WO2013086951A1 - Dynamic contrast enhanced magnetic resonance imaging method and system - Google Patents

Dynamic contrast enhanced magnetic resonance imaging method and system Download PDF

Info

Publication number
WO2013086951A1
WO2013086951A1 PCT/CN2012/086197 CN2012086197W WO2013086951A1 WO 2013086951 A1 WO2013086951 A1 WO 2013086951A1 CN 2012086197 W CN2012086197 W CN 2012086197W WO 2013086951 A1 WO2013086951 A1 WO 2013086951A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
magnetic resonance
support
reconstructed image
resonance imaging
Prior art date
Application number
PCT/CN2012/086197
Other languages
French (fr)
Chinese (zh)
Inventor
梁栋
张娜
刘新
郑海荣
Original Assignee
中国科学院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 中国科学院深圳先进技术研究院 filed Critical 中国科学院深圳先进技术研究院
Publication of WO2013086951A1 publication Critical patent/WO2013086951A1/en

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R33/00—Arrangements or instruments for measuring magnetic variables
    • G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
    • G01R33/44—Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
    • G01R33/48—NMR imaging systems
    • G01R33/54—Signal processing systems, e.g. using pulse sequences ; Generation or control of pulse sequences; Operator console
    • G01R33/56—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution
    • G01R33/561—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution by reduction of the scanning time, i.e. fast acquiring systems, e.g. using echo-planar pulse sequences
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R33/00—Arrangements or instruments for measuring magnetic variables
    • G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
    • G01R33/44—Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
    • G01R33/48—NMR imaging systems
    • G01R33/54—Signal processing systems, e.g. using pulse sequences ; Generation or control of pulse sequences; Operator console
    • G01R33/56—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution
    • G01R33/563—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution of moving material, e.g. flow contrast angiography
    • G01R33/56308—Characterization of motion or flow; Dynamic imaging

