WO2020140912A1 - 一种非线性梯度成像的图像重建方法及其相关设备 - Google Patents
一种非线性梯度成像的图像重建方法及其相关设备 Download PDFInfo
- Publication number
- WO2020140912A1 WO2020140912A1 PCT/CN2019/130827 CN2019130827W WO2020140912A1 WO 2020140912 A1 WO2020140912 A1 WO 2020140912A1 CN 2019130827 W CN2019130827 W CN 2019130827W WO 2020140912 A1 WO2020140912 A1 WO 2020140912A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- space
- magnetic resonance
- target image
- image
- resonance imaging
- Prior art date
- Legal status (The legal status 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 status listed.)
- Ceased
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T12/00—Tomographic reconstruction from projections
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—Two-dimensional [2D] image generation
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10088—Magnetic resonance imaging [MRI]
Definitions
- the present invention relates to the technical field of software, and more specifically, to an image reconstruction method, device, electronic device, and readable storage medium for nonlinear gradient imaging.
- nonlinear gradient magnetic field imaging technology which is an emerging technology that uses additional nonlinear gradient magnetic fields to accelerate and improve magnetic resonance imaging.
- This technology can produce a mixture of linear and nonlinear The gradient magnetic field improves the imaging coding efficiency, but this technology has the problems of excessive smoothing of the reconstructed signal and loss of image details when the image is reconstructed, which leads to the problem of low quality of the reconstructed image, which will seriously affect the nonlinear gradient magnetic field imaging technology in the actual clinical Further promote and apply.
- the invention provides an image reconstruction method, device, electronic equipment and readable storage medium for nonlinear gradient imaging, which can solve the reconstruction image caused by the excessive smoothing of the reconstruction signal and the loss of image details existing in the prior art when reconstructing the image Low-quality technical issues.
- the present invention provides an image reconstruction method.
- the method includes:
- the O-Space nonlinear gradient magnetic resonance imaging signal placed at each center is calculated
- the minimization function is calculated to obtain the reconstructed target image.
- the encoding phase is:
- ⁇ (t) represents the gradient vector at time t
- the echo of O-Space nonlinear gradient magnetic resonance imaging is:
- ⁇ represents the gyromagnetic ratio
- t represents time
- the lth center placement of the target image space position is (x l , y l );
- G x ( ⁇ ), G y ( ⁇ ) and G z ( ⁇ ) are the gradients along the x, y, z 2 directions;
- the z 2 direction refers to a circumferential direction in which echoes are mapped onto concentric circles.
- the magnetic resonance signal is:
- ⁇ is the region of interest
- t is time
- the O-Space nonlinear gradient magnetic resonance imaging signals are:
- G x ( ⁇ ), G y ( ⁇ ) and G z ( ⁇ ) are the gradients along the x, y, z 2 directions;
- the z 2 direction refers to a circumferential direction in which echoes are mapped onto concentric circles.
- the O-Space nonlinear gradient magnetic resonance imaging signals placed at the centers are reduced to a minimum function, including:
- E the O-space coding matrix
- the target image be M ⁇ C N ⁇ N
- the O-Space nonlinear gradient magnetic resonance imaging signal at all center positions of the target image space be: S ⁇ C N
- the O-Space nonlinear gradient magnetic resonance imaging signal placed at each center is reduced to a minimum function:
- C represents the complex number space
- N represents the dimension of the complex number space
- ⁇ represents the regular parameter
- R(M)
- 1 , ⁇ represents the wavelet transform, the target image is M ⁇ C N ⁇ N ,
- 1 represents the L1 norm, and
- 2 represents the L2 norm.
- calculating the minimization function to obtain the reconstructed target image includes the following steps:
- Step 1 Calculate the n+1th first dual parameter according to the following formula:
- n is an integer and the initial value of n is 0, the initial value of M n is 0;
- Step 2 Calculate the n+1th second dual parameter according to the following formula:
- Step 3 Calculate the intermediate variable obtained at the n+1th iteration according to the following formula:
- Step 4 Calculate the result obtained by the n+1 iteration according to the following formula:
- Step 5 Calculate the relative error r according to the following formula:
- p, q are the first dual parameter and second dual parameter respectively;
- G 0, ⁇ , ⁇ , ⁇ are preset coefficients;
- the target image is M ⁇ C N ⁇ N
- the O-Space nonlinear gradient magnetic resonance imaging signal is S ⁇ C N ;
- C represents a complex number space
- N represents a dimension of the complex number space
- ⁇ represents a regular parameter.
- the present invention also provides an image reconstruction device.
