WO2020114329A1 - 磁共振快速参数成像方法及装置 - Google Patents

磁共振快速参数成像方法及装置 Download PDF

Info

Publication number
WO2020114329A1
WO2020114329A1 PCT/CN2019/122002 CN2019122002W WO2020114329A1 WO 2020114329 A1 WO2020114329 A1 WO 2020114329A1 CN 2019122002 W CN2019122002 W CN 2019122002W WO 2020114329 A1 WO2020114329 A1 WO 2020114329A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
parameter
magnetic resonance
target
space data
Prior art date
Application number
PCT/CN2019/122002
Other languages
English (en)
French (fr)
Inventor
刘元元
朱燕杰
梁栋
程静
刘新
郑海荣
Original Assignee
深圳先进技术研究院
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳先进技术研究院 filed Critical 深圳先进技术研究院
Publication of WO2020114329A1 publication Critical patent/WO2020114329A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R33/00Arrangements or instruments for measuring magnetic variables
    • G01R33/20Arrangements or instruments for measuring magnetic variables involving magnetic resonance
    • G01R33/44Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
    • G01R33/48NMR imaging systems
    • G01R33/54Signal processing systems, e.g. using pulse sequences ; Generation or control of pulse sequences; Operator console
    • G01R33/56Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R33/00Arrangements or instruments for measuring magnetic variables
    • G01R33/20Arrangements or instruments for measuring magnetic variables involving magnetic resonance
    • G01R33/44Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
    • G01R33/48NMR imaging systems
    • G01R33/4818MR characterised by data acquisition along a specific k-space trajectory or by the temporal order of k-space coverage, e.g. centric or segmented coverage of k-space