Definitions

  • the present invention relates to magnetic resonance imaging techniques, and more particularly to a dynamic contrast enhanced magnetic resonance imaging method and system. Background technique
  • MRI Magnetic Resonance Imaging
  • CT computed tomography
  • It is a rapid development with the development of computer technology, electronic circuit technology and superconductor technology.
  • the spine motion characteristics of a particular nucleus in human tissue and the biomagnetic nuclear spin imaging technique of nuclear magnetic resonance phenomena are characterized by non-invasive, non-ionizing radiation and high tissue contrast.
  • MRI can obtain the morphological information and functional information of the examination site at the same time. It has the unparalleled advantages of other technologies and has become an important means of medical imaging examination today.
  • the DCE-MRI Dynamic Contrast Enhanced MRI method based on rapid imaging technology is a functional MRI technique based on the injection of a paramagnetic contrast agent into a blood vessel to shorten the longitudinal relaxation time T1 of the tissue.
  • the imaging records the change in tissue signal intensity to track the diffusion of contrast agent into the surrounding tissue over time.
  • Dynamic imaging is performed at intervals of approximately 5-10 minutes after the first passage of the contrast agent. It is a functional MRI method for quantitatively studying microvascular endothelial permeability. .
  • DCE-MRI requires high spatial resolution and time resolution to ensure the accuracy of quantitative measurements.
  • time resolution and spatial resolution are in a certain relationship under certain conditions, and it is difficult to achieve high temporal and spatial resolution imaging at the same time.
  • a dynamic contrast-enhanced magnetic resonance imaging method comprising: scanning and obtaining ⁇ spatial data; performing nonlinear image reconstruction on the ⁇ spatial data to obtain a reconstructed image; and performing a branching coefficient on the reconstructed image or the reconstructed image Set detection; the reconstructed image and the support detect repeated iterations to convergence; generate a magnetic resonance image.
  • the step of performing support detection on the refinement image or the sparse coefficient of the reconstructed image is: a preset threshold; acquiring an image reconstruction value or a sparse coefficient of the reconstructed image according to the reconstructed image And if the image reconstruction value or the sparse coefficient value is greater than the preset threshold, the support information is acquired.
  • the step of repeating the iterative to convergence of the image reconstruction and the support detection is: in the image reconstruction process, the solution is truncated by the support information, and the calculation formula is :
  • the position of the non-zero element outside the known support is the image sequence in the spatial and temporal frequency domains, d is the transform domain signal, and F is the two-dimensional Fourier transform of the spatial frequency domain and the time domain direction, which is the noise level;
  • the calculation formula of the truncation minimization problem is converted into the weight minimization problem by the support information, and the reconstructed image is solved by the FOCUSS algorithm, and the calculation formula is:
  • w is a diagonal weighting matrix
  • the calculation formula for obtaining the support information is
  • the threshold is calculated as: r (0 is the threshold, ) > 0 is the exponential function of the number of external iterations i.
  • a dynamic contrast-enhanced magnetic resonance imaging system comprising: an acquisition module for scanning and obtaining K-space data; and an image reconstruction module for performing nonlinear image reconstruction on the K-space data to obtain reconstruction
  • An image detection module configured to perform support detection on a sparse coefficient of the reconstructed image or the reconstructed image; a repeating iterative module, configured to repeatedly iterate to the reconstructed image and the support to converge;
  • a generation module for generating a magnetic resonance image.
  • the support detection module includes: a preset threshold unit, configured to preset a threshold; and an acquisition support information unit, configured to acquire an image reconstruction value or a sparse image of the reconstructed image according to the reconstructed image The coefficient value, the image reconstruction value or the sparse coefficient value is greater than the preset threshold, and the support information is acquired.
  • the repeated iteration module includes: an image reconstruction calculation unit, which is used to solve the minimization problem by using the support information, and the calculation formula is:
  • the position of the non-zero element outside the known support is the image sequence in the spatial and temporal frequency domains
  • d is the transform domain signal
  • F is the two-dimensional Fourier transform of the spatial frequency domain and the time domain direction
  • s is the noise level
  • the calculation formula for solving the truncation minimization problem by the support information is converted into the weight minimization problem, and the reconstructed image is solved by the FOCUSS algorithm, and the calculation formula is:
  • w is a diagonal weighting matrix
  • the support detection unit calculates the calculation formula of the support information as
  • the threshold is calculated as: r (0 is the threshold, ) > 0 is the exponential function of the number of external iterations i.
  • the dynamic contrast-enhanced magnetic resonance imaging method and system described above repeats iteration to convergence by image reconstruction and support detection, that is, obtains multiple reconstructed images and supports, and updates the support to the image reconstruction in the next iteration.
  • image reconstruction and support detection that is, obtains multiple reconstructed images and supports, and updates the support to the image reconstruction in the next iteration.
  • the more support information is detected the less the measurement data is applied to reconstruct the accurate image, which shortens the scanning time and improves the image quality.
  • FIG. 1 is a flow chart of a dynamic contrast enhanced magnetic resonance imaging method in an embodiment
  • FIG. 2 is a flow chart of a method for detecting a sparse coefficient of a reconstructed image or a reconstructed image in FIG.
  • Figure 3 is a magnetic resonance image obtained by repeating iteration to convergence and fitting reconstruction based on image reconstruction and support detection;
  • Figure 4 is a magnetic resonance image obtained only by the conventional FOCUSS algorithm
  • Figure 5 is a magnetic resonance image obtained by the OMP method
  • FIG. 6 is a schematic structural view of a dynamic contrast enhanced magnetic resonance imaging system in an embodiment
  • FIG. 7 is a schematic structural view of the support detecting module in FIG. 7;
  • FIG. 8 is a schematic structural diagram of a repeating iterative module in FIG. 7.
  • a dynamic contrast enhanced magnetic resonance imaging method includes the following steps:
  • Step S10 Scan and obtain K-space data.
  • the K-space data is received and obtained by the scanning signal transmitted by the tissue (e.g., breast) of the radio frequency coil of the magnetic resonance apparatus.
  • Step S20 performing nonlinear image reconstruction on the K-space data to obtain a reconstructed image.
  • the signal released by the tissue to be inspected is received, K-space data is obtained, and nonlinear image reconstruction is performed according to the K-space data to obtain a reconstructed image.
  • Step S30 performing branch detection on the sparse coefficients of the reconstructed image or the reconstructed image.
  • the dynamic MRI image is sparse in the spatial domain and the time-frequency domain (xf) domain, and is a priori information of dynamic imaging.
  • the initial data is reconstructed using Compress Sensing (CS) technology.
  • the data of the spatial domain and the time-frequency domain are detected by the support, or the sparse coefficients of the reconstructed image are detected.
  • the support set is the position of the non-zero element in the xf domain, and the partial known support information (Partial Known Support, PKS) of the dynamic MRI image is a priori information, so the support information is obtained through the support detection. .
  • PKS Partial Known Support
  • step S30 is:
  • Step S310 preset a threshold.
  • the selection of the preset threshold if the smaller threshold will cause many errors in the detection of the support, it will not be corrected in the subsequent image reconstruction iteration; if the larger threshold will cause few position values for the detection support , need ⁇ multiple iterations.
  • the threshold can be set according to the following formula:
  • Step S320 Acquire an image reconstruction value or a sparse coefficient value of the reconstructed image according to the reconstructed image. If the image reconstruction value or the sparse coefficient value is greater than a preset threshold, the support information is acquired.
  • selecting the image reconstruction value or the location where the sparse coefficient value is greater than the preset threshold value learns and acquires the support information of the x-f domain, that is, the process of the support detection.
  • Step S40 Image reconstruction and support detection repeat iteration to convergence.
  • step S20 and step S30 are iteratively repeated to convergence.
  • w is a diagonal weighting matrix, which is an image sequence in the spatial and temporal domains
  • d is a transform domain signal
  • F is a two-dimensional Fourier transform in the spatial frequency domain and the time domain (kt) direction
  • s is the noise level, i For the number of iterations.
  • the calculation formula for obtaining the support information is:
  • Step S50 Generate a magnetic resonance image.
  • the image reconstruction and the support detection are subjected to a plurality of iterations to converge and a reconstructed image is obtained, which is a magnetic resonance image.
  • the above dynamic contrast-enhanced magnetic resonance imaging method uses dynamic MRI images to be sparse in the xf domain and partially known support information as a priori information, and repeats iteration to convergence by image reconstruction and support detection, thereby improving spatial resolution and The time resolution, in turn, improves the dynamic contrast and enhances the quality of magnetic resonance imaging.
  • FIGS. 3 to 5 are a magnetic resonance image obtained by repeating iteration to convergence and fitting reconstruction based on image reconstruction and support detection, that is, a magnetic resonance image obtained by using the scheme;
  • FIG. 4 is a magnetic resonance image obtained by using only the conventional FOCUSS algorithm.
  • Figure 5 shows the magnetic resonance image obtained using the OMP (Two-Step Orthogonal Matching Tracking) method.
  • OMP Two-Step Orthogonal Matching Tracking
  • the acquisition module 10 is configured to scan and obtain K-space data.
  • the K-space data is received and obtained by the radio frequency coil of the magnetic resonance apparatus, the scanning signal emitted by the tissue to be examined (for example, the breast).
  • the image reconstruction module 20 is configured to perform nonlinear image reconstruction on the K-space data to obtain a reconstructed image.
  • the signal released by the organization to be inspected is received, K-space data is obtained, and the data is spatially transformed to obtain the data of the spatial domain and the time-frequency domain (x-f).
  • the support detection module 30 is configured to perform branch detection on the reconstructed image or the sparse coefficient of the reconstructed image.
  • the dynamic MRI image is sparse in the spatial domain and the time-frequency domain (x-f) domain, and is a priori information of dynamic imaging.
  • the initial data is first reconstructed using Compress Sensing (CS) technology.
  • CS Compress Sensing
  • the support is the position of the non-zero element in the xf domain, and the part of the dynamic MRI image in the xf domain.
  • Known support information (Partial Known Support, PKS) is a priori information, so the support information is obtained through the support detection.
  • the support detection module 30 further includes:
  • the preset threshold unit 310 is configured to preset a threshold.
  • the position value of the preset threshold is selected. If the smaller threshold will cause many errors in the detection support, it will not be corrected in the subsequent image reconstruction iteration; if the larger threshold will cause the detection support Very few position values require many iterations.
  • ') is the threshold, ')>0 is the exponential function of the external iteration number i, which is determined by the exponential law: ae— bz b ⁇ 0, ⁇ 0.
  • the acquisition support information unit 320 is configured to obtain an image reconstruction value or a sparse coefficient value of the reconstructed image according to the reconstructed image. If the image reconstruction value or the sparse coefficient value is greater than a preset threshold, the support information is acquired.
  • selecting the image reconstruction value or the location where the sparse coefficient value is greater than the preset threshold value learns and acquires the support information of the x-f domain, that is, the process of the support detection.
  • the iterative module 40 is used to iterate to convergence for image reconstruction and branch detection.
  • the iterative to convergence is repeated for image reconstruction and support detection, that is, the process of support detection.
  • the iterative repeating module 40 includes:
  • the image reconstruction calculation unit 410 in the image reconstruction process, solves the minimization problem by the support information, and obtains the intermediate reconstruction ⁇ , and the calculation formula is:
  • the position of the non-zero element outside the known support is the image sequence in the spatial and temporal frequency domains
  • d is the transform domain signal
  • F is the two-dimensional Fourier transform of the spatial frequency domain and the time domain (kt) direction
  • ⁇ For the noise level, i is the number of iterations.
  • Cut through the branch Set Information Solutions formula minimization problem i.e., the above equation 1) is converted to a weight minimization problem, and by diversity measure of FOCUSS (aggregation underdetermined system resolution algorithm Focal Underdetermined System Solver, cylindrical called diversity measure of FOCUSS) algorithm for the reconstruction (0,
  • the calculation formula is:
  • W is a diagonal weighting matrix, which is a sequence of images in the spatial and temporal domains, and d is a transform domain signal.
  • F is a two-dimensional Fourier transform in the spatial frequency domain and the time domain (k-t) direction, where s is the noise level and i is the number of iterations.
  • the support detection calculation unit 420 in the support detection process, obtains the calculation formula of the support information as:
  • the image generation module 50 is configured to generate a magnetic resonance image.
  • the image reconstruction and the support detection are subjected to a plurality of iterations to converge and a reconstructed image is obtained, which is a magnetic resonance image.
  • the above dynamic contrast-enhanced magnetic resonance imaging system uses dynamic MRI images to be sparse in the xf domain and partially known support information as a priori information, and repeats iteration to convergence for image reconstruction and support detection, thereby improving spatial resolution and The time resolution, in turn, achieves the goal of improving the quality of dynamic contrast-enhanced magnetic resonance imaging. It is not to be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the spirit and scope of the invention. Therefore, the scope of the invention is to be determined by the appended claims.

Landscapes

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

Abstract

A dynamic contrast enhanced magnetic resonance imaging method and system. The method comprises: scanning and obtaining Κ space data (S10); performing non-linear image reconstruction on the K space data, and obtaining a reconstructed image (S20); performing subset detection on the reconstructed image or a sparsity coefficient of the reconstructed image (S30); performing repetitive iteration on the image reconstruction and the subset detection to converge (S40); and generating a magnetic resonance image (S50). The imaging method and system perform repetitive iteration on image reconstruction and subset detection to converge, so as to obtain a plurality of reconstructed images and subsets, and the subsets are updated into image reconstruction performed by next iteration. When more subset information is detected, the measurement data contained in a signal used by a reconstructed accurate image becomes less, thereby shortening the scanning time, so as to improve image quality.

