WO2016090686A1 - 磁共振弥散张量去噪方法和装置以及计算机程序产品 - Google Patents
磁共振弥散张量去噪方法和装置以及计算机程序产品 Download PDFInfo
- Publication number
- WO2016090686A1 WO2016090686A1 PCT/CN2014/095362 CN2014095362W WO2016090686A1 WO 2016090686 A1 WO2016090686 A1 WO 2016090686A1 CN 2014095362 W CN2014095362 W CN 2014095362W WO 2016090686 A1 WO2016090686 A1 WO 2016090686A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- diffusion
- diffusion tensor
- tensor
- maximum
- maximum likelihood
- Prior art date
Links
- 238000009792 diffusion process Methods 0.000 title claims abstract description 176
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000004590 computer program Methods 0.000 title claims abstract description 10
- 238000007476 Maximum Likelihood Methods 0.000 claims abstract description 41
- 239000006185 dispersion Substances 0.000 claims description 29
- 230000006870 function Effects 0.000 claims description 15
- 239000011159 matrix material Substances 0.000 claims description 10
- 238000002598 diffusion tensor imaging Methods 0.000 description 10
- 238000004891 communication Methods 0.000 description 5
- 230000009897 systematic effect Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000005070 sampling Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 239000000835 fiber Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 238000002939 conjugate gradient method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000001727 in vivo Methods 0.000 description 1
- 238000002595 magnetic resonance imaging Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
- 210000004885 white matter Anatomy 0.000 description 1
Images
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
- G01R33/563—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
- G01R33/56341—Diffusion imaging
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/05—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves
- A61B5/055—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging
-
- 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
- G01R33/5602—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution by filtering or weighting based on different relaxation times within the sample, e.g. T1 weighting using an inversion pulse
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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]
- G06T2207/10092—Diffusion tensor magnetic resonance imaging [DTI]
Definitions
- the present invention relates to the field of magnetic resonance imaging technologies, and in particular, to a magnetic resonance diffusion tensor denoising method and apparatus, and a computer program product.
- DTI Diffusion Tensor Imaging
- DTI is the only non-invasive imaging method that can display white matter fiber bundles in vivo.
- the signal-to-noise ratio of DTI is relatively low, which greatly limits the wide clinical application of DTI.
- d denotes the signal acquired on the magnetic resonance apparatus
- F denotes the Boller coding matrix
- ⁇ denotes a diffusion-weighted image
- n is usually assumed to be complex Gaussian white noise.
- the mth diffusion-weighted image ⁇ m can be expressed as:
- I 0 ⁇ R N ⁇ 1 represents a reference image without dispersion weighting
- b denotes the dispersion weighting factor (constant)
- g m denotes the diffusion gradient vector corresponding to the m-th diffusion-weighted image
- D Indicates the diffusion tensor, the specific D can be expressed as:
- each D i ⁇ R N ⁇ 1 , N represents the number of pixel points of the magnetic resonance image.
- the process of diffusion tensor imaging mainly includes the following steps:
- Step 1 reconstructing the image ⁇ from the acquired signal d;
- Step 2 Estimating the tensor matrix D corresponding to each spatial position according to the model tensor (ie, the above formula 2);
- Step 3 According to the following formula 4 and formula 5, various diffusion parameters are calculated from the tensor matrix D, for example: MD, FA, etc.:
- ⁇ 1 , ⁇ 2 , and ⁇ 3 are the eigenvalues of the matrix D.
- the current direct method is to average or reduce the sampling area of K-space by multiple sampling. Although these methods can improve the signal-to-noise ratio in practical applications, these methods increase the scan time or affect the spatial resolution.
- the method After acquiring the scan data of the K space, the method first reconstructs the diffusion weighted image. The image is then denoised using signal processing. Finally, the diffusion tensor and various diffusion parameters are calculated from the denoised image. That is, in the process of processing, it is necessary to reconstruct the image first, and then denoise the image. However, the systematic errors generated during the image denoising process are often further transferred to subsequent tensor calculations, thereby affecting the accuracy of the diffusion tensor.
- the embodiment of the invention provides a magnetic resonance diffusion tensor denoising method, which solves the technical problem that the high-precision denoising of the diffusion tensor can be realized without increasing the scanning time and not affecting the spatial resolution in the prior art.
- the method includes:
- the maximum likelihood estimate obtained as the initial value is used to calculate the maximum posterior probability estimate of the diffusion tensor
- the diffusion coefficient is calculated based on the estimated maximum posterior probability estimate.
- the maximum likelihood estimate of the diffusion tensor is obtained from the acquired K-space data, including:
- a maximum likelihood estimate representing the diffusion tensor d m representing the K-space data corresponding to the mth diffusion-weighted image, F representing the Venezuela coding matrix, and I 0 representing the reference image without diffusion weighting,
- d m the K-space data corresponding to the mth diffusion-weighted image
- F the Venezuela coding matrix
- I 0 the reference image without diffusion weighting
- b the diffusion weighting factor
- g m (g xm , g ym , g zm ) T denotes the mth dispersion gradient vector.
- the obtained maximum likelihood estimate is used as an initial value, and the maximum a posteriori probability estimate of the diffusion tensor is calculated, including:
- R( ⁇ ) denotes a penalty function acting on the diffusion tensor D
- G( ⁇ ) denotes a penalty function acting on the diffusion parameter FA
- ⁇ 1 and ⁇ 2 denote regularization parameters
- D 1 , D 2 , D 3 , D 4 , D 5 , D 6 represent 6 diffusion tensors in different directions.
- R( ⁇ ) constrains the sparsity of each diffusion tensor, or R( ⁇ ) constrains the joint sparsity of the diffusion tensor.