Definitions

  • Embodiments of the present invention relate to the technical field of magnetic resonance imaging, and in particular, to a method and device for rapid magnetic resonance imaging.
  • Magnetic resonance parametric imaging has become an important quantitative imaging tool because it can characterize some inherent information of tissues.
  • MRI parametric imaging uses a parameter-dependent model to fit the signal evolution in the acquired weighted image pixel by pixel to obtain the corresponding parameter estimates. Because magnetic resonance parametric imaging needs to acquire multiple contrast-weighted magnetic resonance images, the scan time is often very long, which has become a major bottleneck restricting its rapid clinical development.
  • the purpose of the embodiments of the present invention is to provide a method and apparatus for rapid magnetic resonance parameter imaging to improve the accuracy of the parameter map.
  • an embodiment of the present invention provides a method for rapid parametric imaging of magnetic resonance, the method includes: acquiring a template according to a variable rate and a density, acquiring under-acquisition K-space data corresponding to a magnetic resonance image; Reconstructed reconstruction model of K-space data; solve the reconstruction model, perform signal compensation based on a preset parameter relaxation model during the solution process, and reconstruct a parameter-weighted image from the under-recovered K-space data; The parameter relaxation model fits the reconstructed parameter-weighted image to obtain a parameter map, where the parameter map includes parameter values of various tissues in the magnetic resonance image.
  • an embodiment of the present invention also provides a magnetic resonance fast parameter imaging device, which includes a data acquisition module, a model building module, an image reconstruction module, and an image fitting module.
  • the data collection module is used for collecting templates according to the variable rate and density to collect the under-acquisition K-space data corresponding to the magnetic resonance image
  • the model building module is used to establish a reconstruction model for reconstructing the under-acquisition K-space data
  • the image reconstruction module It is used to solve the reconstructed model, perform signal compensation based on a preset parameter relaxation model during the solution process, and reconstruct a parameter-weighted image from the under-collected K-space data
  • an image fitting module is used to use the The parameter relaxation model fits the reconstructed parameter-weighted image to obtain a parameter map, where the parameter map includes parameter values of various tissues in the magnetic resonance image.
  • a method and device for rapid parameter imaging of magnetic resonance provided by the embodiments of the present invention, first, acquire templates according to the variable rate and density, collect the K-space data corresponding to the magnetic resonance image, and establish the K Spatial data to reconstruct the reconstructed model; then, the reconstructed model is solved, and signal compensation is performed based on the preset parameter relaxation model during the solution process, so as to reconstruct the parameter-weighted image from the under-collected K-space data; finally, The parameter relaxation model is used to fit the reconstructed parameter-weighted image to obtain the parameter map.
  • the embodiment of the present invention introduces signal compensation during the reconstruction process, which can accurately reconstruct the parameter-weighted image from the under-acquired K-space data, further fit the parameter map, and improve the accuracy of the parameter map .
  • FIG. 1 shows a block schematic diagram of an electronic device provided by an embodiment of the present invention.
  • FIG. 2 shows a flowchart of a magnetic resonance fast parameter imaging method provided by an embodiment of the present invention.
  • FIG. 3 is a flowchart of sub-steps of step S105 shown in FIG. 2.
  • FIG. 4 shows a block schematic diagram of a magnetic resonance fast parameter imaging device provided by an embodiment of the present invention.
  • Icons 100-electronic equipment; 101-processor; 102-memory; 103-bus; 104-communication interface; 200-magnetic resonance fast parameter imaging device; 201-setting module; 202-template generation module; 203-data acquisition module ; 204-model building module; 205-image reconstruction module; 206-image fitting module.
  • FIG. 1 shows a block schematic diagram of an electronic device 100 provided by an embodiment of the present invention.
  • the electronic device 100 can be in communication with the magnetic resonance instrument, and the electronic device 100 can reconstruct the parameter map according to the under-acquired K-space data corresponding to the magnetic resonance image collected by the magnetic resonance scanner.
  • the electronic device 100 may be, but not limited to, a notebook computer, a desktop computer, a server, a portable computer, and so on.
  • the electronic device 100 includes a processor 101, a memory 102, a bus 103, and a communication interface 104, and the processor 101, the memory 102, and the communication interface 104 are connected through the bus 103.
  • the memory 102 may include a high-speed random access memory (RAM: Random Access Memory), or may also include a non-volatile memory (non-volatile memory), for example, at least one disk memory.
  • RAM Random Access Memory
  • non-volatile memory non-volatile memory
  • the electronic device 100 realizes the communication connection between the electronic device 100 and the magnetic resonance instrument through at least one communication interface 104 (which may be wired or wireless).
  • the memory 102 is used to store programs, such as the magnetic resonance fast parameter imaging apparatus 200 shown in FIG. 4.
  • the magnetic resonance fast parameter imaging apparatus 200 includes at least one software function module that can be stored in the memory 102 in the form of software or firmware or solidified in the operating system of the electronic device 100.
  • the processor 101 may execute the program stored in the memory 102 to implement the magnetic resonance fast parameter imaging method disclosed in the following embodiments.
  • the processor 101 may be an integrated circuit chip with signal processing capabilities for executing executable modules stored in the memory 102, such as a computer program. During execution, the steps of the magnetic resonance positive contrast imaging method may be processed by The instructions in the form of integrated logic circuits of hardware or software in the device 101 are completed.
  • the processor 101 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP) or an application specific integrated circuit (ASIC) ), ready-made programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components.
  • CPU central processing unit
  • NP network processor
  • ASIC application specific integrated circuit
  • FPGA ready-made programmable gate array
  • the bus 103 may be an ISA bus, a PCI bus, an EISA bus, or the like. In FIG. 1, only one bidirectional arrow is used, but it does not mean that there is only one bus or one type of bus.
  • An embodiment of the present invention also provides a computer-readable storage medium on which a computer program is stored.
  • the computer program is executed by the processor 101, the magnetic resonance fast parameter imaging method disclosed in the following embodiments is implemented.
  • FIG. 2 shows a flowchart of a magnetic resonance fast parameter imaging method provided by an embodiment of the present invention.
  • the magnetic resonance fast parametric imaging method includes the following steps:
  • Step S101 Set a plurality of parameter time points at which the magnetic resonance instrument collects magnetic resonance images, and determine a data reduction rate corresponding to each parameter time point.
  • the magnetic resonance parametric imaging may include longitudinal relaxation (T 1 ), transverse relaxation (T 2 ), and longitudinal relaxation in the rotating coordinate system (spin-lattice relaxation in the rotating frame, T 1 ⁇ ), etc.
  • T 1 imaging, T 2 imaging and T 1 ⁇ imaging parameters The directions are inversion time (TI), echo time (TE), and spin-lock time (TSL) respectively.
  • T 1 ⁇ imaging and TSL are used for subsequent imaging. Give an example at a point in time.
  • the number of parameter time points is the number of magnetic resonance images to be acquired in the direction of the magnetic resonance parameter imaging parameter.
  • T 1 ⁇ imaging assuming that N magnetic resonance images need to be acquired, then Set N TSL time points, the user can flexibly set multiple parameter time points according to the type of magnetic resonance parameter imaging and the number of magnetic resonance images to be acquired.
  • Step S102 Generate a variable rate variable density acquisition template according to the data reduction rate corresponding to each parameter time point and the data acquisition strategy of full acquisition in the frequency encoding direction and variable density acquisition in the phase encoding direction.
  • variable rate variable density collection template is generated according to the data collection strategy of full-frequency sampling in the frequency encoding direction and variable-density sampling in the phase encoding direction.
  • the variable-rate and variable-density acquisition template includes multiple parameter time points, the data reduction rate corresponding to each parameter time point, and the data acquisition strategy of full-frequency acquisition in the frequency encoding direction and variable-density acquisition in the phase encoding direction at each parameter time point.
  • Step S103 Acquire the K-space data corresponding to the magnetic resonance image according to the variable rate and density acquisition template.
  • the frequency encoding direction is fully acquired, and the phase encoding direction is variable density acquisition to obtain each
  • the data at the parameter time point can be accelerated by using the variable rate and density acquisition template to accelerate the scanning speed.
  • all magnetic resonance image data is arranged into a space-parameter matrix according to the parameter direction, that is, the serial number of each parameter time point, and each column of the space-parameter matrix represents a certain parameter time
  • the magnetic resonance image data collected at the point, the space-parameter matrix is the magnetic resonance image data corresponding to the under-collected K-space data.
  • Step S104 Establish a reconstruction model for reconstructing under-collected K-space data.
  • an embodiment of the present invention based on a low-rank plus sparse model (L+S) applied to magnetic resonance dynamic imaging, an embodiment of the present invention establishes a reconstruction model for reconstructing under-collected K-space data.
  • the reconstructed model can be expressed as:
  • 1 represents the l 1 norm
  • C( ⁇ ) is an operation operator, which represents the pixel-level signal compensation of the magnetic resonance image
  • X represents the magnetic resonance image to be reconstructed, and it is expressed as the size Is a matrix of voxel numbers ⁇ parameter time points (for example, TSL time points);
  • L represents the low-rank matrix of the magnetic resonance image, that is, the low-rank part of the magnetic resonance image expressed in matrix form; S represents the sparse matrix of the magnetic resonance image , That is, the magnetic resonance image and the residual image of the low rank part;
  • E represents the multi-channel coil coding matrix, which is equal to the product of the under-Fourier operator and the sensitivity matrix of the coil;
  • Rank(L) represents the low rank matrix L Of rank.
  • Step S105 Solve the reconstructed model, perform signal compensation based on a preset parameter relaxation model during the solution process, and reconstruct a parameter-weighted image from the under-recovered K-space data.
  • the preset parameter relaxation model can be expressed as:
  • M represents the image intensity of the magnetic resonance image corresponding to each parameter time point
  • M 0 is the image intensity of the equilibrium state, which represents the balanced image intensity obtained without the parameter pulse (for example, spin-lock pulse)
  • TSL k represents the k-th parameter time point
  • N represents the number of parameter time points
  • T 1 ⁇ represents the parameter graph.
  • signal compensation can be specifically expressed as multiplying each pixel in the image by a compensation coefficient, and the calculation formula of the compensation coefficient can be expressed as:
  • Coef represents the compensation coefficient
  • step S105 may include a pre-processing sub-step, a first updating sub-step and a first iteration sub-step, the pre-processing sub
  • the step is substep S1051
  • the first update substep includes substeps S1052 to S1054
  • the first iteration substep includes substep S1055, which will be described in detail below:
  • Sub-step S1051 pre-process the under-collected K-space data to obtain the first compensation coefficient and the first image details.
  • the full-collected K-space center data of the under-collected K-space data is subjected to inverse Fourier transform to convert the full-collected K-space center data to the image domain to obtain the initial Image; then, use the preset parameter relaxation model to fit the initial image to obtain the first parameter map, represented by T 1 ⁇ 1 ; then substitute the first parameter map into the preset compensation coefficient calculation formula to calculate the first
  • the compensation coefficient is expressed by Coef 1 ; at the same time, according to the initial image, the first image details are set to the empty set, that is
  • the reconstruction model that is, the parameter-weighted image is reconstructed from the under-collected K-space data, which is used in the process of rebuilding the parameter-weighted image
  • the parameter map is added as a constraint to the parameter-weighted image reconstruction.
  • each iteration will update the parameter map according to the newly reconstructed parameter-weighted image and parameter relaxation model, and use the updated parameter map for signal compensation in the next iteration.
  • sub-step S1052 inverse Fourier transform is performed on the under-collected K-space data to obtain the target magnetic resonance image, and the first compensation coefficient is used to compensate the target magnetic resonance image to obtain the first reference image.
  • the parameter graph in i iterations is represented by T 1 ⁇ i
  • the compensation coefficient is represented by Coef i
  • the image details are represented by feat i .
  • the target magnetic resonance image needs to be compensated according to the compensation coefficient, that is, Where X i represents the target magnetic resonance image, Represents the compensated image.
  • X i represents the target magnetic resonance image