Description

说明书 发明名称： 动态对比度增强磁共振成像方法和*** 技术领域  Description Title: Dynamic Contrast Enhanced Magnetic Resonance Imaging Method and System
本发明涉及磁共振成像技术， 特别是涉及一种动态对比度增强磁共振成 像方法和***。 背景技术  The present invention relates to magnetic resonance imaging techniques, and more particularly to a dynamic contrast enhanced magnetic resonance imaging method and system. Background technique
磁共振成像 MRI ( Magnetic Resonance Imaging )是继计算机断层扫描 ( CT )之后医学影像诊断技术的又一重大进展， 是随着计算机技术、 电子电 路技术、 超导体技术的发展而迅速发展起来的一种利用人体组织中某种特定 原子核的自旋运动特点与核磁共振现象的生物磁学核自旋成像技术， 具有无 创伤、 无电离辐射及较高的组织对比度的特点。 MRI作为目前被广泛采用的 一种诊断技术， 可同时获得***位的形态信息和功能信息， 具有其它技术 所无可比拟的优势， 成为当今医学影像检查的重要手段。  MRI (Magnetic Resonance Imaging) is another major advancement in medical imaging diagnostics following computed tomography (CT). It is a rapid development with the development of computer technology, electronic circuit technology and superconductor technology. The spine motion characteristics of a particular nucleus in human tissue and the biomagnetic nuclear spin imaging technique of nuclear magnetic resonance phenomena are characterized by non-invasive, non-ionizing radiation and high tissue contrast. As a diagnostic technology widely used at present, MRI can obtain the morphological information and functional information of the examination site at the same time. It has the unparalleled advantages of other technologies and has become an important means of medical imaging examination today.
近年来， 基于快速成像技术发展起来的动态对比度增强磁共振成像 ( DCE-MRI Dynamic Contrast Enhanced MRI )方法是基于顺磁性造影剂注 入血管导致组织纵向弛豫时间 T1缩短的功能 MRI技术， 它使用重复成像记 录组织信号强度的变化以跟踪造影剂随时间扩散到周围组织中的情况， 在造 影剂首次通过后大约 5-10分钟的间隔完成动态成像， 是定量研究微血管内皮 通透性的功能 MRI方法。  In recent years, the DCE-MRI Dynamic Contrast Enhanced MRI method based on rapid imaging technology is a functional MRI technique based on the injection of a paramagnetic contrast agent into a blood vessel to shorten the longitudinal relaxation time T1 of the tissue. The imaging records the change in tissue signal intensity to track the diffusion of contrast agent into the surrounding tissue over time. Dynamic imaging is performed at intervals of approximately 5-10 minutes after the first passage of the contrast agent. It is a functional MRI method for quantitatively studying microvascular endothelial permeability. .
DCE-MRI作为一种定量评估方法， 需要较高的空间分辨率和时间分辨 率， 以保证定量测量的准确性。 然而在传统的傅里叶 MRI中， 时间分辨率与 空间分辨率在一定条件下是此消彼长的关系，难以同时实现高时空分辨成像。 发明内容  As a quantitative evaluation method, DCE-MRI requires high spatial resolution and time resolution to ensure the accuracy of quantitative measurements. However, in traditional Fourier MRI, time resolution and spatial resolution are in a certain relationship under certain conditions, and it is difficult to achieve high temporal and spatial resolution imaging at the same time. Summary of the invention
基于此， 有必要针对磁共振图像质量不高的问题， 提供一种磁共振图像 质量较高的动态对比度增强磁共振成像方法。 此外， 还有必要提供一种磁共振图像质量较高的动态对比度增强磁共振 成像***。 Based on this, it is necessary to provide a dynamic contrast-enhanced magnetic resonance imaging method with high magnetic resonance image quality for the problem of low quality of magnetic resonance images. In addition, it is also necessary to provide a dynamic contrast-enhanced magnetic resonance imaging system with high magnetic resonance image quality.
一种动态对比度增强磁共振成像方法， 包括： 扫描并得到 Κ空间数据； 对所述 Κ空间数据进行非线性图像重建， 得到重建图像； 对所述重建图像或 者所述重建图像的稀疏系数进行支集检测； 所述重建图像和所述支集检测重 复迭代至收敛； 生成磁共振图像。  A dynamic contrast-enhanced magnetic resonance imaging method, comprising: scanning and obtaining Κ spatial data; performing nonlinear image reconstruction on the Κ spatial data to obtain a reconstructed image; and performing a branching coefficient on the reconstructed image or the reconstructed image Set detection; the reconstructed image and the support detect repeated iterations to convergence; generate a magnetic resonance image.
在其中一个实施例中， 对所述重建图像或者所述重建图像的稀疏系数进 行支集检测的步骤过程为： 预设阈值； 根据所述重建图像获取图像重建值或 所述重建图像的稀疏系数值， 所述图像重建值或所述稀疏系数值大于所述预 设阈值， 则获取支集信息。  In one embodiment, the step of performing support detection on the refinement image or the sparse coefficient of the reconstructed image is: a preset threshold; acquiring an image reconstruction value or a sparse coefficient of the reconstructed image according to the reconstructed image And if the image reconstruction value or the sparse coefficient value is greater than the preset threshold, the support information is acquired.
在其中一个实施例中， 所述图像重建和所述支集检测重复迭代至收敛的 步骤过程为： 在所述图像重建过程中， 通过所述支集信息解截断 1 最小化问 题， 计算公式为： In one embodiment, the step of repeating the iterative to convergence of the image reconstruction and the support detection is: in the image reconstruction process, the solution is truncated by the support information, and the calculation formula is :
Figure imgf000004_0001
Figure imgf000004_0001
为已知支集外的非零元素位置， 为空域和时频域中的图像序列， d 为变换域信号， F为空间频域和时域方向的二维傅里叶变换， 为噪音水平； 通过所述支集信息解截断 最小化问题的计算公式转换为权重 最小化 问题， 并通过 FOCUSS算法求解重建图像， 计算公式为： The position of the non-zero element outside the known support is the image sequence in the spatial and temporal frequency domains, d is the transform domain signal, and F is the two-dimensional Fourier transform of the spatial frequency domain and the time domain direction, which is the noise level; The calculation formula of the truncation minimization problem is converted into the weight minimization problem by the support information, and the reconstructed image is solved by the FOCUSS algorithm, and the calculation formula is:
Figure imgf000004_0002
Figure imgf000004_0002
w为对角加权矩阵；  w is a diagonal weighting matrix;
在所述支集检测过程中， 获得支集信息的计算公式为  In the support detection process, the calculation formula for obtaining the support information is
f > τ(ί) ] 为 p、、的第 z个元素， 为所述空域和时频域的重建图像数据值 < 在其中一个实施例中， 所述阈值的计算公式为：
Figure imgf000004_0003
r(0为阈值， )>0增序列为外部迭代次数 i的指数函数。 f > τ (ί) ] is the zth element of p, , is the reconstructed image data value of the spatial domain and the time-frequency domain < In one embodiment, the threshold is calculated as:
Figure imgf000004_0003
r (0 is the threshold, ) > 0 is the exponential function of the number of external iterations i.
另外， 还有必要提供一种动态对比度增强磁共振成像***， 包括： 采集 模块， 用于扫描并得到 K空间数据； 图像重建模块， 用于对所述 K空间数据 进行非线性图像重建， 得到重建图像； 支集检测模块， 用于对所述重建图像 或者所述重建图像的稀疏系数进行支集检测； 重复迭代模块， 用于对所述重 建图像和所述支集检测重复迭代至收敛； 图像生成模块， 用于生成磁共振图 像。  In addition, it is also necessary to provide a dynamic contrast-enhanced magnetic resonance imaging system, comprising: an acquisition module for scanning and obtaining K-space data; and an image reconstruction module for performing nonlinear image reconstruction on the K-space data to obtain reconstruction An image detection module, configured to perform support detection on a sparse coefficient of the reconstructed image or the reconstructed image; a repeating iterative module, configured to repeatedly iterate to the reconstructed image and the support to converge; A generation module for generating a magnetic resonance image.