- the image reconstruction device includes:
- the echo calculation module is used to obtain the O-Space nonlinear gradient magnetic resonance imaging echo at each center of the target image space position based on the encoding phase calculation;
- a signal calculation module used to calculate the O-Space nonlinear gradient magnetic resonance imaging signal placed at each center based on the O-Space nonlinear gradient magnetic resonance imaging echo and the magnetic resonance signal algorithm;
- a simplification module which is used to reduce the O-Space nonlinear gradient magnetic resonance imaging signals placed at each center to a minimum function
- the reconstruction module is used to calculate the minimization function based on the original dual algorithm and the approximate mapping algorithm to obtain the reconstructed target image.
- the present invention also provides an electronic device, the electronic device includes a processor, a memory, and a communication bus, wherein:
- the communication bus is used to realize the connection and communication between the processor and the memory
- the processor is used to execute the image reconstruction program stored in the memory to implement the various steps in the image reconstruction method as above.
- the present invention also provides a computer-readable storage medium that stores one or more programs, and the one or more programs can be executed by one or more processors to implement the above image Steps in the reconstruction method.
- the invention provides an image reconstruction method, device, electronic equipment and readable storage medium for nonlinear gradient imaging.
- the method includes: first obtaining the O-Space nonlinear gradient at each center of the target image space position based on the encoding phase calculation Magnetic resonance imaging echo, then based on the O-Space nonlinear gradient magnetic resonance imaging echo and the magnetic resonance signal algorithm, calculate the O-Space nonlinear gradient magnetic resonance imaging signal placed at each center, and then the center The placed O-Space nonlinear gradient magnetic resonance imaging signal is reduced to a minimization function. Finally, based on the original dual algorithm and the approximate mapping algorithm, the minimization function is calculated to obtain the reconstructed target image.
- the image reconstruction method provided by the present invention adopts an original dual algorithm that can handle nonlinear transformation, and an approximate mapping algorithm that can directly process non-smooth functions to reconstruct the target image, which can improve the need to smooth the target image in the prior art
- the approximation causes excessive smoothing of the reconstructed signal and loss of image detail, which in turn leads to the problem of low quality of the reconstructed image, which can effectively improve image clarity and maintain image detail.
- FIG. 2 is a schematic diagram of the center of a second-order nonlinear gradient magnetic field provided by the present invention
- FIG. 3 is a schematic structural diagram of an image reconstruction device provided by the present invention.
- FIG. 4 is a schematic structural diagram of an electronic device provided by the present invention.
- the invention provides an image reconstruction method, device, electronic equipment and readable storage medium for nonlinear gradient imaging, which can solve the technical problems of excessive smoothing of reconstruction signals and loss of image details when reconstructing images in the prior art.
- the method includes:
- center is placed as the center of the second-order nonlinear gradient magnetic field, as shown in FIG. 2.
- ⁇ (t) represents the gradient vector at time t
- the calculated O-Space nonlinear gradient magnetic resonance imaging echo is:
- ⁇ represents the gyromagnetic ratio
- the lth center placement of the target image space position is (x l , y l );
- G x ( ⁇ ), G y ( ⁇ ) and G z ( ⁇ ) are the gradients along the x, y, z 2 directions;
- the z 2 direction refers to a circumferential direction in which echoes are mapped onto concentric circles.
- G x ( ⁇ ) and G y ( ⁇ ) are the same as in linear imaging with radial gradient, while G z ( ⁇ ) represents a non-linear spatial code whose amplitude does not change in the circumferential direction and only changes in the radial direction magnetic field.
- the O-Space echo is the gradient encoding phase of O-Space imaging
- the O-Space imaging signal is the acquisition signal of O-Space nonlinear gradient magnetic resonance imaging.
- the magnetic resonance signal of the target image is:
- G x ( ⁇ ), G y ( ⁇ ) and G z ( ⁇ ) are the gradients along the x, y, z 2 directions;
- the z 2 direction refers to a circumferential direction in which echoes are mapped onto concentric circles.
- F(Kx) and G(x) represent the function of x
- K in F(Kx) represents the forward operation operator, which can represent either a linear operator or a non-linear operator.
- C represents the complex number space
- N represents the dimension of the complex number space
- ⁇ represents the regular parameter
- R(M)
- 1 , ⁇ represents the wavelet transform, the target image is M ⁇ C N ⁇ N ,
- 1 represents the L1 norm, and
- 2 represents the L2 norm.