  • inverse Fourier transform is performed on the under-collected K-space data to obtain the target magnetic resonance image
  • X 1 is compensated using the first compensation coefficient Coef 1 to obtain the first reference image
  • the first reference image is reconstructed using the first compensation coefficient and the first image details to obtain a target parameter weighted image.
  • the first reference image includes a low-rank component (L) and a sparse component (Sparse component).
  • L low-rank component
  • Sparse component sparse component
  • the low rank of the first reference image needs to be determined.
  • Part L performs singular value threshold operation and sparse part S performs soft threshold operation to obtain iteratively updated low-rank part L and sparse part S, and then sums the updated low-rank part L and sparse part S to obtain update After the target magnetic resonance image.
  • the process of reconstructing the first reference image may include a second update sub-step and a second iteration sub-step, where,
  • the second update substep includes:
  • the first reference image is divided into a low-rank part L and a sparse part S;
  • the auxiliary sparse matrix is obtained, and the auxiliary sparse matrix is used to perform a singular value threshold operation on the low rank part of the first reference image to obtain the target low rank matrix.
  • SVT( ⁇ ) represents the singular value threshold operator, which is defined as:
  • U and V are the matrix of left and right singular value vectors respectively
  • V H represents the conjugate transpose of V
  • is the diagonal composed of the singular values of M Matrix
  • ⁇ ⁇ ( ⁇ ) means that the largest singular value in ⁇ remains unchanged, and all others are 0.
  • Rank(L) 1.
  • the third step is to determine whether the first image detail is an empty set.
  • use the target low rank matrix and the first image detail to perform a soft threshold operation on the sparse part of the first reference image to obtain the target sparse matrix.
  • the soft threshold operation is performed on the sparse part S of the first reference image, and the process of obtaining the target sparse matrix S j can be expressed by the following formula:
  • ST( ⁇ ) represents the soft threshold operation operator, which is defined as:
  • p is an element of the image matrix
  • v is the threshold, the value of which is linearly related to the value in the first image detail.
  • the fourth step is to obtain the second reference image according to the target low rank matrix and the target sparse matrix. This process can be expressed by the following formula:
  • Equation (8) represents the second parameter image
  • L j represents the target low rank matrix
  • S j represents the target sparse matrix
  • E * represents the inverse operation of E, that is, Equation (8) is after inverse Fourier transform of the multi-channel coil K-space data Then the coils are combined to obtain the second reference image.
  • the inverse process of performing signal compensation on the second reference image using the first compensation coefficient to obtain an auxiliary parameter weighted image In the jth iteration, the process of reconstructing the second reference image using the first compensation coefficient can be used The following expression:
  • C -1 ( ⁇ ) means that The image is based on each pixel divided by the compensation coefficient Coef i , X i j represents the auxiliary parameter weighted image, Indicates the second reference image.
  • the second iteration substep includes:
  • the iteration termination condition can be It means that the number of iterations reaches a preset threshold or the reconstruction error is less than a certain preset value.
  • the preset threshold is the set number of iterations J, which can be flexibly set by the user according to the actual situation, and is not limited here.
  • the process of reconstructing the first reference image may include:
  • the reconstructed target parameter weighted image is used as the target magnetic resonance image, and the first reference image is obtained by using sub-step S1052 After that, right
  • sub-step S1054 the image is weighted according to the target parameter to determine the second compensation coefficient and the details of the second image.
  • a preset parameter relaxation model is used to fit the target parameter-weighted image to obtain a second parameter map; then, the second parameter map Substitute the preset compensation coefficient calculation formula to calculate the second compensation coefficient; and then use an iterative feature extraction operator (iterative feature refinement, IFR) to extract image details from the target parameter-weighted image to obtain second image details.
  • IFR iterative feature refinement
  • Sub-step S1055 using the reconstructed target parameter weighted image as the target magnetic resonance image, replacing the first compensation coefficient and the first image detail with the second compensation coefficient and the second image detail, respectively, and performing the first update sub-step until the iteration is reached
  • This process may include:
  • step S106 a parameter relaxation model is used to fit the reconstructed parameter-weighted image to obtain a parameter map, where the parameter map includes parameter values of various tissues in the magnetic resonance image.
  • variable-rate and variable-density acquisition module is used to collect under-collected K-space data, which accelerates the scanning speed
  • the method of signal compensation is introduced in the process of parameter-weighted image reconstruction, and combined with the reconstruction model that reconstructs the under-collected K-space data, the parameter-weighted image can be accurately reconstructed from the under-collected K-space data; at the same time, the parameter weighting
  • the fitted parameter map is used, and the parameter map is added to the reconstruction as a constraint to improve the reconstruction accuracy;
  • the existing model-based rapid magnetic resonance parametric imaging method can only use some simpler parameter relaxation models.
  • the selection of parameter relaxation models in the embodiments of the present invention is more flexible and lenient, and is more practical;
  • an embodiment of the present invention adds an iterative detail extraction operator to extract image details from the residual image, and performs a soft threshold operation on the sparse part of the first reference image according to the value of the image details, which can better remove artifacts .
  • FIG. 4 shows a block diagram of a magnetic resonance fast parameter imaging apparatus 200 provided by an embodiment of the present invention.
  • the magnetic resonance rapid parameter imaging apparatus 200 includes a setting module 201, a template generation module 202, a data acquisition module 203, a model building module 204, an image reconstruction module 205, and an image fitting module 206.
  • the setting module 201 is used to set a plurality of parameter time points for the magnetic resonance imager to acquire magnetic resonance images, and determine a data reduction rate corresponding to each parameter time point.
  • the template generation module 202 is used to generate a variable rate variable density acquisition template according to the data reduction rate corresponding to each parameter time point, and the data acquisition strategy of full acquisition in the frequency encoding direction and variable density acquisition in the phase encoding direction.
  • the data collection module 203 is used to collect templates according to the variable rate and density, and collect the under-capture K-space data corresponding to the magnetic resonance image.
  • the model building module 204 is used to build a reconstruction model for reconstructing the under-collected K-space data.
  • the image reconstruction module 205 is used to solve the reconstructed model, perform signal compensation based on a preset parameter relaxation model during the solution process, and reconstruct a parameter-weighted image from the under-recovered K-space data.
  • the image reconstruction module 205 executes a solution to the reconstruction model, performs signal compensation based on a preset parameter relaxation model during the solution process, and reconstructs a parameter-weighted image from the under-recovered K-space data , Including the pre-processing sub-step, the first updating sub-step and the first iteration sub-step;
  • the pre-processing sub-steps include: pre-processing the under-collected K-space data to obtain the first compensation coefficient and the first image details;
  • the first update sub-step includes: performing inverse Fourier transform on the under-recovered K-space data to obtain a target magnetic resonance image, and compensating the target magnetic resonance image using a first compensation coefficient to obtain a first reference image; using the first The compensation coefficient and the first image details are reconstructed from the first reference image to obtain the target parameter weighted image; according to the target parameter weighted image, the second compensation coefficient and the second image details are determined;
  • the first iteration sub-step includes: using the reconstructed target parameter weighted image as the target magnetic resonance image, replacing the first compensation coefficient and the first image detail with the second compensation coefficient and the second image detail, respectively, and executing the first update sub Step until the iteration termination condition is reached, and the finally reconstructed target parameter-weighted image is taken as the parameter-weighted image.
  • the manner in which the image reconstruction module 205 performs the pre-processing sub-steps includes: taking the full-capture K-space center data of the under-capture K-space data and performing an inverse Fourier transform to obtain an initial image; using preset parameters The relaxation model fits the initial image to obtain the first parameter map; the first parameter map is substituted into a preset compensation coefficient calculation formula to calculate the first compensation coefficient; according to the initial image, the details of the first image are set to an empty set.
  • the image reconstruction module 205 uses the first compensation coefficient and the first image details in the first update sub-step to reconstruct the first reference image to obtain the target parameter-weighted image, including the second update sub Step and second iteration sub-step, where,
  • the second update sub-step includes: dividing the first reference image into a low-rank part and a sparse part; acquiring an auxiliary sparse matrix, and using the auxiliary sparse matrix to perform a singular value threshold operation on the low-rank part of the first reference image to obtain a low target Rank matrix; when the first image detail is not an empty set, use the target low rank matrix and the first image detail to perform soft threshold operation on the sparse part of the first reference image to obtain the target sparse matrix; according to the target low rank matrix and the target sparse Matrix to obtain the second reference image; using the first compensation coefficient to perform the signal inverse process of the second reference image to obtain the auxiliary parameter weighted image;
  • the second iteration sub-step includes: replacing the auxiliary sparse matrix with the target sparse matrix, replacing the first reference image with the second reference image, and performing the second update sub-step until the iteration termination condition is reached, using the auxiliary parameter weighted image obtained as the target Parameter weighted image.
  • the image reconstruction module 205 executes the first update sub-step to weight the image according to the target parameter, and determines the second compensation coefficient and the second image details, including: using a preset parameter relaxation model to target Fit the parameter-weighted image to obtain the second parameter map; substitute the second parameter map into the preset compensation coefficient calculation formula to calculate the second compensation coefficient; use the iterative detail extraction operator to perform image detail extraction on the target parameter-weighted image To get the second image details.
  • the image fitting module 206 is used to fit the reconstructed parameter-weighted image using the parameter relaxation model to obtain a parameter map, where the parameter map includes parameter values of various tissues in the magnetic resonance image.
  • a method and apparatus for rapid parametric magnetic resonance imaging include: acquiring templates according to variable-rate and variable-density acquisition of under-acquisition K-space data corresponding to magnetic resonance images; Reconstructed reconstruction model using K-space data; solve the reconstruction model, perform signal compensation based on a preset parameter relaxation model during the solution process, and reconstruct a parameter-weighted image from the under-recovered K-space data; use parameter relaxation
  • the Yu model fits the reconstructed parameter-weighted image to obtain a parameter map, where the parameter map includes parameter values of various tissues in the magnetic resonance image.
  • the embodiment of the present invention introduces signal compensation during the reconstruction process, which can accurately reconstruct the parameter-weighted image from the under-acquired K-space data, further fit the parameter map, and improve the accuracy of the parameter map .
  • each block in the flowchart or block diagram may represent a module, program segment, or part of code that contains one or more of the Executable instructions.
  • the functions noted in the block may occur out of the order noted in the figures. For example, two consecutive blocks can actually be executed substantially in parallel, and sometimes they can also be executed in reverse order, depending on the functions involved.
  • each block in the block diagrams and/or flowcharts, and combinations of blocks in the block diagrams and/or flowcharts can be implemented with dedicated hardware-based systems that perform specified functions or actions Or, it can be realized by a combination of dedicated hardware and computer instructions.
  • the functional modules in the embodiments of the present invention may be integrated together to form an independent part, or each module may exist alone, or two or more modules may be integrated to form an independent part.
  • the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium.
  • the technical solution of the present invention essentially or part of the contribution to the existing technology or part of the technical solution can be embodied in the form of a software product
  • the computer software product is stored in a storage medium, including Several instructions are used to enable a computer device (which may be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in the embodiments of the present invention.
  • the aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program code .