在其中一个实施例中， 所述支集检测模块包括： 预设阈值单元， 用于预 设阈值； 获取支集信息单元， 用于根据所述重建图像获取图像重建值或所述 重建图像的稀疏系数值， 所述图像重建值或所述稀疏系数值大于所述预设阈 值， 则获取支集信息。  In one embodiment, the support detection module includes: a preset threshold unit, configured to preset a threshold; and an acquisition support information unit, configured to acquire an image reconstruction value or a sparse image of the reconstructed image according to the reconstructed image The coefficient value, the image reconstruction value or the sparse coefficient value is greater than the preset threshold, and the support information is acquired.
在其中一个实施例中， 所述重复迭代模块包括： 图像重建计算单元， 通 过所述支集信息解截断 最小化问题， 计算公式为： In one embodiment, the repeated iteration module includes: an image reconstruction calculation unit, which is used to solve the minimization problem by using the support information, and the calculation formula is:
Figure imgf000005_0001
Figure imgf000005_0001
为已知支集外的非零元素位置， 为空域和时频域中的图像序列， d 为变换域信号， F为空间频域和时域方向的二维傅里叶变换， s为噪音水平； 通过所述支集信息解截断 最小化问题的计算公式转换为权重 最小化 问题， 并通过 FOCUSS算法求解重建图像， 计算公式为： The position of the non-zero element outside the known support is the image sequence in the spatial and temporal frequency domains, d is the transform domain signal, F is the two-dimensional Fourier transform of the spatial frequency domain and the time domain direction, and s is the noise level The calculation formula for solving the truncation minimization problem by the support information is converted into the weight minimization problem, and the reconstructed image is solved by the FOCUSS algorithm, and the calculation formula is:
Figure imgf000005_0002
Figure imgf000005_0002
w为对角加权矩阵；  w is a diagonal weighting matrix;
支集检测计算单元， 获得支集信息的计算公式为 The support detection unit calculates the calculation formula of the support information as
Figure imgf000005_0003
Figure imgf000005_0003
0为 ^的第 z个元素， w为所述空域和时频域的重建图像数据值 < 在其中一个实施例中所述阈值的计算公式为：
Figure imgf000005_0004
r(0为阈值， )>0增序列为外部迭代次数 i的指数函数。 0 is the zth element of ^, w is the reconstructed image data value of the spatial domain and the time-frequency domain < In one of the embodiments, the threshold is calculated as:
Figure imgf000005_0004
r (0 is the threshold, ) > 0 is the exponential function of the number of external iterations i.
上述动态对比度增强磁共振成像方法和***， 通过对图像重建和支集检 测重复迭代至收敛， 即得到多个重建的图像和支集， 并将支集更新到下一次 迭代进行的图像重建中， 所探测得到的支集信息越多， 重建出精确图像所应 用的信号包含的测量数据也就越少， 从而缩短了扫描时间， 进而达到了提高 了图像质量的目的。  The dynamic contrast-enhanced magnetic resonance imaging method and system described above repeats iteration to convergence by image reconstruction and support detection, that is, obtains multiple reconstructed images and supports, and updates the support to the image reconstruction in the next iteration. The more support information is detected, the less the measurement data is applied to reconstruct the accurate image, which shortens the scanning time and improves the image quality.
附图说明 图 1为一个实施例中动态对比度增强磁共振成像方法的流程图； 图 2为图 1中对重建图像或者重建图像的稀疏系数进行支集检测的方法 流程图； BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a flow chart of a dynamic contrast enhanced magnetic resonance imaging method in an embodiment; FIG. 2 is a flow chart of a method for detecting a sparse coefficient of a reconstructed image or a reconstructed image in FIG.
图 3为基于图像重建和支集检测重复迭代至收敛并进过拟合重建得到的 磁共振图像；  Figure 3 is a magnetic resonance image obtained by repeating iteration to convergence and fitting reconstruction based on image reconstruction and support detection;
图 4为仅用传统的 FOCUSS算法得到的磁共振图像；  Figure 4 is a magnetic resonance image obtained only by the conventional FOCUSS algorithm;
图 5为采用 OMP方法所得到磁共振图像；  Figure 5 is a magnetic resonance image obtained by the OMP method;
图 6为一个实施例中动态对比度增强磁共振成像***的结构示意图； 图 7为图 7中支集检测模块的结构示意图；  6 is a schematic structural view of a dynamic contrast enhanced magnetic resonance imaging system in an embodiment; FIG. 7 is a schematic structural view of the support detecting module in FIG. 7;
图 8为图 7中重复迭代模块的结构示意图。  FIG. 8 is a schematic structural diagram of a repeating iterative module in FIG. 7.
具体实施方式 如图 1所示， 在一个实施例中， 动态对比度增强磁共振成像方法， 包括 如下步骤： DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS As shown in FIG. 1, in one embodiment, a dynamic contrast enhanced magnetic resonance imaging method includes the following steps:
步骤 S10: 扫描并得到 K空间数据。 通过磁共振设备的射频线圈对待检 组织 （例如乳腺）所发射的扫描信号， 接收并得到 K空间数据。  Step S10: Scan and obtain K-space data. The K-space data is received and obtained by the scanning signal transmitted by the tissue (e.g., breast) of the radio frequency coil of the magnetic resonance apparatus.
步骤 S20: 对 K空间数据进行非线性图像重建， 得到重建图像。  Step S20: performing nonlinear image reconstruction on the K-space data to obtain a reconstructed image.
本实施例中， 在图像的动态扫描过程中接收到待检组织释放的信号， 得 到 K空间数据， 并根据 K空间数据进行非线性图像重建， 得到重建图像。  In this embodiment, during the dynamic scanning process of the image, the signal released by the tissue to be inspected is received, K-space data is obtained, and nonlinear image reconstruction is performed according to the K-space data to obtain a reconstructed image.
步骤 S30: 对重建图像或者重建图像的稀疏系数进行支集检测。 本实施例中， 动态 MRI图像在空域和时频域（x-f )域是稀疏， 是动态成 像的先验信息。 在 x-f域中首先采用压缩感知（ Compress Sensing, 筒称 CS ) 技术对初始数据进行图像重建。 Step S30: performing branch detection on the sparse coefficients of the reconstructed image or the reconstructed image. In this embodiment, the dynamic MRI image is sparse in the spatial domain and the time-frequency domain (xf) domain, and is a priori information of dynamic imaging. In the xf domain, the initial data is reconstructed using Compress Sensing (CS) technology.
在首次对空域和时频域的数据进行图像重建后， 对空域和时频域的数据 进行支集检测，或者对重建图像的的稀疏系数进行支集检测。支集是 x-f域中 非零元素的位置， 而动态 MRI图像在 x-f域中的部分已知支集信息 （Partial Known Support, 筒称 PKS )为先验信息， 故通过支集检测获得支集信息。  After image reconstruction of the data in the spatial domain and the time-frequency domain for the first time, the data of the spatial domain and the time-frequency domain are detected by the support, or the sparse coefficients of the reconstructed image are detected. The support set is the position of the non-zero element in the xf domain, and the partial known support information (Partial Known Support, PKS) of the dynamic MRI image is a priori information, so the support information is obtained through the support detection. .
进一步地， 结合附图 2, 步骤 S30的过程为：  Further, with reference to Figure 2, the process of step S30 is:
步骤 S310: 预设阈值。  Step S310: preset a threshold.
预设阈值的选取， 若较小的阈值会对检测支集造成很多错误的位置， 在 之后的图像重建迭代中得不到校正； 若较大的阈值会对检测支集造成很少的 位置值， 需要^艮多次的迭代。  The selection of the preset threshold, if the smaller threshold will cause many errors in the detection of the support, it will not be corrected in the subsequent image reconstruction iteration; if the larger threshold will cause few position values for the detection support , need ^ multiple iterations.