- ⁇ is an operator, indicating a sparse transform
- ⁇ i represents the ith column of ⁇
- 1 represents the L1 norm
- 2 represents the L2 norm.
- the embodiment of the invention further provides a magnetic resonance diffusion tensor denoising device, which solves the technology of realizing high-precision denoising of the diffusion tensor without increasing the scanning time and not affecting the spatial resolution in the prior art.
- the device includes:
- a memory for storing instructions
- processor coupled to the memory, the processor being arranged to execute instructions stored in the memory, wherein the processor is configured to:
- the maximum likelihood estimate obtained is taken as the initial value, and the maximum posterior probability estimate of the diffusion tensor is calculated;
- the diffusion coefficient is calculated based on the obtained maximum posterior probability estimate.
- the processor is specifically configured to calculate a maximum likelihood estimate of the diffusion tensor according to the following formula:
- a maximum likelihood estimate representing the diffusion tensor d m representing the K-space data corresponding to the mth diffusion-weighted image, F representing the Venezuela coding matrix, and I 0 representing the reference image without diffusion weighting,
- d m the K-space data corresponding to the mth diffusion-weighted image
- F the Venezuela coding matrix
- I 0 the reference image without diffusion weighting
- b the diffusion weighting factor
- g m (g xm , g ym , g zm ) T denotes the mth gradient vector.
- the processor is specifically configured to calculate a maximum a posteriori probability estimate of the diffusion tensor according to the following formula:
- R( ⁇ ) denotes a penalty function acting on the diffusion tensor D
- G( ⁇ ) denotes a penalty function acting on the diffusion parameter FA
- ⁇ 1 and ⁇ 2 denote regularization parameters
- D 1 , D 2 , D 3 , D 4 , D 5 , D 6 represent 6 diffusion tensors in different directions.
- R( ⁇ ) constrains the sparsity of each diffusion tensor, or R( ⁇ ) constrains the joint sparsity of the diffusion tensor.
- ⁇ is an operator, indicating a sparse transform
- ⁇ i represents the ith column of ⁇
- 1 represents the L1 norm
- 2 represents the L2 norm.
- the embodiment of the invention further provides a computer program product for magnetic resonance diffusion tensor denoising, comprising:
- a computer readable storage medium having computer readable program code embodied in the medium, the computer readable program code comprising:
- Computer readable program code configured to acquire K-space data
- a computer readable program code configured to obtain a maximum likelihood estimate of the diffusion tensor based on the acquired K-space data
- Computer readable program code configured to utilize the sparsity of the diffusion tensor and the dispersion parameter, the maximum likelihood estimate obtained as an initial value, and the maximum a posteriori probability estimate of the diffusion tensor;
- a computer readable program code configured to calculate a diffusion coefficient based on the determined maximum a posteriori probability estimate.
- the maximum likelihood estimation value of the diffusion tensor is directly determined according to the K-space data, and then the diffusion tensor and the diffusion parameter are combined according to the obtained maximum likelihood estimation value.
- Sparseness calculates the maximum a posteriori probability estimate of the diffusion tensor to determine the diffusion tensor after denoising. This is different from the method of directly denoising the diffusion-weighted image, avoiding the systematic error introduced by image denoising, and solving the problem that the prior art can not achieve the dispersion without increasing the scanning time and not affecting the spatial resolution.
- the technical problem of high-precision denoising of tensor effectively suppresses the noise in the diffusion tensor. Further, since the denoising is performed by the dispersion of the tensor and the dispersion parameter, the estimation accuracy of the diffusion tensor is more effectively improved.
- FIG. 1 is a flow chart of a method of a magnetic resonance diffusion tensor denoising method according to an embodiment of the present invention
- FIG. 2 is a block diagram showing the structure of a magnetic resonance diffusion tensor denoising device according to an embodiment of the present invention
- FIG. 3 is a schematic block diagram of a computer program product in accordance with an embodiment of the present invention.
- Step 101 Acquire K-space data
- Step 102 Calculate a maximum likelihood estimation value of the diffusion tensor according to the collected K-space data
- Step 103 calculating the maximum posterior probability estimate of the diffusion tensor by using the sparseness of the diffusion tensor and the dispersion parameter to obtain the maximum likelihood estimation value as the initial value;
- Step 104 Calculate a diffusion coefficient according to the obtained maximum posterior probability estimation value, where the diffusion coefficient may include: MD, FA.
- the maximum likelihood estimation value of the diffusion tensor is directly determined from the K-space data. Then, the maximum likelihood estimate obtained is taken as the initial value, and the maximum posterior probability estimate of the diffusion tensor is calculated according to the sparsity of the diffusion tensor and the dispersion parameter, and then the diffusion tensor after denoising is determined.
- This is different from the method of denoising the diffusion-weighted image, avoiding the systematic error introduced by image denoising, and solving the problem that the prior art cannot realize the diffusion tensor without increasing the scanning time and not affecting the spatial resolution.
- the technical problem of high-precision denoising effectively suppresses the noise in the diffusion tensor. Further, since the denoising is performed by the dispersion of the tensor and the dispersion parameter, the estimation accuracy of the diffusion tensor is more effectively improved.
- a maximum likelihood estimate representing the diffusion tensor d m representing the K-space data corresponding to the mth diffusion-weighted image, F representing the Venezuela coding matrix, and I 0 representing the reference image without diffusion weighting,
- d m the K-space data corresponding to the mth diffusion-weighted image
- F the Venezuela coding matrix
- I 0 the reference image without diffusion weighting
- b the diffusion weighting factor
- g m (g xm , g ym , g zm ) T denotes the diffusion gradient vector corresponding to the mth diffusion-weighted image.