Landscapes

  • Physics & Mathematics (AREA)
  • High Energy & Nuclear Physics (AREA)
  • Condensed Matter Physics & Semiconductors (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Radiology & Medical Imaging (AREA)
  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Magnetic Resonance Imaging Apparatus (AREA)

Abstract

一种磁共振快速参数成像方法及装置,包括:按照变速率变密度采集模板,采集磁共振图像对应的欠采K空间数据(S103);建立对欠采K空间数据进行重建的重建模型(S104);对重建模型进行求解,在求解过程中基于预设的参数弛豫模型进行信号补偿,从欠采K空间数据中重建出参数加权图像(S105);利用参数弛豫模型对重建出的参数加权图像进行拟合得到参数图,其中,参数图包括磁共振图像中各种组织的参数值(S106)。与现有技术相比,本方法在重建过程中引入了信号补偿,可以从欠采K空间数据中精确的重建出参数加权图像,进一步拟合出参数图,提高了参数图的准确度。

Description

磁共振快速参数成像方法及装置 技术领域
本发明实施例涉及磁共振成像技术领域,具体而言,涉及一种磁共振快速参数成像方法及装置。
背景技术
磁共振参数成像由于其能表征组织的一些固有信息,已成为一种重要的定量成像工具。磁共振参数成像通过一个参数相关的模型,对采集到的加权图像中的信号演变进行逐个像素的拟合,就可以得到相应的参数估计。由于磁共振参数成像需要采集多幅对比度加权的磁共振图像,因此其扫描时间往往很长,这成为制约其在临床上快速发展的一大瓶颈。
现有的快速成像方法主要是通过采集少量的相位编码线来减少扫描时间的,这些方法通过发掘磁共振图像或K空间数据之间的冗余来得到没有伪影的参数图,其拟合得到的参数图的准确度不足。
发明内容
本发明实施例的目的在于提供一种磁共振快速参数成像方法及装置,用以提高参数图的准确度。
为了实现上述目的,本发明实施例采用的技术方案如下:
第一方面,本发明实施例提供了一种磁共振快速参数成像方法,所述方法包括:按照变速率变密度采集模板,采集磁共振图像对应的欠采K空间数据; 建立对所述欠采K空间数据进行重建的重建模型;对所述重建模型进行求解,在求解过程中基于预设的参数弛豫模型进行信号补偿,从所述欠采K空间数据中重建出参数加权图像;利用所述参数弛豫模型对重建出的参数加权图像进行拟合得到参数图,其中,参数图包括磁共振图像中各种组织的参数值。
第二方面,本发明实施例还提供了一种磁共振快速参数成像装置,所述装置包括数据采集模块、模型建立模块、图像重建模块及图像拟合模块。其中,数据采集模块用于按照变速率变密度采集模板,采集磁共振图像对应的欠采K空间数据;模型建立模块用于建立对所述欠采K空间数据进行重建的重建模型;图像重建模块用于对所述重建模型进行求解,在求解过程中基于预设的参数弛豫模型进行信号补偿,从所述欠采K空间数据中重建出参数加权图像;图像拟合模块用于利用所述参数弛豫模型对重建出的参数加权图像进行拟合得到参数图,其中,参数图包括磁共振图像中各种组织的参数值。
相对现有技术,本发明实施例提供的一种磁共振快速参数成像方法及装置,首先,按照变速率变密度采集模板,采集磁共振图像对应的欠采K空间数据,并建立对欠采K空间数据进行重建的重建模型;然后,对重建模型进行求解,并在求解过程中基于预设的参数弛豫模型进行信号补偿,以此从欠采K空间数据中重建出参数加权图像;最后,利用参数弛豫模型对重建出的参数加权图像进行拟合得到参数图。与现有技术相比,本发明实施例在重建过程中引入了信号补偿,可以从欠采K空间数据中精确的重建出参数加权图像,进一步拟合出参数图,提高了参数图的准确度。
为使本发明的上述目的、特征和优点能更明显易懂,下文特举较佳实施例,并配合所附附图,作详细说明如下。
附图说明
为了更清楚地说明本发明实施例的技术方案,下面将对实施例中所需要使用的附图作简单地介绍,应当理解,以下附图仅示出了本发明的某些实施例,因此不应被看作是对范围的限定,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图获得其他相关的附图。
图1示出了本发明实施例提供的电子设备的方框示意图。
图2示出了本发明实施例提供的磁共振快速参数成像方法流程图。
图3为图2示出的步骤S105的子步骤流程图。
图4示出了本发明实施例提供的磁共振快速参数成像装置的方框示意图。
图标:100-电子设备;101-处理器;102-存储器;103-总线;104-通信接口;200-磁共振快速参数成像装置;201-设置模块;202-模板生成模块;203-数据采集模块;204-模型建立模块;205-图像重建模块;206-图像拟合模块。
具体实施方式
下面将结合本发明实施例中附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例仅仅是本发明一部分实施例,而不是全部的实施例。通常在此处附图中描述和示出的本发明实施例的组件可以以各种不同的配置来布置和设计。因此,以下对在附图中提供的本发明的实施例的详细描述并非旨在限制要求保护的本发明的范围,而是仅仅表示本发明的选定实施例。基于本发明的实施例,本领域技术人员在没有做出创造性劳动的前提下所获得的所有其他实施例,都属于本发明保护的范围。
应注意到:相似的标号和字母在下面的附图中表示类似项,因此,一旦某一项在一个附图中被定义,则在随后的附图中不需要对其进行进一步定义和解释。同时,在本发明的描述中,术语“第一”、“第二”等仅用于区分描述,而不 能理解为指示或暗示相对重要性。
请参照图1,图1示出了本发明实施例提供的电子设备100的方框示意图。电子设备100可以与磁共振仪通信连接,电子设备100可以依据磁共振扫描仪采集的磁共振图像对应的欠采K空间数据重建出参数图。电子设备100可以是,但不限于笔记本电脑、台式机、服务器、便携计算机等等。电子设备100包括处理器101、存储器102、总线103和通信接口104,处理器101、存储器102和通信接口104通过总线103连接。
存储器102可能包括高速随机存取存储器(RAM:Random Access Memory),也可能还包括非不稳定的存储器(non-volatile memory),例如至少一个磁盘存储器。电子设备100通过至少一个通信接口104(可以是有线或者无线)实现该电子设备100与磁共振仪之间的通信连接。
存储器102用于存储程序,例如图4所示的磁共振快速参数成像装置200。磁共振快速参数成像装置200包括至少一个能以软件或固件(firmware)的形式存储于存储器102中或固化在电子设备100的操作系统中的软件功能模块。处理器101可以在接收到执行指令后,执行存储器102中存储的程序以实现下述实施例揭示的磁共振快速参数成像方法。
处理器101可能是一种集成电路芯片,具有信号的处理能力,用于执行存储器102中存储的可执行模块,例如计算机程序,在执行过程中,磁共振正对比成像方法的各步骤可以通过处理器101中的硬件的集成逻辑电路或者软件形式的指令完成。处理器101可以是通用处理器,包括中央处理器(Central Processing Unit,简称CPU)、网络处理器(Network Processor,简称NP)等;还可以是数字信号处理器(DSP)、专用集成电路(ASIC)、现成可编程门阵列(FPGA)或者其它可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬 件组件。
总线103可以是ISA总线、PCI总线或EISA总线等。图1中仅用一个双向箭头表示,但并不表示仅有一根总线或一种类型的总线。
本发明实施例还提供了一种计算机可读存储介质,其上存储有计算机程序,计算机程序被处理器101执行时实现下述实施例揭示的磁共振快速参数成像方法。
第一实施例
请参照图2,图2示出了本发明实施例提供的磁共振快速参数成像方法流程图。磁共振快速参数成像方法包括以下步骤:
步骤S101,设置磁共振仪采集磁共振图像的多个参数时间点,并确定每个参数时间点对应的数据降采率。
在本发明实施例中,磁共振参数成像可以包括纵向弛豫(longitudinal relaxation,T 1)、横向弛豫(transverse relaxation,T 2)、旋转坐标系下的纵向弛豫(spin-lattice relaxation in the rotating frame,T )等,每一种磁共振参数成像均有各自对应的参数方向,参数方向的一些变化可以表示病理组织的改变,例如,T 1成像、T 2成像和T 成像的参数方向分别为反转时间(inversion time,TI)、回波时间(echo time,TE)、自旋-锁时间(spin-lock time,TSL),本发明实施例后续均以T 成像及其TSL时间点进行举例。
一个参数时间点采集一个磁共振图像,参数时间点的个数就是磁共振参数成像参数方向需要采集的磁共振图像的个数,以T 成像为例,假设需要采集N个磁共振图像,则设定N个TSL时间点,用户可以根据磁共振参数成像的类型和需要采集的磁共振图像个数,灵活设置多个参数时间点。
设置好多个参数时间点之后,还需要确定每个参数时间点对应的数据降采 率,数据降采率的设置原则可以是:较小的参数时间点,采用较小的加速倍数进行采集;较大的参数时间点,采用较大的加速倍数进行采集。例如,以T 成像为例,TSL参数方向需要采集的图像的个数为N,则设定第i=1,2,…,N个TSL参数时间点的数据降采率为R i,R i是变化的。
步骤S102,根据每个参数时间点对应的数据降采率,以及频率编码方向全采、相位编码方向变密度采集的数据采集策略,生成变速率变密度采集模板。
在本发明实施例中,确定每个参数时间点对应的数据降采率之后,按照频率编码方向全采、相位编码方向变密度采集的数据采集策略,生成变速率变密度采集模板。变速率变密度采集模板包括多个参数时间点、每个参数时间点对应的数据降采率、以及每个参数时间点均频率编码方向全采、相位编码方向变密度采集的数据采集策略。
步骤S103,按照变速率变密度采集模板,采集磁共振图像对应的欠采K空间数据。
在本发明实施例中,得到变速率变密度采集模板之后,在每个参数时间点,按照该参数时间点对应的数据降采率频率编码方向全采、相位编码方向变密度采集,得到每个参数时间点的数据,采用变速率变密度采集模板进行加速采集,可以加快扫描速度。
采集到每个参数时间点的数据之后,将所有磁共振图像数据按照参数方向即每个参数时间点的序号,排成一个空间-参数矩阵,该空间-参数矩阵的每一列表示某一参数时间点采集到的磁共振图像数据,该空间-参数矩阵即为欠采K空间数据所对应的磁共振图像数据。
步骤S104,建立对欠采K空间数据进行重建的重建模型。
在本发明实施例中,基于应用于磁共振动态成像的低秩加稀疏模型 (low-rank plus sparse,L+S),本发明实施例建立对欠采K空间数据进行重建的重建模型,该重建模型可以表示为:
min {X,L,S}||S|| 1s.t.C(X)=L+S,E(X)=d,Rank(L)=1    (1)
其中,||·|| 1表示l 1范数,C(·)是一个操作算子,表示对磁共振图像进行像素级的信号补偿,X表示要重建的磁共振图像,且其表示成大小为体素数×参数时间点数(例如,TSL时间点数)的矩阵;L表示磁共振图像的低秩矩阵,即,用矩阵形式表示的磁共振图像的低秩部分;S表示磁共振图像的稀疏矩阵,即,磁共振图像和低秩部分的残差图像;E表示多通道线圈编码矩阵,其等于欠采傅里叶算子与线圈的敏感度矩阵的乘积;Rank(L)表示低秩矩阵L的秩。