故， 阈值的设定可以根据如下计算公式：  Therefore, the threshold can be set according to the following formula:
、('■) 11  , ('■) 11
7(0 0为阈值， )>0增序列为外部迭代次数 i的指数函数，该指数函数由指 数定律确定： xz = ae-bz; b≥ 0, "≥ 0。 7 (0 0 is the threshold, ) > 0 is the exponential function of the number of external iterations i, which is determined by the exponential law: x z = ae- bz ; b ≥ 0, "≥ 0.
步骤 S320:根据重建图像获取得到图像重建值或重建图像的稀疏系数值， 图像重建值或稀疏系数值大于预设阈值， 则获取支集信息。  Step S320: Acquire an image reconstruction value or a sparse coefficient value of the reconstructed image according to the reconstructed image. If the image reconstruction value or the sparse coefficient value is greater than a preset threshold, the support information is acquired.
在本实施例中， 选取图像重建值或稀疏系数值大于预设阈值的位置学习 并获取 x-f域的支集信息， 即为支集检测的过程。  In this embodiment, selecting the image reconstruction value or the location where the sparse coefficient value is greater than the preset threshold value learns and acquires the support information of the x-f domain, that is, the process of the support detection.
步骤 S40: 图像重建和支集检测重复迭代至收敛。  Step S40: Image reconstruction and support detection repeat iteration to convergence.
在本实施例中， 对图像重建和支集检测的步骤（即步骤 S20和步骤 S30 ) 重复迭代至收敛。  In the present embodiment, the steps of image reconstruction and branch detection (i.e., step S20 and step S30) are iteratively repeated to convergence.
在图像重建过程中， 通过支集信息解截断 最小化问题， 进而获得中间 P(i 计算公式为： In the image reconstruction process, the minimization problem is solved by the support information to obtain the intermediate P (i is calculated as:
s.t. ¥ - Fp\\2 为已知支集外的非零元素位置， 为空域和时频域中的图像序列， d 为变换域信号， F为空间频域和时域（k-t ) 方向的二维傅里叶变换， ε为噪 音水平， i为迭代次数。 通过支集信息解截断 最小化问题的计算公式（即上述公式 1 )转换为 权重 最小化问题， 并通过 FOCUSS (聚集欠定***解决算法 Focal Underdetermined System Solver, 筒称 FOCUSS )算法求解重建 (0 , 计算公式 为：St ¥ - Fp\\ 2 The position of the non-zero element outside the known support is the image sequence in the spatial and temporal frequency domains, d is the transform domain signal, and F is the two-dimensional Fourier transform of the spatial frequency domain and the time domain (kt) direction, ε For the noise level, i is the number of iterations. Cut through the branch Set Information Solutions formula minimization problem (i.e., the above equation 1) is converted to a weight minimization problem, and by diversity measure of FOCUSS (aggregation underdetermined system resolution algorithm Focal Underdetermined System Solver, cylindrical called diversity measure of FOCUSS) algorithm for the reconstruction (0, The calculation formula is:
Figure imgf000008_0001
w为对角加权矩阵， 为空域和时频域中的图像序列， d为变换域信号， F为空间频域和时域（k-t )方向的二维傅里叶变换， s为噪音水平， i为迭代 次数。
Figure imgf000008_0001
w is a diagonal weighting matrix, which is an image sequence in the spatial and temporal domains, d is a transform domain signal, F is a two-dimensional Fourier transform in the spatial frequency domain and the time domain (kt) direction, s is the noise level, i For the number of iterations.
在支集检测过程中， 获得支集信息的计算公式为： In the support detection process, the calculation formula for obtaining the support information is:
Figure imgf000008_0002
Figure imgf000008_0002
0为 ^的第 z个元素， w为空域和时频域（x-f ) 的重建图像数据值。 步骤 S50: 生成磁共振图像。 0 is the zth element of ^, w is the reconstructed image data value of the spatial domain and the time-frequency domain (xf). Step S50: Generate a magnetic resonance image.
在本实施例中， 图像重建和支集检测经过了多次迭代至收敛并得到重建 图像， 该重建图像即为磁共振图像。 上述动态对比度增强磁共振成像方法， 利用动态 MRI图像在 x-f域是稀 疏的以及部分已知支集信息作为先验信息， 通过对图像重建和支集检测重复 迭代至收敛， 提高了空间分辨率和时间分辨率， 进而达到了提高了动态对比 度， 增强磁共振成像的质量的目的。  In the present embodiment, the image reconstruction and the support detection are subjected to a plurality of iterations to converge and a reconstructed image is obtained, which is a magnetic resonance image. The above dynamic contrast-enhanced magnetic resonance imaging method uses dynamic MRI images to be sparse in the xf domain and partially known support information as a priori information, and repeats iteration to convergence by image reconstruction and support detection, thereby improving spatial resolution and The time resolution, in turn, improves the dynamic contrast and enhances the quality of magnetic resonance imaging.
结合附图 3~5, 通过实验效果图进一步的阐述本方案的效果。 图 3为基 于图像重建和支集检测重复迭代至收敛并进过拟合重建得到的磁共振图像， 即采用本方案所得到的磁共振图像； 图 4为仅用传统的 FOCUSS算法得到的 磁共振图像； 图 5为采用 OMP (两步正交匹配追踪）方法所得到磁共振图像。 通过对比可知（重点关注图像中箭头所指区域）， 图 4所采用的 FOCUSS 方法得到的磁共振图像往往会在相位编码方向造成更多的欠采样伪影； 图 5 所采用的 OMP 方法得到的磁共振图像会造成更多的噪声， 导致图像质量较 差。 而采用本方案所获得的图像， 即图 3就克服了传统方案的缺陷， 提供了 较好的磁共振图像。 基于上述所提供的方法， 结合附图 6, 还有必要提供一种动态对比度增 强磁共振成像***， 包括： The effects of the solution are further illustrated by the experimental renderings in conjunction with FIGS. 3 to 5. 3 is a magnetic resonance image obtained by repeating iteration to convergence and fitting reconstruction based on image reconstruction and support detection, that is, a magnetic resonance image obtained by using the scheme; FIG. 4 is a magnetic resonance image obtained by using only the conventional FOCUSS algorithm. Figure 5 shows the magnetic resonance image obtained using the OMP (Two-Step Orthogonal Matching Tracking) method. By contrast (focusing on the area indicated by the arrow in the image), the magnetic resonance image obtained by the FOCUSS method used in Figure 4 tends to cause more undersampling artifacts in the phase encoding direction; Figure 5 is obtained by the OMP method. Magnetic resonance images cause more noise, resulting in poorer image quality. The image obtained by the scheme, that is, Fig. 3 overcomes the defects of the conventional scheme and provides a better magnetic resonance image. Based on the method provided above, in conjunction with FIG. 6, it is also necessary to provide a dynamic contrast enhanced magnetic resonance imaging system, including:
采集模块 10, 用于扫描并得到 K空间数据。 通过磁共振设备的射频线圈 对待检组织 （例如乳腺）所发射的扫描信号， 接收并得到 K空间数据。  The acquisition module 10 is configured to scan and obtain K-space data. The K-space data is received and obtained by the radio frequency coil of the magnetic resonance apparatus, the scanning signal emitted by the tissue to be examined (for example, the breast).
图像重建模块 20, 用于对 K空间数据进行非线性图像重建， 得到重建图 像。  The image reconstruction module 20 is configured to perform nonlinear image reconstruction on the K-space data to obtain a reconstructed image.
本实施例中， 在图像的动态扫描过程中接收到待检组织释放的信号， 得 到 K空间数据， 并对其进行稀疏变换得到空域和时频域（x-f ) 的数据。  In this embodiment, during the dynamic scanning process of the image, the signal released by the organization to be inspected is received, K-space data is obtained, and the data is spatially transformed to obtain the data of the spatial domain and the time-frequency domain (x-f).