- the maximum a posteriori probability estimate of the diffusion tensor can be calculated according to the following formula:
- R( ⁇ ) denotes a penalty function acting on the diffusion tensor D
- G( ⁇ ) denotes a penalty function acting on the diffusion parameter
- ⁇ 1 and ⁇ 2 denote regularization parameters
- D 1 , D 2 , D 3 , D 4 , D 5 , D 6 represent 6 diffusion tensors in different directions. Assuming that both D and FA are sparse, then you can make:
- the operator ⁇ represents a sparse transform (for example, wavelet transform, etc.)
- ⁇ i represents the ith column of ⁇
- 1 represents the L1 norm, which is used to constrain the diffusion tensor in the sparse transform domain.
- 2 means to find the L2 norm.
- the denoising problem of the diffusion tensor is equivalent to the above optimization problem.
- the specific can be solved by the nonlinear conjugate gradient method.
- the diffusion tensor is directly denoised by the acquired K-space data, which skips the tradition.
- the image denoising step avoids the influence of the image denoising error on the tensor estimation, can more effectively suppress the noise in the diffusion tensor, and improve the estimation accuracy of the diffusion tensor.
- a magnetic resonance diffusion tensor denoising device is also provided in the embodiment of the present invention, as described in the following embodiments. Since the principle of solving the problem of the magnetic resonance diffusion tensor denoising device is similar to the magnetic resonance diffusion tensor denoising method, the implementation of the magnetic resonance diffusion tensor denoising device can be implemented by the implementation of the magnetic resonance diffusion tensor denoising method. I won't go into details here.
- the term "unit” or “module” may implement a combination of software and/or hardware of a predetermined function.
- the apparatus described in the following embodiments is preferably implemented in software, hardware, or a combination of software and hardware, is also possible and contemplated.
- FIG. 2 is a structural block diagram of a magnetic resonance diffusion tensor denoising apparatus 200 according to an embodiment of the present invention.
- the system includes: a memory 201 for storing instructions; a processor 202 coupled to the memory 201, and a processor 202 is arranged to execute an instruction stored in the memory 201, wherein the processor 202 is configured to: acquire K-space data; to obtain a maximum likelihood estimate of the diffusion tensor based on the acquired K-space data; The sparsity of the tensor and dispersion parameters, the maximum likelihood to be obtained The estimated value is used as an initial value, and a maximum posterior probability estimation value of the diffusion tensor is calculated; and the diffusion coefficient is calculated according to the obtained maximum posterior probability estimation value.
- the processor 202 is specifically configured to calculate a maximum likelihood estimate of the dispersion tensor according to the following formula:
- a maximum likelihood estimate representing the diffusion tensor d m representing the K-space data corresponding to the mth diffusion-weighted image, F representing the Venezuela coding matrix, and I 0 representing the reference image without diffusion weighting,
- d m the K-space data corresponding to the mth diffusion-weighted image
- F the Venezuela coding matrix
- I 0 the reference image without diffusion weighting
- b the diffusion weighting factor
- g m (g xm , g ym , g zm ) T denotes the mth gradient vector.
- the processor 202 is specifically configured to calculate a maximum a posteriori probability estimate of the diffusion tensor according to the following formula:
- R( ⁇ ) denotes a penalty function acting on the diffusion tensor D
- G( ⁇ ) denotes a penalty function acting on the diffusion parameter FA
- ⁇ 1 and ⁇ 2 denote regularization parameters
- D 1 , D 2 , D 3 , D 4 , D 5 , D 6 represent 6 diffusion tensors in different directions.
- R( ⁇ ) constrains the sparsity of each diffusion tensor, or R( ⁇ ) constrains the joint sparsity of the diffusion tensor.
- ⁇ is an operator, indicating a sparse transform
- ⁇ i represents the ith column of ⁇
- 1 represents the L1 norm
- 2 represents the L2 norm.
- FIG. 3 is a block diagram showing an exemplary computer program product 300.
- computer program product 300 includes a computer readable storage medium 3001 having stored thereon computer readable program code 30011 embodied in the medium,
- the computer readable program code 30011 includes computer readable program code configured to acquire K-space data, computer readable program code configured to obtain a maximum likelihood estimate of a diffusion tensor based on the acquired K-space data , configured to utilize the sparseness of the diffusion tensor and the dispersion parameter, using the maximum likelihood estimate obtained as the initial value, and calculating the maximum posterior summary of the diffusion tensor
- Computer readable program code for a rate estimate and computer readable program code configured to calculate a diffusion coefficient based on the determined maximum a posteriori probability estimate.
- computer program product 300 can include one or more of recordable medium 3002 and communication medium 3003.
- the computer readable storage medium 3001 and the recordable medium 3002 may include, but are not limited to, a floppy disk, a hard disk drive (HDD), a compact disk (CD), a digital versatile disk (DVD), a digital tape, a computer memory, and the like.
- Communication medium 3003 may include, but is not limited to, digital and/or analog communication media (eg, fiber optic cables, waveguides, wired communication links, wireless communication links, and the like).
- the embodiment of the present invention achieves the following technical effects: after acquiring the K-space data, the maximum likelihood estimation value of the diffusion tensor is directly determined according to the K-space data, and then the maximum is obtained according to the method.
- the likelihood estimate combined with the sparsity of the diffusion tensor and the dispersion parameter, calculates the maximum posterior probability estimate of the diffusion tensor to determine the diffusion tensor after denoising. This is different from the method of directly denoising the diffusion-weighted image, avoiding the systematic error introduced by image denoising, and solving the problem that the prior art can not achieve the dispersion without increasing the scanning time and not affecting the spatial resolution.