- step S104 may be performing the following steps:
- Step 1 Calculate the n+1th first dual parameter according to the following formula:
- n is an integer and the initial value of n is 0, the initial value of M n is 0;
- Step 2 Calculate the n+1th second dual parameter according to the following formula:
- Step 3 Calculate the intermediate variable obtained at the n+1th iteration according to the following formula:
- Step 4 Calculate the result obtained by the n+1 iteration according to the following formula:
- Step 5 Calculate the relative error r according to the following formula:
- p, q are the first dual parameter and second dual parameter respectively;
- G 0, ⁇ , ⁇ , ⁇ are preset coefficients;
- the target image is M ⁇ C N ⁇ N
- the O-Space nonlinear gradient magnetic resonance imaging signal is S ⁇ C N ;
- C represents a complex number space
- N represents the dimension of the complex number space
- ⁇ represents a regular parameter, which is used to balance the data fidelity term and the regular term.
- H(x) function As an example to introduce the approximate mapping algorithm.
- the specific meaning of H(x) needs to be determined according to the specific problem.
- the approximate mapping prox ⁇ H of H(x) can be obtained:
- the image reconstruction method provided in this embodiment adopts an original dual algorithm that can handle nonlinear transformation, and an approximate mapping algorithm that can directly process non-smooth functions to reconstruct the target image, which can improve the need to perform target image reconstruction in the prior art.
- Smooth approximation leads to excessive smoothing of the reconstructed signal and loss of image details, which leads to the problem of low quality of the reconstructed image, which can effectively improve image clarity and maintain image details.
- the present invention also provides an image reconstruction device.
- the image reconstruction device includes:
- the echo calculation module 301 is used to obtain the O-Space nonlinear gradient magnetic resonance imaging echo at each center of the target image space position based on the encoding phase calculation;
- the signal calculation module 302 is used to calculate the O-Space nonlinear gradient magnetic resonance imaging signal placed at each center based on the O-Space nonlinear gradient magnetic resonance imaging echo and the magnetic resonance signal algorithm;
- the simplification module 303 is used to reduce the O-Space nonlinear gradient magnetic resonance imaging signals placed in the centers to a minimum function