支集检测模块 30, 用于对所述重建图像或者所述重建图像的稀疏系数进 行支集检测。  The support detection module 30 is configured to perform branch detection on the reconstructed image or the sparse coefficient of the reconstructed image.
本实施例中， 动态 MRI图像在空域和时频域（x-f )域是稀疏， 是动态成 像的先验信息。 在 x-f域中首先采用压缩感知（ Compress Sensing, 筒称 CS ) 技术对初始数据进行图像重建。  In this embodiment, the dynamic MRI image is sparse in the spatial domain and the time-frequency domain (x-f) domain, and is a priori information of dynamic imaging. In the x-f domain, the initial data is first reconstructed using Compress Sensing (CS) technology.
在首次对空域和时频域的数据进行图像重建后， 对空域和时频域的数据 进行支集检测， 支集是 x-f域中非零元素的位置， 而动态 MRI图像在 x-f域 中的部分已知支集信息（ Partial Known Support , 筒称 PKS )为先验信息， 故 通过支集检测获得支集信息。  After the image reconstruction of the spatial and temporal frequency data is performed for the first time, the data of the spatial domain and the time-frequency domain are detected, and the support is the position of the non-zero element in the xf domain, and the part of the dynamic MRI image in the xf domain. Known support information (Partial Known Support, PKS) is a priori information, so the support information is obtained through the support detection.
进一步地， 结合附图 7, 支集检测模块 30还包括：  Further, in conjunction with FIG. 7, the support detection module 30 further includes:
预设阈值单元 310 , 用于预设阈值。  The preset threshold unit 310 is configured to preset a threshold.
预设阈值的位置值选取， 若较小的阈值会对检测支集造成很多错误的位 置， 在之后的图像重建迭代中得不到校正； 若较大的阈值会对检测支集造成 很少的位置值， 需要很多次的迭代。 The position value of the preset threshold is selected. If the smaller threshold will cause many errors in the detection support, it will not be corrected in the subsequent image reconstruction iteration; if the larger threshold will cause the detection support Very few position values require many iterations.
故， 设定可以根据如下计算公式
Figure imgf000010_0001
Therefore, the setting can be calculated according to the following formula
Figure imgf000010_0001
')为阈值， ')>0增序列为外部迭代次数 i的指数函数，该指数函数由指 数定律确定: ae—bz b≥0, ≥0。 ') is the threshold, ')>0 is the exponential function of the external iteration number i, which is determined by the exponential law: ae— bz b≥0, ≥0.
获取支集信息单元 320, 用于根据重建图像获取得到图像重建值或重建 图像的稀疏系数值， 图像重建值或稀疏系数值大于预设阈值， 则获取支集信 息。  The acquisition support information unit 320 is configured to obtain an image reconstruction value or a sparse coefficient value of the reconstructed image according to the reconstructed image. If the image reconstruction value or the sparse coefficient value is greater than a preset threshold, the support information is acquired.
在本实施例中， 选取图像重建值或稀疏系数值大于预设阈值的位置学习 并获取 x-f域的支集信息， 即为支集检测的过程。  In this embodiment, selecting the image reconstruction value or the location where the sparse coefficient value is greater than the preset threshold value learns and acquires the support information of the x-f domain, that is, the process of the support detection.
重复迭代模块 40, 用于对图像重建和支集检测重复迭代至收敛。  The iterative module 40 is used to iterate to convergence for image reconstruction and branch detection.
在本实施例中， 对图像重建和支集检测重复迭代至收敛， 即为支集检测 的过程。  In this embodiment, the iterative to convergence is repeated for image reconstruction and support detection, that is, the process of support detection.
进一步地， 结合附图 8, 重复迭代模块 40包括：  Further, in conjunction with FIG. 8, the iterative repeating module 40 includes:
图像重建计算单元 410, 在图像重建过程中， 通过支集信息解截断 1 最 小化问题， 进而获得中间重建 ^， 计算公式为： The image reconstruction calculation unit 410, in the image reconstruction process, solves the minimization problem by the support information, and obtains the intermediate reconstruction ^, and the calculation formula is:
Figure imgf000010_0002
Figure imgf000010_0002
为已知支集外的非零元素位置， 为空域和时频域中的图像序列， d 为变换域信号， F为空间频域和时域（k-t ) 方向的二维傅里叶变换， ε为噪 音水平， i为迭代次数。 通过支集信息解截断 最小化问题的计算公式（即上述公式 1 )转换为 权重 最小化问题， 并通过 FOCUSS (聚集欠定***解决算法 Focal Underdetermined System Solver, 筒称 FOCUSS )算法求解重建 (0 , 计算公式 为： The position of the non-zero element outside the known support is the image sequence in the spatial and temporal frequency domains, d is the transform domain signal, and F is the two-dimensional Fourier transform of the spatial frequency domain and the time domain (kt) direction, ε For the noise level, i is the number of iterations. Cut through the branch Set Information Solutions formula minimization problem (i.e., the above equation 1) is converted to a weight minimization problem, and by diversity measure of FOCUSS (aggregation underdetermined system resolution algorithm Focal Underdetermined System Solver, cylindrical called diversity measure of FOCUSS) algorithm for the reconstruction (0, The calculation formula is:
s.t.\\d - Fp\\ < s (2) W为对角加权矩阵， 为空域和时频域中的图像序列， d为变换域信号，St\\d - Fp\\ < s (2) W is a diagonal weighting matrix, which is a sequence of images in the spatial and temporal domains, and d is a transform domain signal.
F为空间频域和时域（k-t )方向的二维傅里叶变换， s为噪音水平， i为迭代 次数。 F is a two-dimensional Fourier transform in the spatial frequency domain and the time domain (k-t) direction, where s is the noise level and i is the number of iterations.
支集检测计算单元 420, 在支集检测过程中， 获得支集信息的计算公式 为： The support detection calculation unit 420, in the support detection process, obtains the calculation formula of the support information as:
Figure imgf000011_0001
Figure imgf000011_0001
0为 ^的第 z个元素， w为空域和时频域（x-f ) 的重建图像数据值。 图像生成模块 50, 用于生成磁共振图像。 0 is the zth element of ^, w is the reconstructed image data value of the spatial domain and the time-frequency domain (xf). The image generation module 50 is configured to generate a magnetic resonance image.
在本实施例中， 图像重建和支集检测经过了多次迭代至收敛并得到重建 图像， 该重建图像即为磁共振图像。  In the present embodiment, the image reconstruction and the support detection are subjected to a plurality of iterations to converge and a reconstructed image is obtained, which is a magnetic resonance image.
上述动态对比度增强磁共振成像***， 利用动态 MRI图像在 x-f域是稀 疏的以及部分已知支集信息作为先验信息， 通过对图像重建和支集检测重复 迭代至收敛， 提高了空间分辨率和时间分辨率， 进而达到了提高了动态对比 度增强磁共振成像的质量的目的。 细， 但并不能因此而理解为对本发明专利范围的限制。 应当指出的是， 对于 本领域的普通技术人员来说， 在不脱离本发明构思的前提下， 还可以做出若 干变形和改进， 这些都属于本发明的保护范围。 因此， 本发明专利的保护范 围应以所附权利要求为准。  The above dynamic contrast-enhanced magnetic resonance imaging system uses dynamic MRI images to be sparse in the xf domain and partially known support information as a priori information, and repeats iteration to convergence for image reconstruction and support detection, thereby improving spatial resolution and The time resolution, in turn, achieves the goal of improving the quality of dynamic contrast-enhanced magnetic resonance imaging. It is not to be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the spirit and scope of the invention. Therefore, the scope of the invention is to be determined by the appended claims.