步骤S105,对重建模型进行求解,在求解过程中基于预设的参数弛豫模型进行信号补偿,从所述欠采K空间数据中重建出参数加权图像。
在本发明实施例中,预设的参数弛豫模型可以表示为:
M=M 0exp(-TSL k/T ) k=1,2,...,N     (2)
其中,M表示每个参数时间点对应的磁共振图像的图像强度;M 0为平衡态的图像强度,其表示不带参数脉冲(例如,自旋-锁脉冲)情况下得到的平衡图像强度;TSL k表示第k个参数时间点,N表示参数时间点的个数,T 表示参数图。利用公式(2)对磁共振图像中的所有像素进行非线性拟合,即可得到磁共振图像中各种组织的参数值,也就是得到了参数图。
基于预设的参数弛豫模型,信号补偿可以具体表示为将图像中的每个像素均乘以一个补偿系数,补偿系数计算公式可以表示为:
Coef=exp(TSL k/T ) k=1,2,...,N     (3)
其中,Coef表示补偿系数。
建立对欠采K空间数据进行重建的重建模型之后,需要对重建模型进行 求解,请参照图3,步骤S105可以包括预处理子步骤、第一更新子步骤和第一迭代子步骤,预处理子步骤为子步骤S1051,第一更新子步骤包括子步骤S1052~S1054,第一迭代子步骤包括子步骤S1055,下面进行详细描述:
子步骤S1051,对欠采K空间数据进行预处理,得到第一补偿系数及第一图像细节。
在本发明实施例中,对于欠采K空间数据,首先,取欠采K空间数据的全采K空间中心数据进行傅里叶逆变换,将全采K空间中心数据转换到图像域,得到初始图像;然后,利用预设的参数弛豫模型对初始图像进行拟合,得到第一参数图,用T 1表示;再将第一参数图代入预设的补偿系数计算公式,计算出第一补偿系数,用Coef 1表示;同时,依据初始图像,设置第一图像细节为空集,即
Figure PCTCN2019122002-appb-000001
在本发明实施例中,对欠采K空间数据预处理之后,需要对重建模型进行迭代求解,即,从欠采K空间数据中重建出参数加权图像,在重建参数加权图像的过程中用到了拟合的参数图,将参数图作为一个约束加入到参数加权图像重建中。换句话说,在重建的迭代过程中,每次迭代都会根据新重建出来的参数加权图像和参数弛豫模型,更新参数图,并将更新后的参数图用于下一次迭代中的信号补偿,如此反复迭代,直至算法收敛,停止重建,得到最终重建出来的参数加权图像,下面通过子步骤S1052~S1054进行详细描述。利用参数弛豫模型的先验信息来增加参数加权图像的数据冗余性,以此来提高重建精度。
子步骤S1052,对欠采K空间数据进行傅里叶逆变换得到目标磁共振图像,利用第一补偿系数对目标磁共振图像进行补偿,得到第一参考图像。
在本发明实施例中,设定迭代次数为i=1,2...I,在i=1次迭代时,目标磁 共振图像是对欠采K空间数据进行傅里叶逆变换得到的,故i=1次迭代时的目标磁共振图像有伪影;在i=k次迭代时,将第k-1次迭代重建出的目标参数加权图像作为目标磁共振图像,由于第k-1次迭代重建出的目标参数加权图像伪影减少,故i=k次迭代时的目标磁共振图像伪影减少。同时,在i次迭代中的参数图用T i表示,补偿系数用Coef i表示,图像细节用feat i表示。
在迭代过程中,需要根据补偿系数对目标磁共振图像进行补偿,即,
Figure PCTCN2019122002-appb-000002
其中,X i表示目标磁共振图像,
Figure PCTCN2019122002-appb-000003
表示补偿后的图像。在第i=1时,对欠采K空间数据进行傅里叶逆变换得到目标磁共振图像,利用第一补偿系数Coef 1对X 1进行补偿,得到第一参考图像
Figure PCTCN2019122002-appb-000004
子步骤S1053,利用第一补偿系数及第一图像细节,对所述第一参考图像进行重建,得到目标参数加权图像。
在本发明实施例中,第一参考图像包括低秩部分(low-rank component,L)和稀疏部分(sparse component,S),在参数加权图像重建过程中,需要对第一参考图像的低秩部分L做奇异值阈值操作、稀疏部分S做软阈值操作,得到迭代更新后的低秩部分L和稀疏部分S,再对更新后的低秩部分L和稀疏部分S求和,就能得到更新后的目标磁共振图像。
因此,对所述第一参考图像进行重建的过程可以包括第二更新子步骤及第二迭代子步骤,其中,
第二更新子步骤,包括:
第一步,将第一参考图像划分为低秩部分L和稀疏部分S;
第二步,获取辅助稀疏矩阵,并利用辅助稀疏矩阵对第一参考图像的低秩部分进行奇异值阈值操作,得到目标低秩矩阵。辅助稀疏矩阵用S j-1表示,设定迭代次数为j=1,2,...J,在j=1时设定S j-1=0,在第j次迭代中,对第一参 考图像的低秩部分L进行奇异值阈值操作,得到目标低秩矩阵L j的过程可以用下式表示:
Figure PCTCN2019122002-appb-000005
其中,
Figure PCTCN2019122002-appb-000006
表示第一参数图像,SVT(·)表示奇异值阈值操作算子,其定义为:
SVT l(M)=UΛ λ(Σ)V H     (5)
其中,M=UΣV H表示奇异值分解(SVD),U、V分别为左、右奇异值向量组成的矩阵,V H表示V的共轭转置,Σ是由M的奇异值组成的对角矩阵;Λ λ(Σ)表示保留Σ中最大的奇异值不变,其他全为0。在本发明实施例中,只取L的最大奇异值,使得做奇异值阈值操作后低秩部分L的秩为Rank(L)=1。
第三步,判断第一图像细节是否为空集,当第一图像细节不为空集时,利用目标低秩矩阵及第一图像细节对第一参考图像的稀疏部分进行软阈值操作,得到目标稀疏矩阵。在第j次迭代中,对第一参考图像的稀疏部分S进行软阈值操作,得到目标稀疏矩阵S j的过程可以用下式表示:
Figure PCTCN2019122002-appb-000007
其中,
Figure PCTCN2019122002-appb-000008
表示第一参数图像,ST(·)表示软阈值操作算子,其定义为:
Figure PCTCN2019122002-appb-000009
其中,p是图像矩阵的一个元素,v是阈值,其值与第一图像细节中的值线性相关。
第四步,依据目标低秩矩阵及目标稀疏矩阵,得到第二参考图像,这一过程可以用下式表示:
Figure PCTCN2019122002-appb-000010
其中,
Figure PCTCN2019122002-appb-000011
表示第二参数图像,L j表示目标低秩矩阵,S j表示目标稀疏矩阵,E *表示E的逆操作,即,式(8)就是对多通道线圈K空间数据做傅里叶逆变换后再进行线圈组合,得到第二参考图像。
第四步,利用第一补偿系数对第二参考图像进行信号补偿的逆过程,得到辅助参数加权图像,在第j次迭代中,利用第一补偿系数对第二参考图像进行重建的过程可以用下式表示:
Figure PCTCN2019122002-appb-000012
其中,C -1(·)表示将
Figure PCTCN2019122002-appb-000013
图像基于每个像素除以补偿系数Coef i,X i j表示辅助参数加权图像,
Figure PCTCN2019122002-appb-000014
表示第二参考图像。
第二迭代子步骤,包括:
利用目标稀疏矩阵替代辅助稀疏矩阵、第二参考图像替代第一参考图像并执行第二更新子步骤,直至达到迭代终止条件,将重建得到的辅助参数加权图像作为目标参数加权图像,迭代终止条件可以是迭代次数达到预设阈值、或者重建误差小于某个预设值,预设阈值就是设定的迭代次数J,其可以由用户根据实际情况灵活设置,在此不做限定。
下面换一种方式对子步骤S1053介绍的迭代过程进行描述:
在第i次迭代的目标参数加权图像重建过程中,对第一参考图像进行重建的过程可以包括:
设定迭代次数为j=1,2,...J,并设定j=1时S j-1=0,在第j次迭代中:
1)将第一参考图像
Figure PCTCN2019122002-appb-000015
划分为低秩部分L和稀疏部分S;
2)更新L j,用公式(4)即
Figure PCTCN2019122002-appb-000016
对第一参考图像的低秩部分L进行奇异值阈值操作;
3)更新S j,判断图像细节feat i是否为空集,当
Figure PCTCN2019122002-appb-000017
时,用公式(6) 即
Figure PCTCN2019122002-appb-000018
对第一参考图像的稀疏部分S进行软阈值操作;
4)利用2)和3)更新后的L j和S j和公式(8)即
Figure PCTCN2019122002-appb-000019
得到第二参考图像
Figure PCTCN2019122002-appb-000020
5)利用公式(9)即
Figure PCTCN2019122002-appb-000021
对4)中的第二参考图像
Figure PCTCN2019122002-appb-000022
进行信号补偿的逆过程,得到辅助参数加权图像X i j
6)当j=J或重建误差小于某个预设值时终止迭代,将最终得到的辅助参数加权图像X i J作为目标加权图像。
当i=1时,将重建出的目标参数加权图像作为目标磁共振图像,利用子步骤S1052得到第一参考图像
Figure PCTCN2019122002-appb-000023
之后,对
Figure PCTCN2019122002-appb-000024
进行重建,得到目标参数加权图像X 1的过程为:j=1时,
Figure PCTCN2019122002-appb-000025
由于
Figure PCTCN2019122002-appb-000026
则S 1不更新,
Figure PCTCN2019122002-appb-000027
Figure PCTCN2019122002-appb-000028
基于每个像素除以补偿系数Coef 1得到
Figure PCTCN2019122002-appb-000029
j=2时,
Figure PCTCN2019122002-appb-000030
由于
Figure PCTCN2019122002-appb-000031
则S 2不更新,
Figure PCTCN2019122002-appb-000032
Figure PCTCN2019122002-appb-000033
基于每个像素除以补偿系数Coef 1得到
Figure PCTCN2019122002-appb-000034
j=3时……,重复上述步骤,直到第j=J或重建误差小于某个预设值时,将最终得到的X 1 J作为目标加权图像。
子步骤S1054,依据目标参数加权图像,确定出第二补偿系数及第二图像细节。
在本发明实施例中,对于子步骤S1053得到的目标参数加权图像,首先,利用预设的参数弛豫模型对目标参数加权图像进行拟合,得到第二参数图;然后,将第二参数图代入预设的补偿系数计算公式,计算出第二补偿系数;再利用迭代细节提取算子(iterative feature refinement,IFR),对目标参数加权图像进行图像细节提取,得到第二图像细节。也就是说,在第i次迭代中,对于子步骤S1053得到的目标参数加权图像X 1 J,先利用公式(2)即 M=M 0exp(-TSL k/T 1p) k=1,2,...,N对X 1 J进行拟合,得到参数图
Figure PCTCN2019122002-appb-000035
再将参数图
Figure PCTCN2019122002-appb-000036