- the reconstruction module 304 is configured to calculate the minimization function based on the original dual algorithm and the approximate mapping algorithm to obtain the reconstructed target image.
- the present invention also provides an image reconstruction electronic device.
- the electronic device includes a processor 401, a memory 402, and a communication bus 403, where:
- the communication bus 403 is used to implement connection communication between the processor 401 and the memory 402;
- the processor 401 is used to execute the image reconstruction program stored in the memory 402 to implement the various steps in the image reconstruction method as above.
- the present invention also provides a computer-readable storage medium that stores one or more programs, and the one or more programs can be executed by one or more processors to implement the above image Steps in the reconstruction method.
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
一种非线性梯度成像的图像重建方法、装置及计算机可读存储介质,该方法包括:先基于编码相位,计算得到目标图像空间位置的各中心放置处的O-Space非线性梯度磁共振成像回波(S101),又再基于磁共振信号计算得到O-Space非线性梯度磁共振成像信号,之后将其化简为最小化函数,基于原始对偶算法以及近似映射算法,计算最小化函数得到重建后的目标图像(S104)。上述图像重建方法,采用了可处理非线性变换的原始对偶算法,以及可直接处理非光滑函数的近似映射算法对目标图像进行图像重建,可以改善现有技术中需对目标图像进行光滑近似而造成重建信号过度平滑以及丢失图像细节,进而导致的重建图像质量低的问题,能够有效地提高图像清晰度和保持图像的细节。
Description
本发明涉及软件技术领域,更具体地说,涉及一种非线性梯度成像的图像重建方法、装置、电子设备及可读存储介质。
随着临床上人脑医学影像诊断的不断发展,对快速成像技术的图像质量要求越来越高,而传统的快速成像技术由于技术条件限制,而具有较差的成像效果,随着新兴的磁共振成像技术的出现和发展,提出了非线性梯度磁场成像技术,该技术是一种利用额外的非线性梯度磁场,加速和提高磁共振成像的新兴技术,该技术可以通过产生线性和非线性混合梯度磁场,提高成像编码效率,但是该技术存在重建图像时,存在重建信号过度平滑、图像细节丢失等缺陷进而导致的重建图像质量低的问题,这将严重影响非线性梯度磁场成像技术在实际临床中进一步推广和应用。
发明内容
本发明提供了一种非线性梯度成像的图像重建方法、装置、电子设备及可读存储介质,可以解决现有技术在重建图像时,存在的重建信号过度平滑和图像细节丢失而导致的重建图像质量低的技术问题。
为实现上述目的,本发明提供一种图像重建方法,该方法包括:
基于编码相位计算得到目标图像空间位置的各中心放置处的O-Space非线性梯度磁共振成像回波;
基于所述O-Space非线性梯度磁共振成像回波以及磁共振信号算法,计算得到各中心放置的O-Space非线性梯度磁共振成像信号;
将所述各中心放置的O-Space非线性梯度磁共振成像信号化简为最小化函数;
基于原始对偶算法以及近似映射算法,对所述最小化函数进行计算,得到重建后的所述目标图像。
可选的,编码相位为:
Φ(x',t)=κ
T(t)Ψ(x')
其中,所述κ(t)表示时间t的梯度向量,Ψ(x')表示所述目标图像在空间位置x'=(x,y,z)处的梯度编码场。
可选的,O-Space非线性梯度磁共振成像回波为:
其中,所述目标图像空间位置为x'=(x,y,z);
γ表示旋磁比,t表示时间;
所述目标图像空间位置的第l个中心放置处为(x
l,y
l);
G
x(τ),G
y(τ)和G
z(τ)分别是沿x,y,z
2方向的梯度;
所述z
2方向指回波映射到同心圆周上的圆周方向。
可选的,磁共振信号为:
s
q(t)=∫
ΩM(x')·C
q(x')·e
jΦ(x',t)dx'
其中,M(x')表示所述目标图像在空间位置x'=(x,y,z)处的磁化矢量;
C
q(x')表示所述目标图像在空间位置x'=(x,y,z)处的第q个线圈的线圈敏感度;
Φ(x',t)表示所述目标图像在空间位置x'=(x,y,z)处的编码相位;
Ω为感兴趣区域,t表示时间。
可选的,O-Space非线性梯度磁共振成像信号为:
其中,所述目标图像的空间位置为x'=(x,y,z);
M(x')表示所述目标图像在空间位置x'=(x,y,z)处的磁化矢量;
C
q(x')表示所述目标图像在空间位置x'=(x,y,z)处的第q个线圈的线圈敏感度;
Ω为感兴趣区域;γ为旋磁比;t表示时间;
G
x(τ),G
y(τ)和G
z(τ)分别是沿x,y,z
2方向的梯度;
所述z
2方向指回波映射到同心圆周上的圆周方向。
可选的,将所述各中心放置的O-Space非线性梯度磁共振成像信号化简为最小化函数,包括:
在离散情况下,令E为O-space编码矩阵,令所述目标图像为M∈C
N×N,令所述目标图像空间位置所有中心放置处的O-Space非线性梯度磁共振成像信号为S∈C