Claims

权利要求书 Claim
1、 一种动态对比度增强磁共振成像方法， 包括： 1. A dynamic contrast enhanced magnetic resonance imaging method, comprising:
扫描并得到 K空间数据；  Scan and get K space data;
对所述 K空间数据进行非线性图像重建， 得到重建图像；  Performing nonlinear image reconstruction on the K-space data to obtain a reconstructed image;
对所述重建图像或者所述重建图像的稀疏系数进行支集检测；  Performing support detection on the reconstructed image or the sparse coefficient of the reconstructed image;
所述重建图像和所述支集检测重复迭代至收敛；  The reconstructed image and the support detect repeated iterations to convergence;
生成磁共振图像。  A magnetic resonance image is generated.
2、根据权利要求 1所述的动态对比度增强磁共振成像方法，其特征在于， 对所述重建图像或者所述重建图像的稀疏系数进行支集检测的步骤过程为： 预设阈值；  The dynamic contrast-enhanced magnetic resonance imaging method according to claim 1, wherein the step of performing support detection on the refinement image or the sparse coefficient of the reconstructed image is: a preset threshold;
根据所述重建图像获取图像重建值或所述重建图像的稀疏系数值， 所述 图像重建值或所述稀疏系数值大于所述预设阈值， 则获取支集信息。  And acquiring, according to the reconstructed image, an image reconstruction value or a sparse coefficient value of the reconstructed image, where the image reconstruction value or the sparse coefficient value is greater than the preset threshold, acquiring the support information.
3、根据权利要求 2所述的动态对比度增强磁共振成像方法，其特征在于， 所述图像重建和所述支集检测重复迭代至收敛的步骤过程为：  3. The dynamic contrast enhanced magnetic resonance imaging method according to claim 2, wherein the step of repeating the iterative to convergence of the image reconstruction and the support detection is:
在所述图像重建过程中， 通过所述支集信息解截断 最小化问题， 计算 公式为： In the image reconstruction process, the minimization problem is solved by the support information, and the calculation formula is:
Figure imgf000012_0001
为已知支集外的非零元素位置， 为空域和时频域中的图像序列， d 为变换域信号， F为空间频域和时域方向的二维傅里叶变换， s为噪音水平； 通过所述支集信息解截断 最小化问题的计算公式转换为权重 最小化 问题， 并通过 FOCUSS算法求解重建图像， 计算公式为：
Figure imgf000012_0001
The position of the non-zero element outside the known support is the image sequence in the spatial and temporal frequency domains, d is the transform domain signal, F is the two-dimensional Fourier transform of the spatial frequency domain and the time domain direction, and s is the noise level The calculation formula for solving the truncation minimization problem by the support information is converted into the weight minimization problem, and the reconstructed image is solved by the FOCUSS algorithm, and the calculation formula is:
Figure imgf000012_0002
w为对角加权矩阵；
Figure imgf000012_0002
w is a diagonal weighting matrix;
在所述支集检测过程中， 获得支集信息的计算公式为  In the support detection process, the calculation formula for obtaining the support information is
> r(;) 0为 ^的第 z个元素， w为所述空域和时频域的重建图像数据值。 > r (;) 0 is the zth element of ^, and w is the reconstructed image data value of the spatial domain and the time-frequency domain.
4、根据权利要求 3所述的动态对比度增强磁共振成像方法，其特征在于, 所述阈值的计 公式为：
Figure imgf000013_0001
4. The dynamic contrast enhanced magnetic resonance imaging method according to claim 3, wherein the formula of the threshold is:
Figure imgf000013_0001
r(0为阈值， ') >0增序列为外部迭代次数 i的指数函数。 r (0 is the threshold, ') > 0 is the exponential function of the number of external iterations i.
5、 一种动态对比度增强磁共振成像***， 其特征在于， 包括： 5. A dynamic contrast enhanced magnetic resonance imaging system, comprising:
采集模块， 用于扫描并得到 K空间数据；  An acquisition module for scanning and obtaining K-space data;
图像重建模块， 用于对所述 K空间数据进行非线性图像重建， 得到重建 图像；  An image reconstruction module, configured to perform nonlinear image reconstruction on the K-space data to obtain a reconstructed image;
支集检测模块， 用于对所述重建图像或者所述重建图像的稀疏系数进行 支集检测；  a support detection module, configured to perform support detection on a sparse coefficient of the reconstructed image or the reconstructed image;
重复迭代模块， 用于对所述重建图像和所述支集检测重复迭代至收敛； 图像生成模块， 用于生成磁共振图像。  And repeating an iterative module, configured to repeat the iteration to the reconstructed image and the support to convergence; and an image generation module, configured to generate a magnetic resonance image.
6、根据权利要求 5所述的动态对比度增强磁共振成像***，其特征在于， 所述支集检测模块包括：  The dynamic contrast-enhanced magnetic resonance imaging system according to claim 5, wherein the support detection module comprises:
预设阈值单元， 用于预设阈值；  a preset threshold unit for preset thresholds;
获取支集信息单元， 用于根据所述重建图像获取图像重建值或所述重建 图像的稀疏系数值， 所述图像重建值或所述稀疏系数值大于所述预设阈值， 则获取支集信息。  Obtaining a support information unit, configured to acquire an image reconstruction value or a sparse coefficient value of the reconstructed image according to the reconstructed image, where the image reconstruction value or the sparse coefficient value is greater than the preset threshold, acquiring support information .
7、根据权利要求 6所述的动态对比度增强磁共振成像***，其特征在于， 所述重复迭代模块包括：  The dynamic contrast-enhanced magnetic resonance imaging system according to claim 6, wherein the repeated iteration module comprises:
图像重建计算单元， 通过所述支集信息解截断 最小化问题， 计算公式 min Δ(,— " s.t.\d - Fp\ 为已知支集外的非零元素位置， 为空域和时频域中的图像序列， d 为变换域信号， F为空间频域和时域方向的二维傅里叶变换， s为噪音水平； 通过所述支集信息解截断 最小化问题的计算公式转换为权重 最小化 问题， 并通过 FOCUSS算法求解重建图像， 计算公式为： An image reconstruction calculation unit is configured to solve the minimization problem by using the support information, and calculate a formula min Δ (, - " st\d - Fp\ The position of the non-zero element outside the known support is the image sequence in the spatial and temporal frequency domains, d is the transform domain signal, F is the two-dimensional Fourier transform of the spatial frequency domain and the time domain direction, and s is the noise level The calculation formula for solving the truncation minimization problem by the support information is converted into the weight minimization problem, and the reconstructed image is solved by the FOCUSS algorithm, and the calculation formula is:
w为对角加权矩阵； w is a diagonal weighting matrix;
支集检测计算单元， 获得支集信息的计算公式为 The support detection unit calculates the calculation formula of the support information as
Figure imgf000014_0001
Figure imgf000014_0001
0为 ^的第 Ζ个元素， w为所述空域和时频域的重建图像数据值。 0 is the third element of ^, and w is the reconstructed image data value of the spatial domain and the time-frequency domain.
8、根据权利要求 7所述的动态对比度增强磁共振成像***，其特征在于 所述阈 公式为：
Figure imgf000014_0002
8. A dynamic contrast enhanced magnetic resonance imaging system according to claim 7 wherein said threshold formula is:
Figure imgf000014_0002
为阈值， ')>0增序列为外部迭代次数 i的指数函数。  For the threshold, the sequence of ')>0 is an exponential function of the number of external iterations i.
PCT/CN2012/086197 2011-12-12 2012-12-07 Dynamic contrast enhanced magnetic resonance imaging method and system WO2013086951A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201110411985.3 2011-12-12
CN201110411985 2011-12-12