- the technical problem of high-precision denoising of tensor effectively suppresses the noise in the diffusion tensor. Further, since the denoising is performed by the dispersion of the tensor and the dispersion parameter, the estimation accuracy of the diffusion tensor is more effectively improved.
- modules or steps of the embodiments of the present invention can be implemented by a general computing device, which can be concentrated on a single computing device or distributed in multiple computing devices. Alternatively, they may be implemented by program code executable by the computing device such that they may be stored in the storage device by the computing device and, in some cases, may be different from The steps shown or described are performed sequentially, or they are separately fabricated into individual integrated circuit modules, or a plurality of modules or steps thereof are fabricated into a single integrated circuit module. Thus, embodiments of the invention are not limited to any specific combination of hardware and software.
Landscapes
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Radiology & Medical Imaging (AREA)
- High Energy & Nuclear Physics (AREA)
- General Physics & Mathematics (AREA)
- Signal Processing (AREA)
- Condensed Matter Physics & Semiconductors (AREA)
- Vascular Medicine (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Theoretical Computer Science (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
- Image Processing (AREA)
Abstract
一种磁共振弥散张量去噪方法和装置(200)以及计算机程序产品(300),其中,磁共振弥散张量去噪方法包括:采集K空间数据;根据采集的K空间数据,求取弥散张量的最大似然估计值;利用弥散张量和弥散参数的稀疏性,将求取的最大似然估计值作为初始值,计算弥散张量的最大后验概率估计值;根据求取的所述最大后验概率估计值,计算弥散系数。磁共振弥散张量去噪方法和装置(200)以及计算机程序产品(300)解决了现有技术中在不增加扫描时间和不影响空间分辨率的同时实现弥散张量的高精度去噪的技术问题,有效抑制了弥散张量中的噪声,提高了弥散张量的估计精度。
Description
本申请要求2014年12月11日递交的申请号为201410758290.6、发明名称为“磁共振弥散张量去噪方法和装置”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
本发明涉及磁共振成像技术领域,特别涉及一种磁共振弥散张量去噪方法和装置以及计算机程序产品。
磁共振弥散张量成像(Diffusion Tensor Imaging,简称为DTI)是利用组织中水分子自由热运动的各向异性原理,探测组织的微观结构,以达到研究人体功能的目的。
目前,DTI是唯一一种可以活体显示脑白质纤维束的无创性成像方法。然而,DTI的信噪比比较低,这在很大程度上限制了DTI在临床上的广泛应用。
磁共振弥散张量成像的物理机制可以表示为:
d=Fρ+n (公式1)
其中,d表示在磁共振仪上所采集的信号,F表示博立叶编码矩阵,ρ表示弥散加权图像,n通常假定为复高斯白噪声。
在弥散张量成像中,第m个弥散加权图像ρm可以表示为:
其中,I0∈RN×1表示无弥散加权的参考图像,表示第m个弥散加权图像的相位,b表示弥散加权因子(常量),gm表示第m个弥散加权图像所对应的弥散梯度向量gm=(gxm,gym,gzm)T,D表示弥散张量,具体的D可以表示为:
其中,每个Di∈RN×1,N表示磁共振图像像素点的个数。
具体的,弥散张量成像的过程,主要包括以下步骤:
步骤1:由采集信号d重建图像ρ;
步骤2:根据模型张量(即上述公式2)估计每一空间位置所对应的张量矩阵D;
步骤3:根据下述公式4和公式5,由张量矩阵D计算得到各种弥散参数,例如:MD、FA等:
MD=(λ1+λ2+λ3)/3 (公式4)
其中,λ1,λ2,λ3为矩阵D的特征值。
为了提高弥散张量成像的信噪比,目前较为直接的方法是通过多次采样取平均或者是减小K空间的采样区域。虽然这些方法在实际应用中可以提高信噪比,但是这些方法会增加扫描时间或者影响空间分辨率。
目前,还有一种常用的提高弥散张量成像质量的方法。该方法在获取K空间的扫描数据后,首先重建出弥散加权图像。然后运用信号处理的方法对图像进行去噪。最后,通过去噪后的图像计算弥散张量及各种弥散参数。即,在处理的过程中需要先重建图像,然后对图像进行去噪。可是在图像去噪过程中产生的***误差往往会进一步传递到后续的张量计算中,从而影响弥散张量的精度。
针对如何在不增加扫描时间和不影响空间分辨率的同时,实现弥散张量的高精度去噪的技术问题,目前尚未提出有效的解决方案。
发明内容
本发明实施例提供了一种磁共振弥散张量去噪方法,以解决现有技术中无法在不增加扫描时间和不影响空间分辨率的同时,实现弥散张量的高精度去噪的技术问题,该方法包括:
采集K空间数据;
根据采集的K空间数据,求取弥散张量的最大似然估计值;
利用弥散张量和弥散参数的稀疏性,以求取的最大似然估计值作为初始值,计算弥散张量的最大后验概率估计值;
根据求取的最大后验概率估计值,计算弥散系数。
在一个实施例中,根据采集的K空间数据,求取弥散张量的最大似然估计值,包括:
按照以下公式计算弥散张量的最大似然估计值:
其中,表示弥散张量的最大似然估计值,dm表示第m个弥散加权图像所对应的K空间数据,F表示博立叶编码矩阵,I0表示无弥散加权的参考图像,表示第m个弥散加权图像的相位,b表示弥散加权因子,gm=(gxm,gym,gzm)T表示第m个弥散梯度向量。
在一个实施例中,利用弥散张量和弥散参数的稀疏性,将求取的最大似然估计值作为初始值,计算弥散张量的最大后验概率估计值,包括:
按照以下公式计算弥散张量的最大后验概率估计值:
在一个实施例中,R(·)约束每个弥散张量的稀疏性,或者R(·)约束弥散张量的联合稀疏性。
其中,Ψ为运算符,表示稀疏变换,ψi表示Ψ的第i列,||·||1表示求L1范数,||·||2表示求L2范数。
本发明实施例还提供了一种磁共振弥散张量去噪装置,以解决现有技术中无法在不增加扫描时间和不影响空间分辨率的同时,实现弥散张量的高精度去噪的技术问题,该装置包括:
存储器,用于存储指令;
处理器,其与所述存储器耦合,该处理器被设置为执行存储在所述存储器中的指令,其中,所述处理器被设置为用于:
采集K空间数据;
根据采集的K空间数据,求取弥散张量的最大似然估计值;
利用弥散张量和弥散参数的稀疏性,将求取的最大似然估计值作为初始值,计算弥散张量的最大后验概率估计值;
根据求取的所述最大后验概率估计值,计算弥散系数。
在一个实施例中,所述处理器具体用于按照以下公式计算弥散张量的最大似然估计值:
其中,表示弥散张量的最大似然估计值,dm表示第m个弥散加权图像所对应的K空间数据,F表示博立叶编码矩阵,I0表示无弥散加权的参考图像,表示第m个弥散加权图像的相位,b表示弥散加权因子,gm=(gxm,gym,gzm)T表示第m个梯度向量。
在一个实施例中,所述处理器具体用于按照以下公式计算弥散张量的最大后验概率估计值:
在一个实施例中,R(·)约束每个弥散张量的稀疏性,或者R(·)约束弥散张量的联合稀疏性。
其中,Ψ为运算符,表示稀疏变换,ψi表示Ψ的第i列,||·||1表示求L1范数,||·||2表示求L2范数。
本发明实施例还提供了一种用于磁共振弥散张量去噪的计算机程序产品,包括:
计算机可读存储介质,其具有在所述介质中实现的计算机可读程序代码,所述计算机可读程序代码包括:
被配置为采集K空间数据的计算机可读程序代码;
被配置为根据采集的K空间数据,求取弥散张量的最大似然估计值的计算机可读程序代码;
被配置为利用弥散张量和弥散参数的稀疏性,将求取的最大似然估计值作为初始值,计算弥散张量的最大后验概率估计值的计算机可读程序代码;以及
被配置为根据求取的所述最大后验概率估计值,计算弥散系数的计算机可读程序代码。
在本发明实施例中,在采集到K空间数据后,直接根据K空间数据确定弥散张量的最大似然估计值,然后依据求取的最大似然估计值,结合弥散张量、弥散参数的稀疏性计算弥散张量的最大后验概率估计值,从而确定去噪后的弥散张量。这有别于直接对弥散加权图像进行去噪的方法,避免了因为图像去噪而引入的***误差,解决了现有技术中无法在不增加扫描时间和不影响空间分辨率的同时,实现弥散张量的高精度去噪的技术问题,有效地抑制了弥散张量中的噪声。进一步的,因为通过弥散张量、弥散参数的稀疏性来进行去噪,从而更有效地提高了弥散张量的估计精度。
为了更清楚地说明本发明实施例中的技术方案,下面将对实施例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的附图。在附图中:
图1是根据本发明实施例的磁共振弥散张量去噪方法的方法流程图;
图2是根据本发明实施例的磁共振弥散张量去噪装置的结构框图;
图3是根据本发明实施例的计算机程序产品的示意性框图。
为使本发明的目的、技术方案和优点更加清楚明白,下面结合实施方式和附图,对本发明做进一步详细说明。在此,本发明的示意性实施方式及其说明用于解释本发明,但并不作为对本发明的限定。
发明人考虑到,在现有的弥散张量成像过程中,需要先得到弥散加权图像,再对弥散加权图像进行去噪。然而这样会引入***误差,从而导致最终的弥散张量估计出现误差。因此,在本例中提出了一种磁共振弥散张量去噪方法。该方法直接通过获取的K空
间数据对弥散张量进行去噪,无需求取弥散加权图像,也不需要对弥散加权图像进行去噪。如图1所示,该方法包括以下步骤:
步骤101:采集K空间数据;
步骤102:根据采集的K空间数据,求取弥散张量的最大似然估计值;
步骤103:利用弥散张量和弥散参数的稀疏性,以求取的最大似然估计值作为初始值,计算弥散张量的最大后验概率估计值;
步骤104:根据求取的最大后验概率估计值,计算弥散系数,其中,弥散系数可以包括:MD、FA。
在上述实施例中,在采集到K空间数据后,直接根据K空间数据确定弥散张量的最大似然估计值。然后以求取的最大似然估计值作为初值,结合弥散张量、弥散参数的稀疏性计算弥散张量的最大后验概率估计值,进而确定去噪后的弥散张量。这不同于对弥散加权图像进行去噪的方法,避免了因为图像去噪而引入的***误差,解决了现有技术中无法在不增加扫描时间和不影响空间分辨率的同时,实现弥散张量的高精度去噪的技术问题,有效抑制了弥散张量中的噪声。进一步的,因为通过弥散张量、弥散参数的稀疏性来进行去噪,从而更有效地提高了弥散张量的估计精度。
具体的,假设采样噪声服从高斯分布,结合上述公式1和公式2,弥散张量的最大似然估计值可以写作:
其中,表示弥散张量的最大似然估计值,dm表示第m个弥散加权图像所对应的K空间数据,F表示博立叶编码矩阵,I0表示无弥散加权的参考图像,表示第m个弥散加权图像的相位,b表示弥散加权因子,gm=(gxm,gym,gzm)T表示第m个弥散加权图像所对应的弥散梯度向量。
可以按照以下公式计算弥散张量的最大后验概率估计值:
因此,上述公式可以表示为:
其中,R(·)表示作用于弥散张量D的惩罚函数,G(·)表示作用于弥散参数的惩罚函数,γ1、γ2表示正则化参数、D1,D2,D3,D4,D5,D6代表不同方向的6个弥散张量。假定D和FA都是稀疏的,那么可以令:
(D1,D2,D3,D4,D5,D6)和弥散参数FA的稀疏性,
或者,令:
其中,运算符Ψ表示某种稀疏变换(例如:小波变换等),ψi表示Ψ的第i列,||·||1表示求L1范数,用于约束弥散张量在稀疏变换域内的稀疏性,||·||2表示求L2范数。
这样,弥散张量的去噪问题等价于上述的最优化求解问题。具体的可以采用非线性共轭梯度法进行求解。
即,在本例中,利用弥散张量模型和弥散张量、弥散参数FA的固有特性(如稀疏性),直接通过获取的K空间数据对弥散张量进行去噪,该方法略过了传统的图像去噪的步骤,避免了图像去噪的误差对张量估计的影响,可以更有效地抑制弥散张量中的噪声,提高弥散张量的估计精度。
基于同一发明构思,本发明实施例中还提供了一种磁共振弥散张量去噪装置,如下面的实施例所述。由于磁共振弥散张量去噪装置解决问题的原理与磁共振弥散张量去噪方法相似,因此磁共振弥散张量去噪装置的实施可以参见磁共振弥散张量去噪方法的实施,重复之处不再赘述。以下所使用的,术语“单元”或者“模块”可以实现预定功能的软件和/或硬件的组合。尽管以下实施例所描述的装置较佳地以软件来实现,但是硬件,或者软件和硬件的组合的实现也是可能并被构想的。图2是本发明实施例的磁共振弥散张量去噪装置200的一种结构框图,如图2所示,包括:存储器201,用于存储指令;处理器202,与存储器201耦合,处理器202被设置为执行存储在存储器201中的指令,其中,处理器202被设置为用于:采集K空间数据;根据采集的K空间数据,求取弥散张量的最大似然估计值;利用弥散张量和弥散参数的稀疏性,将求取的最大似然
估计值作为初始值,计算弥散张量的最大后验概率估计值;根据求取的所述最大后验概率估计值,计算弥散系数。
在一个实施例中,处理器202具体用于按照以下公式计算弥散张量的最大似然估计值:
其中,表示弥散张量的最大似然估计值,dm表示第m个弥散加权图像所对应的K空间数据,F表示博立叶编码矩阵,I0表示无弥散加权的参考图像,表示第m个弥散加权图像的相位,b表示弥散加权因子,gm=(gxm,gym,gzm)T表示第m个梯度向量。