N,将所述各中心放置的O-Space非线性梯度磁共振成像信号化简为最小化函数:
其中,C表示复数空间,N表示所述复数空间的维度;λ表示正则参数;
R(M)=||ΓM||
1,Γ表示小波变换,所述目标图像为M∈C
N×N,||||
1表示L1范数,||||
2表示L2范数。
可选的,基于原始对偶算法以及近似映射算法,对所述最小化函数进行计算,得到重建后的所述目标图像,包括以下步骤:
步骤1、按照如下公式计算得到第n+1个第一对偶参数:
其中,n为整数且n的初始值为0,M
n初值为0;
步骤2、按照如下公式计算得到第n+1个第二对偶参数:
步骤3、按照如下公式计算第n+1次迭代得到的中间变量:
步骤4、按照如下公式进行计算第n+1次迭代得到的结果:
步骤5、按照以下公式计算得到相对误差r:
当所述相对误差r满足预设收敛范围时,确定步骤4得到的第n+1次迭代后得到的结果作为所述目标图像,否则,重复步骤1至5;
p,q分别为第一对偶参数、第二对偶参数;G=0,σ,τ,θ为预设系数;
所述目标图像为M∈C
N×N,所述O-Space非线性梯度磁共振成像信号为S∈C
N;
C表示复数空间,N表示所述复数空间的维度;λ表示正则参数。
进一步的,本发明还提供了一种图像重建装置,该图像重建装置包括:
回波计算模块,用于基于编码相位计算得到目标图像空间位置的各中心放置处的O-Space非线性梯度磁共振成像回波;
信号计算模块,用于基于所述O-Space非线性梯度磁共振成像回波以及磁共振信号算法,计算得到各中心放置的O-Space非线性梯度磁共振成像信号;
化简模块,用于将所述各中心放置的O-Space非线性梯度磁共振成像信号化简为最小化函数;
重建模块,用于基于原始对偶算法以及近似映射算法,对所述最小化函数进行计算,得到重建后的所述目标图像。
进一步的,本发明还提供了一种电子设备,该电子设备包括处理器、存储器及通信总线,其中:
通信总线用于实现处理器和存储器之间的连接通信;
处理器用于执行存储器中存储的图像重建程序,以实现如上的图像重建方法中的各个步骤。
进一步的,本发明还提供了一种计算机可读存储介质,该计算机可读存储介质存储有一个或者多个程序,一个或者多个程序可被一个或者多个处理器执行,以实现如上的图像重建方法中的各个步骤。
本发明提供一种非线性梯度成像的图像重建方法、装置、电子设备及可读存储介质,该方法包括:先基于编码相位计算得到目标图像空间位置的各中心放置处的O-Space非线性梯度磁共振成像回波,随后基于所述O-Space非线性梯度磁共振成像回波以及磁共振信号算法,计算得到各中心放置的O-Space非 线性梯度磁共振成像信号,再将所述各中心放置的O-Space非线性梯度磁共振成像信号化简为最小化函数,最后,基于原始对偶算法以及近似映射算法,对所述最小化函数进行计算,得到重建后的所述目标图像。本发明提供的图像重建方法,采用了可处理非线性变换的原始对偶算法,以及可直接处理非光滑函数的近似映射算法对目标图像进行图像重建,可以改善现有技术中需对目标图像进行光滑近似而造成重建信号过度平滑以及丢失图像细节,进而导致的重建图像质量低的问题,能够有效地提高图像清晰度和保持图像的细节。
为了更清楚地说明本发明实施例或现有技术中的技术方案,下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本发明的一些实施例,对于本领域技术人员来讲,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的附图。
图1为本发明提供的一种图像重建方法的基本流程图;
图2为本发明提供的二阶非线性梯度磁场中心示意图;
图3为本发明提供的一种图像重建装置的结构示意图;
图4为本发明提供的一种电子设备的结构示意图。
为使得本发明的发明目的、特征、优点能够更加的明显和易懂,下面将结合本发明实施例中的附图,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,而非全部实施例。基于本发明中的实施例,本领域技术人员在没有做出创造性劳动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。
本发明提供了一种非线性梯度成像的图像重建方法、装置、电子设备及可读存储介质,可以解决现有技术中重建图像时,重建信号过度平滑和图像细节丢失的技术问题。参见图1,该方法包括:
S101、基于编码相位计算得到目标图像空间位置的各中心放置处的O-Space非线性梯度磁共振成像回波。
需要理解的是,中心放置为二阶非线性梯度磁场中心,可参见图2所示。
需要了解的是,编码相位为:
Φ(x',t)=κ
T(t)Ψ(x')
其中,κ(t)表示时间t的梯度向量,Ψ(x')表示目标图像在空间位置x'=(x,y,z)处的梯度编码场。
计算得到的O-Space非线性梯度磁共振成像回波为:
其中,目标图像空间位置为x'=(x,y,z),t表示时间;
γ表示旋磁比,目标图像空间位置的第l个中心放置处为(x
l,y
l);
G
x(τ),G
y(τ)和G
z(τ)分别是沿x,y,z
2方向的梯度;
所述z
2方向指回波映射到同心圆周上的圆周方向。
需要了解的是,G
x(τ),G
y(τ)与线性梯度的径向成像中一样,而G
z(τ)表示幅度在周向不变,只沿径向改变的非线性空间编码磁场。
S102、基于所述O-Space非线性梯度磁共振成像回波以及磁共振信号算法,计算得到各中心放置的O-Space非线性梯度磁共振成像信号。
需要了解的是,O-Space回波为O-Space成像的梯度编码相位,O-Space成像信号为O-Space非线性梯度磁共振成像的采集信号。
在一些示例下,目标图像的磁共振信号为:
s
q(t)=∫
ΩM(x')·C
q(x')·e
jΦ(x',t)dx'
其中,M(x')表示目标图像在空间位置x'=(x,y,z)处的磁化矢量;
C
q(x')表示所述目标图像在空间位置x'=(x,y,z)处的第q个线圈的线圈敏感度;
Φ(x',t)表示所述目标图像在空间位置x'=(x,y,z)处的编码相位;
Ω为感兴趣区域;γ为旋磁比;t表示时间
G
x(τ),G