Publications (1)

Publication Number Publication Date
WO2013086951A1 true WO2013086951A1 (en) 2013-06-20

Family

ID=48015465

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2012/086197 WO2013086951A1 (en) 2011-12-12 2012-12-07 Dynamic contrast enhanced magnetic resonance imaging method and system

Country Status (2)

Country Link
CN (1) CN103027682A (en)
WO (1) WO2013086951A1 (en)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103349550A (en) * 2013-07-04 2013-10-16 华东师范大学 Method and device for integrating magnetic resonance imaging scanning and compressive sensing reconstruction
CN103654789B (en) * 2013-12-10 2015-12-30 深圳先进技术研究院 Fast magnetic resonance parametric formation method and system
DE102016207314A1 (en) * 2016-04-28 2017-11-02 Siemens Healthcare Gmbh Method for displaying quantitative magnetic resonance image data
CN108802648B (en) * 2018-04-03 2020-12-01 上海东软医疗科技有限公司 Magnetic resonance quantitative imaging method and device based on gradient echo
CN111090068B (en) * 2018-10-23 2021-03-23 中国科学院深圳先进技术研究院 Magnetic resonance positive contrast imaging method and device
CN110391016B (en) * 2019-07-10 2021-11-09 浙江大学 Analysis method for dynamic contrast enhanced magnetic resonance image
CN111784793B (en) * 2020-07-01 2022-05-03 厦门大学 Dynamic magnetic resonance imaging reconstruction method

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1653348A (en) * 2002-05-13 2005-08-10 皇家飞利浦电子股份有限公司 Prior-information-enhanced dynamic magnetic resonance imaging
US20080175452A1 (en) * 2007-01-19 2008-07-24 Korea Advanced Institute Of Science And Technology Method for super-resolution reconstruction using focal underdetermined system solver algorithm
US7592808B1 (en) * 2008-05-06 2009-09-22 General Electric Company System and method for reducing MR scan time using partial fourier acquisition and compressed sensing
CN101975936A (en) * 2010-09-03 2011-02-16 杭州电子科技大学 Rapid magnetic resonance imaging (MRI) method based on CS ( compressed sensing ) technique
CN102085097A (en) * 2010-12-16 2011-06-08 中国科学院深圳先进技术研究院 Magnetic resonance dynamic imaging method

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7688068B2 (en) * 2008-05-06 2010-03-30 General Electric Company System and method for using parallel imaging with compressed sensing
US20100246920A1 (en) * 2009-03-31 2010-09-30 Iowa State University Research Foundation, Inc. Recursive sparse reconstruction
US8897515B2 (en) * 2009-09-08 2014-11-25 Mayo Foundation For Medical Education And Research Method for compressed sensing image reconstruction using a priori knowledge of spatial support
US8760572B2 (en) * 2009-11-19 2014-06-24 Siemens Aktiengesellschaft Method for exploiting structure in sparse domain for magnetic resonance image reconstruction
CN101975935B (en) * 2010-09-03 2012-12-05 杭州电子科技大学 Partial echo compressed sensing-based quick magnetic resonance imaging method
CN103035017A (en) * 2011-12-08 2013-04-10 中国科学院深圳先进技术研究院 Method of magnetic resonance parameter imaging and system of magnetic resonance parameter imaging

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1653348A (en) * 2002-05-13 2005-08-10 皇家飞利浦电子股份有限公司 Prior-information-enhanced dynamic magnetic resonance imaging
US20080175452A1 (en) * 2007-01-19 2008-07-24 Korea Advanced Institute Of Science And Technology Method for super-resolution reconstruction using focal underdetermined system solver algorithm
US7592808B1 (en) * 2008-05-06 2009-09-22 General Electric Company System and method for reducing MR scan time using partial fourier acquisition and compressed sensing
CN101975936A (en) * 2010-09-03 2011-02-16 杭州电子科技大学 Rapid magnetic resonance imaging (MRI) method based on CS ( compressed sensing ) technique
CN102085097A (en) * 2010-12-16 2011-06-08 中国科学院深圳先进技术研究院 Magnetic resonance dynamic imaging method

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
HONG JUNG ET AL.: "Radial k-t FOCUSS for High-Resolution Cardiac Cine MRI.", MAGNETIC RESONANCE IN MEDICINE., vol. 63, 2010, pages 68 - 78 *
MICHAEL LUSTIG ET AL.: "Sparse MRI: The Application of Compressed Sensing for Rapid MR Imaging.", MAGNETIC RESONANCE IN MEDICINE., vol. 58, 2007, pages 1182 - 1195 *

Also Published As

Publication number Publication date
CN103027682A (en) 2013-04-10

Similar Documents

Publication Publication Date Title
Dimov et al. Bone quantitative susceptibility mapping using a chemical species–specific signal model with ultrashort and conventional echo data
WO2013086951A1 (en) Dynamic contrast enhanced magnetic resonance imaging method and system
US9305376B2 (en) Magnetic resonance imaging apparatus and method of acquiring functional image
Zhou et al. Accelerating t1ρ cartilage imaging using compressed sensing with iterative locally adapted support detection and JSENSE
US20180231626A1 (en) Systems and methods for magnetic resonance fingerprinting for quantitative breast imaging
US20150071514A1 (en) Methods and devices for optimization of magnetic resonance imaging protocols
Freiman et al. In vivo assessment of optimal b‐value range for perfusion‐insensitive apparent diffusion coefficient imaging
Aitken et al. Improved UTE‐based attenuation correction for cranial PET‐MR using dynamic magnetic field monitoring
Koolstra et al. Cartesian MR fingerprinting in the eye at 7T using compressed sensing and matrix completion‐based reconstructions
JP2015510804A (en) Reconstruction of magnetic resonance images using detection of respiratory motion during sampling in the middle and periphery of the K-space region
CN110346743B (en) Magnetic resonance diffusion weighted imaging method and device
US9632155B2 (en) Apparatus and method for conductivity and susceptibility reconstruction
JP6912603B2 (en) Dual echo Dixon type water / fat separation MR imaging
Bonanno et al. Self-navigation with compressed sensing for 2D translational motion correction in free-breathing coronary MRI: a feasibility study
JP4698231B2 (en) Magnetic resonance diagnostic equipment
US11835612B2 (en) System and method for motion correction of magnetic resonance image
JP5536358B2 (en) Magnetic resonance imaging apparatus and sensitivity correction method
Kathiravan et al. A review of magnetic resonance imaging techniques
Lugauer et al. Single-breath-hold abdominal T _ 1 T 1 mapping using 3D Cartesian Look-Locker with spatiotemporal sparsity constraints
US20170146627A1 (en) Dynamic contrast enhanced magnetic resonance imaging with flow encoding
Cao et al. Low‐rank plus sparse compressed sensing for accelerated proton resonance frequency shift MR temperature imaging
US8952693B2 (en) Method for principal frequency magnetic resonance elastography inversion
Bödenler et al. Joint multi‐field T1 quantification for fast field‐cycling MRI
US11519987B2 (en) Magnetic resonance fingerprinting thermometry
JP2007260001A (en) Magnetic resonance imaging apparatus and imaging method

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: 12857712

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 12857712

Country of ref document: EP

Kind code of ref document: A1