在一个实施例中,处理器202具体用于按照以下公式计算弥散张量的最大后验概率估计值:
在一个实施例中,R(·)约束每个弥散张量的稀疏性,或者R(·)约束弥散张量的联合稀疏性。
其中,Ψ为运算符,表示稀疏变换,ψi表示Ψ的第i列,||·||1表示求L1范数,||·||2表示求L2范数。
图3是示出示例性计算机程序产品300的框图。在一些示例中,如图3中所示,计算机程序产品300包括计算机可读存储介质3001,在该计算机可读存储介质3001上存储有具有在所述介质中实现的计算机可读程序代码30011,该计算机可读程序代码30011包括:被配置为采集K空间数据的计算机可读程序代码,被配置为根据采集的K空间数据,求取弥散张量的最大似然估计值的计算机可读程序代码,被配置为利用弥散张量和弥散参数的稀疏性,将求取的最大似然估计值作为初始值,计算弥散张量的最大后验概
率估计值的计算机可读程序代码;以及被配置为根据求取的所述最大后验概率估计值,计算弥散系数的计算机可读程序代码。
同样在图3中示出了,在一些示例中,计算机程序产品300可以包括可记录介质3002和通信介质3003中的一种或多种。计算机可读存储介质3001和可记录介质3002可以包括(但不限于)软盘、硬盘驱动器(HDD)、压缩盘(CD)、数字通用盘(DVD)、数字带、计算机存储器等等。通信介质3003可以包括(但不限于)数字和/或模拟通信介质(例如,光纤线缆、波导、有线通信链路、无线通信链路等等)。
从以上的描述中,可以看出,本发明实施例实现了如下技术效果:在采集到K空间数据后,直接根据K空间数据确定弥散张量的最大似然估计值,然后依据求取的最大似然估计值,结合弥散张量、弥散参数的稀疏性计算弥散张量的最大后验概率估计值,从而确定去噪后的弥散张量。这有别于直接对弥散加权图像进行去噪的方法,避免了因为图像去噪而引入的***误差,解决了现有技术中无法在不增加扫描时间和不影响空间分辨率的同时,实现弥散张量的高精度去噪的技术问题,有效抑制了弥散张量中的噪声。进一步的,因为通过弥散张量、弥散参数的稀疏性来进行去噪,从而更有效地提高了弥散张量的估计精度。
显然,本领域的技术人员应该明白,上述的本发明实施例的各模块或各步骤可以用通用的计算装置来实现,它们可以集中在单个的计算装置上,或者分布在多个计算装置所组成的网络上,可选地,它们可以用计算装置可执行的程序代码来实现,从而,可以将它们存储在存储装置中由计算装置来执行,并且在某些情况下,可以以不同于此处的顺序执行所示出或描述的步骤,或者将它们分别制作成各个集成电路模块,或者将它们中的多个模块或步骤制作成单个集成电路模块来实现。这样,本发明实施例不限制于任何特定的硬件和软件结合。
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,本发明实施例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。
Claims (10)
- 一种磁共振弥散张量去噪方法,包括:采集K空间数据;根据采集的K空间数据,求取弥散张量的最大似然估计值;利用弥散张量和弥散参数的稀疏性,将求取的最大似然估计值作为初始值,计算弥散张量的最大后验概率估计值;根据求取的所述最大后验概率估计值,计算弥散系数。
- 如权利要求3所述的方法,其中,R(·)约束每个弥散张量的稀疏性,或者R(·)约束弥散张量的联合稀疏性。
- 一种磁共振弥散张量去噪装置,包括:存储器,用于存储指令;处理器,其与所述存储器耦合,该处理器被设置为执行存储在所述存储器中的指令,其中,所述处理器被设置为用于:采集K空间数据;根据采集的K空间数据,求取弥散张量的最大似然估计值;利用弥散张量和弥散参数的稀疏性,将求取的最大似然估计值作为初始值,计算弥散张量的最大后验概率估计值;根据求取的所述最大后验概率估计值,计算弥散系数。
- 如权利要求8所述的装置,其中,R(·)约束每个弥散张量的稀疏性,或者R(·)约束弥散张量的联合稀疏性。
- 一种用于磁共振弥散张量去噪的计算机程序产品,包括:计算机可读存储介质,其具有在所述介质中实现的计算机可读程序代码,所述计算机可读程序代码包括:被配置为采集K空间数据的计算机可读程序代码;被配置为根据采集的K空间数据,求取弥散张量的最大似然估计值的计算机可读程序代码;被配置为利用弥散张量和弥散参数的稀疏性,将求取的最大似然估计值作为初始值,计算弥散张量的最大后验概率估计值的计算机可读程序代码;以及被配置为根据求取的所述最大后验概率估计值,计算弥散系数的计算机可读程序代码。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US15/618,979 US10620287B2 (en) | 2014-12-11 | 2017-06-09 | Method and apparatus for denoising magnetic resonance diffusion tensor, and computer program product |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410758290.6 | 2014-12-11 | ||
CN201410758290.6A CN104504657B (zh) | 2014-12-11 | 2014-12-11 | 磁共振弥散张量去噪方法和装置 |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US15/618,979 Continuation US10620287B2 (en) | 2014-12-11 | 2017-06-09 | Method and apparatus for denoising magnetic resonance diffusion tensor, and computer program product |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2016090686A1 true WO2016090686A1 (zh) | 2016-06-16 |
Family
ID=52946051
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2014/095362 WO2016090686A1 (zh) | 2014-12-11 | 2014-12-29 | 磁共振弥散张量去噪方法和装置以及计算机程序产品 |
Country Status (3)
Country | Link |
---|---|
US (1) | US10620287B2 (zh) |
CN (1) | CN104504657B (zh) |
WO (1) | WO2016090686A1 (zh) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11079456B2 (en) * | 2018-05-10 | 2021-08-03 | Siemens Healthcare Gmbh | Method of reconstructing magnetic resonance image data |
CN113934766A (zh) * | 2021-10-11 | 2022-01-14 | 网易有道信息技术(江苏)有限公司 | 围棋定式对弈方法、装置、电子设备及存储介质 |