y(τ)和G
z(τ)分别是沿x,y,z
2方向的梯度;
所述z
2方向指回波映射到同心圆周上的圆周方向。
S103、将所述各中心放置的O-Space非线性梯度磁共振成像信号化简为最小化函数。
具体的,在离散情况下,可以令E为O-space编码矩阵,令目标图像为M∈C
N×N,令O-Space非线性梯度磁共振成像信号为S∈C
N,则可以将O-Space非线性梯度磁共振成像信号S
l,q(t)化简为最小化函数:
需要了解的是,
表示重建图像,求解等号右边的最小化函数可以得到重建目标图像的最终结果。一般情况下,可以采用原始对偶算法解决,需要了解的是,在图像处理领域经常用到原始对偶算法,该算法可以用来解决形式如:
的最小化问题,其中,F(Kx)、G(x)表示x的函数,F(Kx)中的K表示前向操作算子,既可以表示线性算子也可以表示非线性算子。
其中,C表示复数空间,N表示复数空间的维度;λ表示正则参数;
R(M)=||ΓM||
1,Γ表示小波变换,目标图像为M∈C
N×N,||||
1表示L1范数,||||
2表示L2范数。
需要了解的是,M与M(x')意义相同,表示目标图像,其中x'=(x,y,z)表示目标图像的空间位置,需要明白的是,图像是空间位置的函数,故目标图像可以表示为M和M(x')。
S104、基于原始对偶算法以及近似映射算法,对所述最小化函数进行计算,得到重建后的所述目标图像。
具体的,在上述的离散情况下,步骤S104可以为执行以下步骤:
步骤1、按照如下公式计算得到第n+1个第一对偶参数:
其中,n为整数且n的初始值为0,M
n初值为0;
步骤2、按照如下公式计算得到第n+1个第二对偶参数:
步骤3、按照如下公式计算第n+1次迭代得到的中间变量:
步骤4、按照如下公式进行计算第n+1次迭代得到的结果:
步骤5、按照以下公式计算得到相对误差r:
当所述相对误差r满足预设收敛范围时,确定步骤4得到的第n+1次迭代后得到的结果作为所述目标图像,否则,重复步骤1至5;
p,q分别为第一对偶参数、第二对偶参数;G=0,σ,τ,θ为预设系数;
所述目标图像为M∈C
N×N,所述O-Space非线性梯度磁共振成像信号为S∈C
N;
C表示复数空间,N表示所述复数空间的维度;λ表示正则参数,用来平衡数据保真项和正则项。
此处以H(x)函数为例介绍近似映射算法,H(x)具体意义则需要根据具体问题确定,基于H(x)函数以及近似映射算法可以得到H(x)的近似映射prox
σH:
需要了解的是,近似映射算法可直接处理非光滑函数,而无需对函数进行光滑近似。
本实施例提供的图像重建方法,采用了可处理非线性变换的原始对偶算法,以及可直接处理非光滑函数的近似映射算法对目标图像进行图像重建,可以改善现有技术中需对目标图像进行光滑近似而造成重建信号过度平滑以及丢失图 像细节,进而导致的重建图像质量低的问题,能够有效地提高图像清晰度和保持图像的细节。
进一步的,本发明还提供了一种图像重建装置,参见图3该图像重建装置包括:
回波计算模块301,用于基于编码相位计算得到目标图像空间位置的各中心放置处的O-Space非线性梯度磁共振成像回波;
信号计算模块302,用于基于所述O-Space非线性梯度磁共振成像回波以及磁共振信号算法,计算得到各中心放置的O-Space非线性梯度磁共振成像信号;
化简模块303,用于将所述各中心放置的O-Space非线性梯度磁共振成像信号化简为最小化函数;
重建模块304,用于基于原始对偶算法以及近似映射算法,对所述最小化函数进行计算,得到重建后的所述目标图像。
进一步的,本发明还提供了一种图像重建电子设备,参见图4该电子设备包括处理器401、存储器402及通信总线403,其中:
通信总线403用于实现处理器401和存储器402之间的连接通信;
处理器401用于执行存储器402中存储的图像重建程序,以实现如上的图像重建方法中的各个步骤。
进一步的,本发明还提供了一种计算机可读存储介质,该计算机可读存储介质存储有一个或者多个程序,一个或者多个程序可被一个或者多个处理器执行,以实现如上的图像重建方法中的各个步骤。
需要说明的是,对于前述的各方法实施例,为了简便描述,故将其都表述为一系列的动作组合,但是本领域技术人员应该知悉,本发明并不受所描述的动作顺序的限制,因为依据本发明,某些步骤可以采用其它顺序或者同时进行。其次,本领域技术人员也应该知悉,说明书中所描述的实施例均属于优选实施例,所涉及的动作和模块并不一定都是本发明所必须的。
在上述实施例中,对各个实施例的描述都各有侧重,某个实施例中没有详述的部分,可以参见其它实施例的相关描述,同时,上述本发明实施例序号仅仅为了描述,不代表实施例的优劣,本领域的普通技术人员在本发明的启示下,在不脱离本发明宗旨和权利要求所保护的范围情况下,还可做出很多形式,这些均属于本发明的保护之内。
Claims (10)
- 一种非线性梯度成像的图像重建方法,其特征在于,所述图像重建方法包括:基于编码相位计算得到目标图像空间位置的各中心放置处的O-Space非线性梯度磁共振成像回波;基于所述O-Space非线性梯度磁共振成像回波以及磁共振信号算法,计算得到各中心放置的O-Space非线性梯度磁共振成像信号;将所述各中心放置的O-Space非线性梯度磁共振成像信号化简为最小化函数;基于原始对偶算法以及近似映射算法,对所述最小化函数进行计算,得到重建后的所述目标图像。
- 如权利要求1所述的图像重建方法,其特征在于,所述编码相位为:Φ(x',t)=κ T(t)Ψ(x')其中,所述κ(t)表示时间t的梯度向量,Ψ(x')表示所述目标图像在空间位置x'=(x,y,z)处的梯度编码场。
- 如权利要求1所述的图像重建方法,其特征在于,所述磁共振信号为:s q(t)=∫ ΩM(x')·C q(x')·e jΦ(x',t)dx'其中,M(x')表示所述目标图像在空间位置x'=(x,y,z)处的磁化矢量;C q(x')表示所述目标图像在空间位置x'=(x,y,z)处的第q个线圈的线圈敏感度;Φ(x',t)表示所述目标图像在空间位置x'=(x,y,z)处的编码相位;Ω为感兴趣区域,t表示时间。