代入公式(3)即Coef=exp(TSL k/T ) k=1,2,...,N,计算出补偿系数Coef i;再利用迭代细节提取算子,对X 1 J进行图像细节提取,得到图像细节feat i
子步骤S1055,将重建出的目标参数加权图像作为目标磁共振图像,利用第二补偿系数、第二图像细节分别替代第一补偿系数及第一图像细节并执行第一更新子步骤,直至达到迭代终止条件,将最终重建出的目标参数加权图像作为参数加权图像。也就是,当i=I或重建误差小于某个预设值时终止迭代,将最终重建出来的目标参数加权图像X N J作为参数加权图像。
下面换一种方式对步骤S105介绍的参数加权图像的重建过程进行描述:
1、设定迭代次数为i=1,2...I,在第i次迭代中,根据补偿系数Coef i对目标磁共振图像进行补偿得到第一参考图像,即
Figure PCTCN2019122002-appb-000037
在i=1次迭代时,目标磁共振图像是对欠采K空间数据进行傅里叶逆变换得到的,i=1次迭代时的目标磁共振图像有伪影;在i=k次迭代时,将第k-1次迭代重建出的目标参数加权图像作为目标磁共振图像,i=k次迭代时的目标磁共振图像伪影减少;
2、对上一步得到的第一参考图像进行重建,这一过程可以包括:
设定迭代次数为j=1,2,...J,并设定j=1时S j-1=0,在第j次迭代中:
1)将第一参考图像
Figure PCTCN2019122002-appb-000038
划分为低秩部分L和稀疏部分S;2)更新L j,用公式(4)即
Figure PCTCN2019122002-appb-000039
对第一参考图像的低秩部分L进行奇异值阈值操作;
3)更新S j,判断图像细节feat i是否为空集,当
Figure PCTCN2019122002-appb-000040
时,用公式(6)即
Figure PCTCN2019122002-appb-000041
对第一参考图像的稀疏部分S进行软阈值操作;
4)利用2)和3)更新后的L j和S j和公式(8)即
Figure PCTCN2019122002-appb-000042
得到第二参考图像
Figure PCTCN2019122002-appb-000043
5)利用公式(9)即
Figure PCTCN2019122002-appb-000044
对4)中的第二参考图像
Figure PCTCN2019122002-appb-000045
进行信号补偿的逆过程,得到辅助参数加权图像X i j
6)当j=J或重建误差小于某个预设值时终止迭代,将最终得到的辅助参数加权图像X i J作为目标参数加权图像;
3、利用公式(2)即M=M 0exp(-TSL k/T 1p) k=1,2,...,N对目标参数加权图像X 1 J进行拟合,得到参数图
Figure PCTCN2019122002-appb-000046
再将参数图
Figure PCTCN2019122002-appb-000047
代入公式(3)即Coef=exp(TSL k/T ) k=1,2,...,N,计算出补偿系数Coef i;再利用迭代细节提取算子,对X 1 J进行图像细节提取,得到图像细节feat i
4、当i=I或重建误差小于某个预设值时终止迭代,将最终重建出来的目标参数加权图像X N J作为参数加权图像。
步骤S106,利用参数弛豫模型对重建出的参数加权图像进行拟合得到参数图,其中,参数图包括磁共振图像中各种组织的参数值。
在本发明实施例中,利用步骤S105介绍的方法重建出参数加权图像X N J之后,利用公式(2)即M=M 0exp(-TSL k/T 1p) k=1,2,...,N对参数加权图像X 1 J进行拟合,得到最终的参数图T
与现有技术相比,本发明实施例具有以下有益效果:
首先,采用了变速率变密度采集模块采集欠采K空间数据,加快了扫描速度;
其次,在参数加权图像重建过程中引入了信号补偿的方法,同时结合对欠采K空间数据进行重建的重建模型,可以从欠采K空间数据中精确重建出参数加权图像;同时,在参数加权图像的重建过程中用到了拟合的参数图,将参数图作为一个约束加入到重建中,提高了重建精度;
第三,现有基于模型的磁共振快速参数成像方法仅能使用一些较为简单的 参数弛豫模型,本发明实施例中参数弛豫模型的选择更为灵活和宽松,实用性更强;
第四,本发明实施例加入了迭代细节提取算子从残差图像中提取出图像细节,并根据图像细节的值对第一参考图像的稀疏部分做软阈值操作,可以更好的去除伪影。
第二实施例
请参照图4,图4示出了本发明实施例提供的磁共振快速参数成像装置200的方框示意图。磁共振快速参数成像装置200包括设置模块201、模板生成模块202、数据采集模块203、模型建立模块204、图像重建模块205及图像拟合模块206。
设置模块201,用于设置磁共振仪采集磁共振图像的多个参数时间点,并确定每个参数时间点对应的数据降采率。
模板生成模块202,用于根据每个参数时间点对应的数据降采率,以及频率编码方向全采、相位编码方向变密度采集的数据采集策略,生成变速率变密度采集模板。
数据采集模块203,用于按照变速率变密度采集模板,采集磁共振图像对应的欠采K空间数据。
模型建立模块204,用于建立对欠采K空间数据进行重建的重建模型。
图像重建模块205,用于对重建模型进行求解,在求解过程中基于预设的参数弛豫模型进行信号补偿,从所述欠采K空间数据中重建出参数加权图像。
在本发明实施例中,图像重建模块205执行对重建模型进行求解,在求解过程中基于预设的参数弛豫模型进行信号补偿,从所述欠采K空间数据中重建出参数加权图像的方式,包括预处理子步骤、第一更新子步骤及第一迭代子 步骤;其中,
预处理子步骤包括:对欠采K空间数据进行预处理,得到第一补偿系数及第一图像细节;
第一更新子步骤,包括:对所述欠采K空间数据进行傅里叶逆变换得到目标磁共振图像,利用第一补偿系数对目标磁共振图像进行补偿,得到第一参考图像;利用第一补偿系数及第一图像细节,对第一参考图像进行重建,得到目标参数加权图像;依据目标参数加权图像,确定出第二补偿系数及第二图像细节;
第一迭代子步骤,包括:将重建出的目标参数加权图像作为作为目标磁共振图像,利用第二补偿系数、第二图像细节分别替代第一补偿系数及第一图像细节并执行第一更新子步骤,直至达到迭代终止条件,将最终重建出的目标参数加权图像作为参数加权图像。
在本发明实施例中,图像重建模块205执行预处理子步骤的方式,包括:取欠采K空间数据的全采K空间中心数据进行傅里叶逆变换,得到初始图像;利用预设的参数弛豫模型对初始图像进行拟合,得到第一参数图;将第一参数图代入预设的补偿系数计算公式,计算出第一补偿系数;依据初始图像,设置第一图像细节为空集。
在本发明实施例中,图像重建模块205执行第一更新子步骤中利用第一补偿系数及第一图像细节,对第一参考图像进行重建,得到目标参数加权图像的方式,包括第二更新子步骤及第二迭代子步骤,其中,
第二更新子步骤,包括:将第一参考图像划分为低秩部分和稀疏部分;获取辅助稀疏矩阵,并利用辅助稀疏矩阵对第一参考图像的低秩部分进行奇异值阈值操作,得到目标低秩矩阵;当第一图像细节不为空集时,利用目标低秩矩 阵及第一图像细节对第一参考图像的稀疏部分进行软阈值操作,得到目标稀疏矩阵;依据目标低秩矩阵及目标稀疏矩阵,得到第二参考图像;利用第一补偿系数对第二参考图像进行信号补偿的逆过程,得到辅助参数加权图像;
第二迭代子步骤,包括:利用目标稀疏矩阵替代辅助稀疏矩阵、第二参考图像替代第一参考图像并执行第二更新子步骤,直至达到迭代终止条件,将重建得到的辅助参数加权图像作为目标参数加权图像。
在本发明实施例中,图像重建模块205执行第一更新子步骤中依据目标参数加权图像,确定出第二补偿系数及第二图像细节的方式,包括:利用预设的参数弛豫模型对目标参数加权图像进行拟合,得到第二参数图;将第二参数图代入预设的补偿系数计算公式,计算出第二补偿系数;利用迭代细节提取算子,对目标参数加权图像进行图像细节提取,得到第二图像细节。
图像拟合模块206,用于利用参数弛豫模型对重建出的参数加权图像进行拟合得到参数图,其中,参数图包括磁共振图像中各种组织的参数值。
综上所述,本发明实施例提供的一种磁共振快速参数成像方法及装置,所述方法包括:按照变速率变密度采集模板,采集磁共振图像对应的欠采K空间数据;建立对欠采K空间数据进行重建的重建模型;对重建模型进行求解,在求解过程中基于预设的参数弛豫模型进行信号补偿,从所述欠采K空间数据中重建出参数加权图像;利用参数弛豫模型对重建出的参数加权图像进行拟合得到参数图,其中,参数图包括磁共振图像中各种组织的参数值。与现有技术相比,本发明实施例在重建过程中引入了信号补偿,可以从欠采K空间数据中精确的重建出参数加权图像,进一步拟合出参数图,提高了参数图的准确度。
在本申请所提供的几个实施例中,应该理解到,所揭露的装置和方法,也 可以通过其它的方式实现。以上所描述的装置实施例仅仅是示意性的,例如,附图中的流程图和框图显示了根据本发明的多个实施例的装置、方法和计算机程序产品的可能实现的体系架构、功能和操作。在这点上,流程图或框图中的每个方框可以代表一个模块、程序段或代码的一部分,所述模块、程序段或代码的一部分包含一个或多个用于实现规定的逻辑功能的可执行指令。也应当注意,在有些作为替换的实现方式中,方框中所标注的功能也可以以不同于附图中所标注的顺序发生。例如,两个连续的方框实际上可以基本并行地执行,它们有时也可以按相反的顺序执行,这依所涉及的功能而定。也要注意的是,框图和/或流程图中的每个方框、以及框图和/或流程图中的方框的组合,可以用执行规定的功能或动作的专用的基于硬件的系统来实现,或者可以用专用硬件与计算机指令的组合来实现。
另外,在本发明各个实施例中的各功能模块可以集成在一起形成一个独立的部分,也可以是各个模块单独存在,也可以两个或两个以上模块集成形成一个独立的部分。
所述功能如果以软件功能模块的形式实现并作为独立的产品销售或使用时,可以存储在一个计算机可读取存储介质中。基于这样的理解,本发明的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的部分可以以软件产品的形式体现出来,该计算机软件产品存储在一个存储介质中,包括若干指令用以使得一台计算机设备(可以是个人计算机,服务器,或者网络设备等)执行本发明各个实施例所述方法的全部或部分步骤。而前述的存储介质包括:U盘、移动硬盘、只读存储器(ROM,Read-Only Memory)、随机存取存储器(RAM,Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。需要说明的是,在本文中,诸如第一和第二等之类的关系术语仅 仅用来将一个实体或者操作与另一个实体或操作区分开来,而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。而且,术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含,从而使得包括一系列要素的过程、方法、物品或者设备不仅包括那些要素,而且还包括没有明确列出的其他要素,或者是还包括为这种过程、方法、物品或者设备所固有的要素。在没有更多限制的情况下,由语句“包括一个……”限定的要素,并不排除在包括所述要素的过程、方法、物品或者设备中还存在另外的相同要素。
以上所述仅为本发明的优选实施例而已,并不用于限制本发明,对于本领域的技术人员来说,本发明可以有各种更改和变化。凡在本发明的精神和原则之内,所作的任何修改、等同替换、改进等,均应包含在本发明的保护范围之内。应注意到:相似的标号和字母在下面的附图中表示类似项,因此,一旦某一项在一个附图中被定义,则在随后的附图中不需要对其进行进一步定义和解释。