CN118261822A (zh) * | 2024-05-30 | 2024-06-28 | 贵州大学 | 基于体素替换的自监督磁共振扩散加权图像去噪方法 |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104545916B (zh) * | 2014-12-23 | 2017-09-22 | 中国科学院深圳先进技术研究院 | 磁共振弥散张量去噪方法和*** |
CN104586394B (zh) * | 2014-12-23 | 2017-04-12 | 中国科学院深圳先进技术研究院 | 去除磁共振弥散张量成像噪声的方法和*** |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2001085230A2 (en) * | 2000-05-08 | 2001-11-15 | Image-Guided Neurologics, Inc. | A method and apparatus for targeting material delivery to tissue |
US20080109171A1 (en) * | 2006-11-02 | 2008-05-08 | Siemens Corporate Research, Inc. | System and method for stochastic dt-mri connectivity mapping on the gpu |
CN102008307A (zh) * | 2010-12-29 | 2011-04-13 | 中国科学院深圳先进技术研究院 | 磁共振弥散张量成像方法和*** |
CN103705239A (zh) * | 2013-12-05 | 2014-04-09 | 深圳先进技术研究院 | 磁共振参数成像方法和*** |
CN103985099A (zh) * | 2014-05-30 | 2014-08-13 | 成都信息工程学院 | 一种弥散张量磁共振图像张量域非局部均值去噪方法 |
Family Cites Families (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1305008C (zh) * | 2003-12-22 | 2007-03-14 | 中国科学院自动化研究所 | 弥散张量核磁成像脑缺血病灶区的自动分割方法 |
US7630530B2 (en) * | 2005-01-14 | 2009-12-08 | Siemens Medical Solutions Usa, Inc. | System and method for fast tensor field segmentation |
US7355403B2 (en) * | 2005-01-27 | 2008-04-08 | Siemens Medical Solutions Usa, Inc. | Noise reduction in diffusion tensor imaging data using bayesian methods |
GB0521640D0 (en) * | 2005-10-24 | 2005-11-30 | Ccbr As | Automatic quantification of a pathology indicating measure from cartilage scan data |
US7889899B2 (en) * | 2006-05-05 | 2011-02-15 | Siemens Medical Solutions Usa, Inc. | Directional statistics visualization tool |
WO2008085473A1 (en) * | 2006-12-27 | 2008-07-17 | The Johns Hopkins University | Mri methods using diffusion tensor imaging techniques and mri systems embodying same |
US7994784B2 (en) * | 2007-07-20 | 2011-08-09 | Medical College Of Georgia Research Institute, Inc. | Systems and methods for rescaling image intensities with diffusion MRI to synthesize complex diffusive geometries |
US9494669B2 (en) * | 2010-05-17 | 2016-11-15 | Washington University | Diagnosis of central nervous system white matter pathology using diffusion MRI |
CN102309328B (zh) * | 2011-10-19 | 2012-11-14 | 中国科学院深圳先进技术研究院 | 弥散张量成像方法及*** |
US9880246B2 (en) * | 2013-01-28 | 2018-01-30 | University Of Southern California | Linear transform for diffusion MRI |
US9234952B2 (en) * | 2013-03-11 | 2016-01-12 | Wisconsin Alumni Research Foundation | System and method for determining k-space views and diffusion weighting directions using centroidal voronoi tessellations |
JP5946800B2 (ja) * | 2013-07-22 | 2016-07-06 | 株式会社日立製作所 | 磁気共鳴イメージング装置、画像処理装置、画像処理方法、及び画像処理プログラム |
-
2014
- 2014-12-11 CN CN201410758290.6A patent/CN104504657B/zh active Active
- 2014-12-29 WO PCT/CN2014/095362 patent/WO2016090686A1/zh active Application Filing
-
2017
- 2017-06-09 US US15/618,979 patent/US10620287B2/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2001085230A2 (en) * | 2000-05-08 | 2001-11-15 | Image-Guided Neurologics, Inc. | A method and apparatus for targeting material delivery to tissue |
US20080109171A1 (en) * | 2006-11-02 | 2008-05-08 | Siemens Corporate Research, Inc. | System and method for stochastic dt-mri connectivity mapping on the gpu |