- 如权利要求1-5任一项所述的图像重建方法,其特征在于,所述基于原始对偶算法以及近似映射算法,对所述最小化函数进行计算,得到重建后的所述目标图像,包括以下步骤:步骤1、按照如下公式计算得到第n+1个第一对偶参数:其中,n为整数且n的初始值为0,M n初值为0;步骤2、按照如下公式计算得到第n+1个第二对偶参数:步骤3、按照如下公式计算第n+1次迭代得到的中间变量:步骤4、按照如下公式进行计算第n+1次迭代得到的结果:步骤5、按照以下公式计算得到相对误差r:当所述相对误差r满足预设收敛范围时,确定步骤4得到的第n+1次迭代后得到的结果作为所述目标图像,否则,重复步骤1至5;p,q分别为第一对偶参数、第二对偶参数;G=0,σ,τ,θ为预设系数;所述目标图像为M∈C N×N,所述O-Space非线性梯度磁共振成像信号为S∈C N;C表示复数空间,N表示所述复数空间的维度;λ表示正则参数。
- 一种图像重建装置,所述图像重建装置包括:回波计算模块,用于基于编码相位计算得到目标图像空间位置的各中心放置处的O-Space非线性梯度磁共振成像回波;信号计算模块,用于基于所述O-Space非线性梯度磁共振成像回波以及磁共振信号算法,计算得到各中心放置的O-Space非线性梯度磁共振成像信号;化简模块,用于将所述各中心放置的O-Space非线性梯度磁共振成像信号化简为最小化函数;重建模块,用于基于原始对偶算法以及近似映射算法,对所述最小化函数进行计算,得到重建后的所述目标图像。
- 一种电子设备,其特征在于,所述电子设备包括处理器、存储器及通信总线,其中:通信总线用于实现处理器和存储器之间的连接通信;处理器用于执行存储器中存储的图像重建程序,以实现权利要求1-7任一项所述的图像重建方法中的各个步骤。
- 一种可读存储介质,其特征在于,所述可读存储介质为计算机可读存储介质,所述计算机可读存储介质存储有一个或者多个程序,所述一个或者多个程序可被一个或者多个处理器执行,以实现如权利要求1-7任一项所述的图像重建方法中的各个步骤。
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811535975 | 2018-12-15 | ||
| CN201910007827.8 | 2019-01-04 | ||
| CN201910007827.8A CN111325808B (zh) | 2018-12-15 | 2019-01-04 | 一种非线性梯度成像的图像重建方法及其相关设备 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2020140912A1 true WO2020140912A1 (zh) | 2020-07-09 |
Family
ID=71166815
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2019/130827 Ceased WO2020140912A1 (zh) | 2018-12-15 | 2019-12-31 | 一种非线性梯度成像的图像重建方法及其相关设备 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN111325808B (zh) |
| WO (1) | WO2020140912A1 (zh) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023028786A1 (zh) * | 2021-08-30 | 2023-03-09 | 深圳先进技术研究院 | 一种设计非线性梯度线圈的方法以及非线性空间编码方法 |
| CN121168178B (zh) * | 2025-11-21 | 2026-03-03 | 东南大学 | 一种用于磁共振成像的梯度线圈设计简化方法及系统 |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2010068299A1 (en) * | 2008-12-12 | 2010-06-17 | Yale University | Parallel magnetic resonance imaging using nonlinear gradient encoding |
| CN102389309A (zh) * | 2011-07-08 | 2012-03-28 | 首都医科大学 | 基于压缩感知理论的磁共振图像重建的方法 |
| CN102947721A (zh) * | 2010-06-23 | 2013-02-27 | 皇家飞利浦电子股份有限公司 | 考虑高阶动态场重建对象的磁共振图像的方法 |
| CN106646302A (zh) * | 2015-10-29 | 2017-05-10 | 上海联影医疗科技有限公司 | 一种磁共振成像方法 |
| CN107690309A (zh) * | 2016-11-08 | 2018-02-13 | 中国科学院深圳先进技术研究院 | 磁共振血管壁成像方法和设备 |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101975935B (zh) * | 2010-09-03 | 2012-12-05 | 杭州电子科技大学 | 一种基于部分回波压缩感知的快速磁共振成像方法 |
| US20120153952A1 (en) * | 2010-12-16 | 2012-06-21 | General Electric Company | Method and system for gradient linear correction |
| CN106772168B (zh) * | 2017-02-24 | 2019-07-02 | 深圳先进技术研究院 | 磁共振成像方法及装置 |
| CN108802648B (zh) * | 2018-04-03 | 2020-12-01 | 上海东软医疗科技有限公司 | 一种基于梯度回波的磁共振定量成像方法和装置 |
-
2019
- 2019-01-04 CN CN201910007827.8A patent/CN111325808B/zh active Active
