the center data from the processed k-space data of the target narrow band according to the size of the k-space data in the target narrow band;

[0115] The acquisition module 332 may be used to rotate the central data to acquire full-sampled k-space data applicable to a reconstruction algorithm.

[0116] In specific application scenarios, such as Figure 4 As shown, the interception module 331 includes:

[0117] The extraction submodule 3311 may be used to extract the width of the k-space center according to the size of the k-space data in the target narrowband in the phase encoding direction;

[0118] The interception submodule 3312 may be used to intercept center data from the processed target narrowband k-space data based on the width of the k-space center.

[0119] In specific application scenarios, such as Figure 4 As shown, the acquisition module 332 includes:

[0120] The rotation submodule 3321 may be used to rotate the center data to form k-space data in the center circle area of ​​the k-space;

[0121] The determination submodule 3322 may be used to intercept the k-space data within the inscribed square of the k-space data in the k-space central circle area, and determine it as the fully sampled k-space data applicable to the reconstruction algorithm.

[0122] In specific application scenarios, such as Figure 4 As shown, the reconstruction unit 34 includes:

[0123] The calculation module 341 may be used to calculate the interpolation weight coefficient corresponding to the downsampling interpolation kernel function by using the full-sampled k-space data of the applicable reconstruction algorithm as a training set;

[0124] The interpolation module 342 may be configured to perform convolution interpolation processing on k-space data of other narrow bands except the target narrow band according to the interpolation weight coefficient, and perform image reconstruction by combining all the deconvolved narrow bands.

[0125] It should be noted that for other corresponding descriptions of the functional units involved in the image reconstruction device applicable to the server side provided in this embodiment, reference can be made to Figure 1 and Figure 2 The corresponding description in will not be repeated here.

[0126] Based on the above Figure 1-Figure 2 The method shown in the embodiment of the present application accordingly provides a storage medium on which a computer program is stored, and when the program is executed by a processor, the above-mentioned Figure 1-Figure 2 The image reconstruction method shown;

[0127] Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present application.

[0128] Based on the above Figure 1-Figure 2 The method shown, and Figure 3-Figure 4 In order to achieve the above-mentioned purpose, the embodiment of the present application further provides a server-side entity device, which can be a computer, a server, or other network devices, etc. The entity device includes a storage medium and a processor; the storage medium is used to store a computer program; the processor is used to execute the computer program to achieve the above-mentioned Figure 1-Figure 2 The image reconstruction method shown.

[0129] Optionally, the above-mentioned physical devices may also include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a WI-FI module, etc. The user interface may include a display screen (Display), an input unit such as a keyboard (Keyboard), etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (such as a WI-FI interface), etc.

[0130] Those skilled in the art will appreciate that the physical device structure for image reconstruction provided in this embodiment does not constitute a limitation on the physical device, and may include more or fewer components, or a combination of certain components, or different arrangements of components.

[0131] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages the hardware and software resources of the physical device for image reconstruction, and supports the operation of information processing programs and other software and / or programs. The network communication module is used to achieve communication between components within the storage medium, and communication with other hardware and software in the information processing physical device.

[0132] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented by means of software plus a necessary general hardware platform, or by hardware. By applying the technical solution of the present application, compared with the current existing methods, the present application uses short-axis non-aliased image domain data to form full-sample k-space data, which can be applied to other narrowband deconvolution without pre-scanning images, saving sampling time while obtaining a stable image deconvolution effect.

[0133] Those skilled in the art will appreciate that the accompanying drawings are only schematic diagrams of a preferred implementation scenario, and the modules or processes in the accompanying drawings are not necessarily necessary for implementing the present application. Those skilled in the art will appreciate that the modules in the devices in the implementation scenario can be distributed in the devices of the implementation scenario according to the description of the implementation scenario, or can be changed accordingly and located in one or more devices different from the present implementation scenario. The modules of the above-mentioned implementation scenario can be combined into one module, or can be further split into multiple submodules.

[0134] The above serial numbers of this application are only for description and do not represent the advantages and disadvantages of the implementation scenarios. The above disclosure is only a few specific implementation scenarios of this application, but this application is not limited to them, and any changes that can be thought of by technicians in this field should fall within the scope of protection of this application.

Claims

1. An image reconstruction method, characterized in that: include: During the acquisition of the object image sequence, the long axis is the target narrow band in the readout direction. When performing a cross-sectional scan, the left-right direction is the long axis and the front-to-back direction is the short axis. When performing a sagittal scan, the front-to-back direction is the short axis and the head-to-foot direction is the long axis. The readout direction rotates with the angle of the narrow band. When the phase encoding direction rotates to the long axis, it is easy to cause image warping artifacts. The k-space data in the target narrow band is virtually expanded in the short-axis direction to obtain the processed k-space data of the target narrow band, and the field of view in the short-axis direction is expanded. The virtual expansion is to fill the data in the image domain with zero or perform white noise processing, specifically retaining the middle data in the image domain, symmetrically filling the data on both sides with zero, or filling the surroundings with zero with the center point of the image domain as the origin; Acquiring full-sample k-space data applicable to a reconstruction algorithm based on the processed target narrow-band k-space data; The k-space data of other narrow bands except the target narrow band are dewarped using the full-sample k-space data of the applicable reconstruction algorithm, and image reconstruction is performed by combining all the dewarped narrow bands.