Claims (10)

  1. 一种磁共振快速参数成像方法,其特征在于,所述方法包括:
    按照变速率变密度采集模板,采集磁共振图像对应的欠采K空间数据;
    建立对所述欠采K空间数据进行重建的重建模型;
    对所述重建模型进行求解,在求解过程中基于预设的参数弛豫模型进行信号补偿,从所述欠采K空间数据中重建出参数加权图像;
    利用所述参数弛豫模型对重建出的参数加权图像进行拟合得到参数图,其中,参数图包括磁共振图像中各种组织的参数值。
  2. 如权利要求1所述的方法,其特征在于,所述按照预设的变速率变密度采集模板,采集磁共振图像对应的欠采K空间数据的步骤之前,所述方法还包括:
    设置磁共振仪采集磁共振图像的多个参数时间点,并确定每个参数时间点对应的数据降采率;
    根据每个参数时间点对应的数据降采率,以及频率编码方向全采、相位编码方向变密度采集的数据采集策略,生成变速率变密度采集模板。
  3. 如权利要求1所述的方法,其特征在于,所述欠采K空间数据包括多个按序排列的参数时间点对应的K空间数据;
    所述对重建模型进行求解,在求解过程中基于预设的参数弛豫模型进行信号补偿,从所述欠采K空间数据中重建出参数加权图像的步骤,包括预处理子步骤、第一更新子步骤及第一迭代子步骤;其中,
    所述预处理子步骤包括:
    对所述欠采K空间数据进行预处理,得到第一补偿系数及第一图像细节;
    所述第一更新子步骤,包括:
    对所述欠采K空间数据进行傅里叶逆变换得到目标磁共振图像,利用第一补偿系数对目标磁共振图像进行补偿,得到第一参考图像;
    利用第一补偿系数及第一图像细节,对所述第一参考图像进行重建,得到目标参数加权图像;依据所述目标参数加权图像,确定出第二补偿系数及第二图像细节;
    所述第一迭代子步骤,包括:
    将重建出的目标参数加权图像作为目标磁共振图像,利用第二补偿系数、第二图像细节分别替代第一补偿系数及第一图像细节并执行所述第一更新子步骤,直至达到迭代终止条件,将最终重建出的目标参数加权图像作为参数加权图像。
  4. 如权利要求3所述的方法,其特征在于,所述对欠采K空间数据进行预处理,得到第一补偿系数及第一图像细节的步骤,包括:
    取所述欠采K空间数据的全采K空间中心数据进行傅里叶逆变换,得到初始图像;
    利用预设的参数弛豫模型对所述初始图像进行拟合,得到第一参数图;
    将第一参数图代入预设的补偿系数计算公式,计算出第一补偿系数;
    依据所述初始图像,设置所述第一图像细节为空集。
  5. 如权利要求3所述的方法,其特征在于,所述利用第一补偿系数及第一图像细节,对所述第一参考图像进行重建,得到目标参数加权图像的步骤,包括第二更新子步骤及第二迭代子步骤,其中,
    所述第二更新子步骤,包括:
    将所述第一参考图像划分为低秩部分和稀疏部分;
    获取辅助稀疏矩阵,并利用辅助稀疏矩阵对所述第一参考图像的低秩部分进行奇异值阈值操作,得到目标低秩矩阵;
    当第一图像细节不为空集时,利用目标低秩矩阵及第一图像细节对所述第一参考图像的稀疏部分进行软阈值操作,得到目标稀疏矩阵;
    依据目标低秩矩阵及目标稀疏矩阵,得到第二参考图像;
    利用第一补偿系数对所述第二参考图像进行信号补偿的逆过程,得到辅助参数加权图像;
    所述第二迭代子步骤,包括:
    利用目标稀疏矩阵替代辅助稀疏矩阵、第二参考图像替代第一参考图像并执行所述第二更新子步骤,直至达到迭代终止条件,将重建得到的辅助参数加权图像作为目标参数加权图像。
  6. 如权利要求3所述的方法,其特征在于,依据所述目标参数加权图像,确定出第二补偿系数及第二图像细节的步骤,包括:
    利用预设的参数弛豫模型对所述目标参数加权图像进行拟合,得到第二参数图;
    将第二参数图代入预设的补偿系数计算公式,计算出第二补偿系数;
    利用迭代细节提取算子,对目标参数加权图像进行图像细节提取,得到第二图像细节。
  7. 如权利要求6所述的方法,其特征在于,所述参数弛豫模型为M=M 0exp(-TSL k/T ) k=1,2,...,N,M表示每个参数时间点对应的磁共振图像的图像强度,M 0为平衡态的图像强度,TSL k表示第k个参数时间点,N表示参数时间点的个数,T 表示参数图;
    所述补偿系数计算公式为Coef=exp(TSL k/T ) k=1,2,...,N,Coef表示补偿系数。
  8. 如权利要求1所述的方法,其特征在于,所述重建模型为min {X,L,S}||S|| 1s.t.C(X)=L+S,E(X)=d,Rank(L)=1,||·|| 1表示l 1范数,C(·)是一个操作算子,表示对磁共振图像进行像素级的信号补偿,X表示要重建的磁共振图像,L表示磁共振图像的低秩矩阵,S表示磁共振图像的稀疏矩阵,E表示多通道线圈编码矩阵,Rank(L)表示低秩矩阵L的秩。
  9. 一种磁共振快速参数成像装置,其特征在于,所述装置包括:
    数据采集模块,用于按照变速率变密度采集模板,采集磁共振图像对应的欠采K空间数据;
    模型建立模块,用于建立对所述欠采K空间数据进行重建的重建模型;
    图像重建模块,用于对所述重建模型进行求解,在求解过程中基于预设的参数弛豫模型进行信号补偿,从所述欠采K空间数据中重建出参数加权图像;
    图像拟合模块,用于利用所述参数弛豫模型对重建出的参数加权图像进行拟合得到参数图,其中,参数图包括磁共振图像中各种组织的参数值。
  10. 如权利要求9所述的装置,其特征在于,所述装置还包括:
    设置模块,用于设置磁共振仪采集磁共振图像的多个参数时间点,并确定每个参数时间点对应的数据降采率;
    模板生成模块,用于根据每个参数时间点对应的数据降采率,以及频率编码方向全采、相位编码方向变密度采集的数据采集策略,生成变速率变密度采集模板。
PCT/CN2019/122002 2018-12-03 2019-11-29 磁共振快速参数成像方法及装置 WO2020114329A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811467405.0 2018-12-03
CN201811467405.0A CN109633502B (zh) 2018-12-03 2018-12-03 磁共振快速参数成像方法及装置