- 2019-12-31 WO PCT/CN2019/130827 patent/WO2020140912A1/zh not_active Ceased
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2010068299A1 (en) * | 2008-12-12 | 2010-06-17 | Yale University | Parallel magnetic resonance imaging using nonlinear gradient encoding |
| CN102947721A (zh) * | 2010-06-23 | 2013-02-27 | 皇家飞利浦电子股份有限公司 | 考虑高阶动态场重建对象的磁共振图像的方法 |
| CN102389309A (zh) * | 2011-07-08 | 2012-03-28 | 首都医科大学 | 基于压缩感知理论的磁共振图像重建的方法 |
| CN106646302A (zh) * | 2015-10-29 | 2017-05-10 | 上海联影医疗科技有限公司 | 一种磁共振成像方法 |
| CN107690309A (zh) * | 2016-11-08 | 2018-02-13 | 中国科学院深圳先进技术研究院 | 磁共振血管壁成像方法和设备 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111325808B (zh) | 2021-02-26 |
| CN111325808A (zh) | 2020-06-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20190369190A1 (en) | Method for processing interior computed tomography image using artificial neural network and apparatus therefor | |
| TW202011893A (zh) | 用於核磁共振影像重建之深度學習技術 | |
| Aghabiglou et al. | Projection-Based cascaded U-Net model for MR image reconstruction | |
| WO2020134826A1 (zh) | 磁共振并行成像方法及相关设备 | |
| WO2020114329A1 (zh) | 磁共振快速参数成像方法及装置 | |
| WO2021077557A1 (zh) | 一种磁共振图像重建方法、装置、设备和介质 | |
| CN101887581B (zh) | 图像融合方法及设备 | |
| CN104739410B (zh) | 一种磁共振图像的迭代重建方法 | |
| CN110148215A (zh) | 一种基于平滑约束和局部低秩约束模型的四维磁共振图像重建方法 | |
| CN103584864A (zh) | 一种磁共振成像方法和装置 | |
| CN106251398A (zh) | 一种图像重建方法和装置 | |
| WO2020140912A1 (zh) | 一种非线性梯度成像的图像重建方法及其相关设备 | |
| CN113050009B (zh) | 三维磁共振快速参数成像方法和装置 | |
| US20230267659A1 (en) | Machine-learning techniques for sparse-to-dense spectral reconstruction | |
| CN114037631B (zh) | 一种图像纹理增强方法、装置、电子设备及存储介质 | |
| Datta et al. | A deep framework for enhancement of diagnostic information in CSMRI reconstruction | |
| CN118379383B (zh) | 基于傅里叶卷积的磁共振图像重建方法及装置 | |
| US20160253825A1 (en) | System and method for model-based reconstruction of quantitative images | |
| 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 | |
| CN119963678A (zh) | 基于灵敏度矩阵约束的隐式神经网络磁共振图像重建方法 | |
| US20230267306A1 (en) | Machine-learning techniques for representing items in a spectral domain | |
| CN116977458A (zh) | 一种产生多通道空间支持图的方法与系统 | |
| CN115272101B (zh) | 基于多尺度频域滤波的红外图像光学非均匀性校正方法 | |
| Mathew et al. | Compressed sensing parallel MRI with adaptive shrinkage TV regularization | |
| KR102163220B1 (ko) | 뉴럴 네트워크를 이용한 자기공명 혈관 조영술 영상 처리 방법 및 장치 |
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: 19907655 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 10.11.2021) |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 19907655 Country of ref document: EP Kind code of ref document: A1 |





