CN102008307A (zh) * | 2010-12-29 | 2011-04-13 | 中国科学院深圳先进技术研究院 | 磁共振弥散张量成像方法和*** |
CN103705239A (zh) * | 2013-12-05 | 2014-04-09 | 深圳先进技术研究院 | 磁共振参数成像方法和*** |
CN103985099A (zh) * | 2014-05-30 | 2014-08-13 | 成都信息工程学院 | 一种弥散张量磁共振图像张量域非局部均值去噪方法 |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11079456B2 (en) * | 2018-05-10 | 2021-08-03 | Siemens Healthcare Gmbh | Method of reconstructing magnetic resonance image data |
CN113934766A (zh) * | 2021-10-11 | 2022-01-14 | 网易有道信息技术(江苏)有限公司 | 围棋定式对弈方法、装置、电子设备及存储介质 |
CN113934766B (zh) * | 2021-10-11 | 2023-04-14 | 网易有道信息技术(江苏)有限公司 | 围棋定式对弈方法、装置、电子设备及存储介质 |
CN118261822A (zh) * | 2024-05-30 | 2024-06-28 | 贵州大学 | 基于体素替换的自监督磁共振扩散加权图像去噪方法 |
Also Published As
Publication number | Publication date |
---|---|
CN104504657B (zh) | 2017-05-03 |
US20170363702A1 (en) | 2017-12-21 |
US10620287B2 (en) | 2020-04-14 |
CN104504657A (zh) | 2015-04-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Johnson et al. | Conditional generative adversarial network for 3D rigid‐body motion correction in MRI | |
US11324418B2 (en) | Multi-coil magnetic resonance imaging using deep learning | |
Liebling et al. | Four-dimensional cardiac imaging in living embryos<? xpp qa?> via postacquisition synchronization of nongated<? xpp qa?> slice sequences | |
US10671939B2 (en) | System, method and computer-accessible medium for learning an optimized variational network for medical image reconstruction | |
WO2016090686A1 (zh) | 磁共振弥散张量去噪方法和装置以及计算机程序产品 | |
CN104574298B (zh) | 一种基于互信息的多b值扩散权重图像的降噪方法 | |
WO2020114329A1 (zh) | 磁共振快速参数成像方法及装置 | |
CN109872376A (zh) | 一种重建动态磁共振图像的方法、装置及可读存储介质 | |
US10605882B2 (en) | Systems and methods for removing background phase variations in diffusion-weighted magnetic resonance imaging | |
CN112790756A (zh) | 任务态功能磁共振扫描数据去噪方法、装置、设备及介质 | |
WO2020198456A1 (en) | Retrospective motion correction using a combined neural network and model-based image reconstruction of magnetic resonance data | |
CN104599244B (zh) | 磁共振弥散张量成像的去噪方法和*** | |
CN112581385B (zh) | 基于多先验约束扩散峰度成像张量估计方法、介质和设备 | |
US20230380714A1 (en) | Method and system for low-field mri denoising with a deep complex-valued convolutional neural network | |
CN104545916B (zh) | 磁共振弥散张量去噪方法和*** | |
CN104586394B (zh) | 去除磁共振弥散张量成像噪声的方法和*** | |
Min et al. | Grid-free localization algorithm using low-rank Hankel matrix for super-resolution microscopy | |
US20220413078A1 (en) | Model-Based Nyquist Ghost Correction for Reverse Readout Echo Planar Imaging | |
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 | |
CN104599245B (zh) | 磁共振弥散张量去噪方法和*** | |
EP4096507B1 (en) | Systems, methods, and media for estimating a mechanical property based on a transformation of magnetic resonance elastography data using a trained artificial neural network | |
KR102215902B1 (ko) | 영상 보정 방법 및 이를 수행하는 장치들 | |
Ye | Fiber orientation estimation using nonlocal and local information | |
Garyfallidis et al. | Constrained spherical deconvolution on signal and ODF values | |
CN114862984A (zh) | 基于锐度增强的磁共振图像重建方法、***、介质及设备 |
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: 14907917 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 13.10.2017) |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 14907917 Country of ref document: EP Kind code of ref document: A1 |