Publications (1)

Publication Number Publication Date
WO2020114329A1 true WO2020114329A1 (zh) 2020-06-11

Family

ID=66070634

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/122002 WO2020114329A1 (zh) 2018-12-03 2019-11-29 磁共振快速参数成像方法及装置

Country Status (2)

Country Link
CN (1) CN109633502B (zh)
WO (1) WO2020114329A1 (zh)

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109633502B (zh) * 2018-12-03 2020-07-24 深圳先进技术研究院 磁共振快速参数成像方法及装置
CN111856365B (zh) * 2019-04-24 2023-03-14 深圳先进技术研究院 磁共振成像方法、装置、系统及存储介质
CN111856364B (zh) * 2019-04-24 2023-03-28 深圳先进技术研究院 一种磁共振成像方法、装置、系统及存储介质
CN113050009B (zh) * 2019-12-27 2022-12-13 深圳先进技术研究院 三维磁共振快速参数成像方法和装置
CN113567901A (zh) * 2020-04-29 2021-10-29 深圳先进技术研究院 一种磁共振旋转坐标系下的自旋晶格弛豫成像方法和系统
CN112336337B (zh) * 2020-11-06 2022-09-02 深圳先进技术研究院 磁共振参数成像模型的训练方法及训练装置、介质和设备
CN112381786A (zh) * 2020-11-12 2021-02-19 深圳先进技术研究院 定量参数值获取方法和装置、可读存储介质、计算机设备
CN112710975A (zh) * 2021-01-25 2021-04-27 东北林业大学 基于稀疏和局部低秩矩阵分解的磁共振扩散图像重建方法
CN113093076B (zh) * 2021-03-22 2022-05-10 上海东软医疗科技有限公司 磁共振图像的处理方法、装置及电子设备

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101975936A (zh) * 2010-09-03 2011-02-16 杭州电子科技大学 一种基于cs压缩感知技术的快速磁共振成像方法
JP2011030673A (ja) * 2009-07-30 2011-02-17 Hitachi Medical Corp 磁気共鳴イメージング装置、および、画像再構成プログラム
CN102309328A (zh) * 2011-10-19 2012-01-11 中国科学院深圳先进技术研究院 弥散张量成像方法及系统
CN103064046A (zh) * 2012-12-25 2013-04-24 深圳先进技术研究院 一种基于稀疏采样的核磁共振成像的图像处理方法
CN103472419A (zh) * 2013-08-30 2013-12-25 深圳先进技术研究院 磁共振快速成像方法及其系统
CN103646410A (zh) * 2013-11-27 2014-03-19 中国科学院深圳先进技术研究院 磁共振快速参数成像方法和系统
CN103654789A (zh) * 2013-12-10 2014-03-26 深圳先进技术研究院 磁共振快速参数成像方法和系统
CN103901376A (zh) * 2012-12-30 2014-07-02 上海联影医疗科技有限公司 磁共振成像方法与装置
CN109633502A (zh) * 2018-12-03 2019-04-16 深圳先进技术研究院 磁共振快速参数成像方法及装置

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108324276B (zh) * 2018-01-11 2021-07-30 上海东软医疗科技有限公司 磁共振成像方法和装置

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2011030673A (ja) * 2009-07-30 2011-02-17 Hitachi Medical Corp 磁気共鳴イメージング装置、および、画像再構成プログラム
CN101975936A (zh) * 2010-09-03 2011-02-16 杭州电子科技大学 一种基于cs压缩感知技术的快速磁共振成像方法
CN102309328A (zh) * 2011-10-19 2012-01-11 中国科学院深圳先进技术研究院 弥散张量成像方法及系统
CN103064046A (zh) * 2012-12-25 2013-04-24 深圳先进技术研究院 一种基于稀疏采样的核磁共振成像的图像处理方法
CN103901376A (zh) * 2012-12-30 2014-07-02 上海联影医疗科技有限公司 磁共振成像方法与装置
CN103472419A (zh) * 2013-08-30 2013-12-25 深圳先进技术研究院 磁共振快速成像方法及其系统
CN103646410A (zh) * 2013-11-27 2014-03-19 中国科学院深圳先进技术研究院 磁共振快速参数成像方法和系统
CN103654789A (zh) * 2013-12-10 2014-03-26 深圳先进技术研究院 磁共振快速参数成像方法和系统
CN109633502A (zh) * 2018-12-03 2019-04-16 深圳先进技术研究院 磁共振快速参数成像方法及装置

Also Published As

Publication number Publication date
CN109633502A (zh) 2019-04-16
CN109633502B (zh) 2020-07-24

Similar Documents

Publication Publication Date Title
WO2020114329A1 (zh) 磁共振快速参数成像方法及装置
Tezcan et al. MR image reconstruction using deep density priors
US10671939B2 (en) System, method and computer-accessible medium for learning an optimized variational network for medical image reconstruction
Ongie et al. A fast algorithm for convolutional structured low-rank matrix recovery
Ongie et al. Off-the-grid recovery of piecewise constant images from few Fourier samples
Song et al. Coupled dictionary learning for multi-contrast MRI reconstruction
US10627470B2 (en) System and method for learning based magnetic resonance fingerprinting
CN109658468B (zh) 磁共振参数成像方法、装置、设备及存储介质
US10180476B2 (en) Systems and methods for segmented magnetic resonance fingerprinting dictionary matching
Montesinos et al. Application of the compressed sensing technique to self‐gated cardiac cine sequences in small animals
WO2019153566A1 (zh) 磁共振参数成像方法、装置、医学设备及存储介质
CN113924503B (zh) 时域磁共振的参数图确定
Kelkar et al. Prior image-constrained reconstruction using style-based generative models
WO2022193379A1 (zh) 图像重建模型生成及装置、图像重建方法及装置、设备、介质
CN117011673B (zh) 基于噪声扩散学习的电阻抗层析成像图像重建方法和装置
Balachandrasekaran et al. Calibration-free B0 correction of EPI data using structured low rank matrix recovery
Ryu et al. K-space refinement in deep learning mr reconstruction via regularizing scan specific spirit-based self consistency
WO2012061475A2 (en) Systems and methods for fast magnetic resonance image reconstruction
WO2021129235A1 (zh) 三维磁共振快速参数成像方法和装置
WO2023050249A1 (zh) 基于深度学习的磁共振成像方法、系统、终端及存储介质
JP2023069890A (ja) 磁気共鳴イメージング装置、画像処理装置、及び、画像処理方法
Hao et al. A fast linearized alternating minimization algorithm for constrained high-order total variation regularized compressive sensing
WO2023029087A1 (zh) 一种低场磁共振快速成像方法、终端设备及计算机存储介质
KR102514804B1 (ko) 비지도 학습 기반 자기 공명 영상 처리 방법 및 그 장치
CN117333571B (zh) 磁共振图像的重建方法、系统、设备及介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19892290

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 05.11.2021)

122 Ep: pct application non-entry in european phase

Ref document number: 19892290

Country of ref document: EP

Kind code of ref document: A1