2. The method according to claim 1, characterized in that The process of acquiring the target narrow band whose long axis is the readout direction in the process of acquiring the object image sequence specifically includes: By reading the imaging positioning parameters among various parameters applicable to the medical imaging protocol record scanning, the imaging plane and phase encoding direction are obtained; According to the imaging plane and the phase encoding direction, a target narrow band with a long axis as a readout direction during acquisition of an object image sequence is determined.

3. The method according to claim 1, characterized in that The target narrow band data is in k-space, and the k-space data in the target narrow band is virtually expanded in the short axis direction to obtain the processed k-space data of the target narrow band, specifically including: Inverse Fourier transforming the k-space data in the target narrowband into the image domain to obtain image domain data in the target narrowband; In the image domain, zero-filling or white noise processing is performed on the image domain data in the target narrow band in the short axis direction to expand the field of view and obtain image domain data of the target narrow band in the image domain; The image domain data of the target narrow band in the image domain is transformed into k-space by Fourier transform to obtain the processed k-space data of the target narrow band.

4. The method according to claim 1, characterized in that: Determining the full-sample k-space data applicable to the reconstruction algorithm according to the processed target narrow-band k-space data specifically includes: According to the size of the k-space data in the target narrow band, extracting the central data from the processed k-space data of the target narrow band; The central data is rotated to obtain full-sample k-space data suitable for a reconstruction algorithm.

5. The method according to claim 4, characterized in that The extracting center data from the processed k-space data of the target narrow band according to the size of the k-space data in the target narrow band specifically includes: Extracting the width of the k-space center according to the size of the k-space data in the target narrowband in the phase encoding direction; Based on the width of the k-space center, center data is intercepted from the processed target narrowband k-space data.

6. The method according to claim 4, characterized in that The step of rotating the central data to obtain full k-space data applicable to the reconstruction algorithm specifically includes: Rotating the center data to form k-space data in the center circle area of ​​k-space; The k-space data in the k-space central circle area is intercepted and the k-space data in the inscribed square is determined as the fully sampled k-space data applicable to the reconstruction algorithm.

7. The method according to any one of claims 1 to 6, characterized in that The step of unwrapping k-space data of other narrow bands except the target narrow band by using the full-sampled k-space data of the applicable reconstruction algorithm, and reconstructing the image by combining all the unwrapped narrow bands, specifically includes: Using the fully sampled k-space data of the applicable reconstruction algorithm as a training set, calculating the interpolation weight coefficient corresponding to the downsampling interpolation kernel function; Convolution interpolation processing is performed on k-space data of other narrow bands except the target narrow band according to the interpolation weight coefficient, and image reconstruction is performed by combining all the deconvolved narrow bands.

8. An image reconstruction device, characterized in that: include: The first acquisition unit is used to acquire a target narrow band with the long axis as the readout direction during the acquisition of the object image sequence. When performing a cross-sectional scan, the left-right direction is the long axis and the front-back direction is the short axis. When performing a sagittal scan, the front-back direction is the short axis and the head-foot direction is the long axis. The readout direction rotates with the angle of the narrow band. When the phase encoding direction rotates to the long axis, it is easy to cause a warping artifact of the image. A processing unit is used to perform virtual expansion processing on the k-space data in the target narrow band in the short axis direction to obtain the processed k-space data of the target narrow band and expand the field of view in the short axis direction, wherein the virtual expansion is to perform zero filling or white noise processing on the data in the image domain, specifically retaining the middle data in the image domain, symmetrically filling the data on both sides with zeros, or filling zeros around the center point of the image domain as the origin; A second acquisition unit is used to acquire full-sample k-space data applicable to a reconstruction algorithm based on the processed target narrow-band k-space data; The reconstruction unit is used to unwrap k-space data of other narrow bands except the target narrow band by using the full-sampled k-space data of the applicable reconstruction algorithm, and to reconstruct an image by combining all the unwrap narrow bands.

9. A storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the image reconstruction method according to any one of claims 1 to 7 is implemented.

10. An image reconstruction device, comprising a storage medium, a processor, and a computer program stored in the storage medium and executable on the processor, characterized in that: When the processor executes the program, the image reconstruction method according to any one of claims 1 to 7 is implemented.

Images