CN101915901A - Magnetic resonance imaging method and device - Google Patents

Magnetic resonance imaging method and device Download PDF

Info

Publication number
CN101915901A
CN101915901A CN 201010255677 CN201010255677A CN101915901A CN 101915901 A CN101915901 A CN 101915901A CN 201010255677 CN201010255677 CN 201010255677 CN 201010255677 A CN201010255677 A CN 201010255677A CN 101915901 A CN101915901 A CN 101915901A
Authority
CN
China
Prior art keywords
image
mask
magnetic resonance
field
composite
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.)
Pending
Application number
CN 201010255677
Other languages
Chinese (zh)
Inventor
翁卓
谢国喜
邹超
刘新
邱本胜
熊承义
郑海荣
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenzhen Institute of Advanced Technology of CAS
Original Assignee
Shenzhen Institute of Advanced Technology of CAS
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 Shenzhen Institute of Advanced Technology of CAS filed Critical Shenzhen Institute of Advanced Technology of CAS
Priority to CN 201010255677 priority Critical patent/CN101915901A/en
Publication of CN101915901A publication Critical patent/CN101915901A/en
Pending legal-status Critical Current

Links

Images

Landscapes

  • Magnetic Resonance Imaging Apparatus (AREA)

Abstract

一种磁共振成像方法，包括如下步骤：对视野校准扫描得到多个复合图像，对多个复合图像进行绝对值求和得到求和图像，并根据所述多个复合图像及求和图像计算得到灵敏度系数；校正所述灵敏度系数，并生成重建图像。上述磁共振成像方法及装置对复合图像的绝对值进行求和得到求和图像，并通过该求和图像计算得到灵敏度系数，从而有效地提高了灵敏度系数的均匀性，灵敏度系数校正后生成重建图像，有效消除了扫描过程中运动等因素对图像造成的伪影，提高了重建后生成的无伪影复合图像的分辨率，加强了重建图像的鲁棒性。

A magnetic resonance imaging method, comprising the steps of: calibrating and scanning the field of view to obtain a plurality of composite images, summing the absolute values of the plurality of composite images to obtain a sum image, and calculating and obtaining a sensitivity coefficient; correcting the sensitivity coefficient, and generating a reconstructed image. The above magnetic resonance imaging method and device sum the absolute values of the composite images to obtain a summed image, and calculate the sensitivity coefficient through the summed image, thereby effectively improving the uniformity of the sensitivity coefficient, and generating a reconstructed image after the sensitivity coefficient is corrected , which effectively eliminates the artifacts caused by factors such as motion during the scanning process, improves the resolution of the artifact-free composite image generated after reconstruction, and strengthens the robustness of the reconstructed image.

Description

MR imaging method and device
[technical field]
The present invention relates to mr imaging technique, particularly relate to a kind of MR imaging method and device.
[background technology]
Magnetic resonance imaging has higher soft tissue contrast and spatial resolution, can select imaging parameters and imaging aspect as required flexibly, be widely used in clinical in.Traditional magnetic resonance imaging usually occurs owing to data acquisition time causes the slower problem of image taking speed than length.This is because the gradient field intensity has seriously restricted the sweep velocity of magnetic resonance imaging near the limit.Therefore, adopted the multi-channel parallel imaging technique in traditional magnetic resonance imaging, promptly multichannel collecting and parallel imaging algorithm make magnetic resonance imaging can be no longer dependent on the also collection of expedited data greatly of raising of gradient field intensity.The multi-channel parallel imaging technique is to utilize the phased-array coil spatial information to replace the gradient coded message, and each coil unit is owed sampling to the K space simultaneously, and the K spacing wave quantity that makes each coil unit gather significantly reduces, and image taking speed improves greatly.This formation method had both improved the speed of imaging, can improve the contrast of image again, therefore had wide practical value in the very high inspection of imaging requirements such as heart, dynamically enhancing, blood vessel imaging.
But, in the multi-channel parallel imaging technique,, each coil gathers the K spatial data, and according to nyquist sampling theorem because all owing sampling, and the data that each coil unit is gathered are directly carried out image reconstruction aliasing can be taken place, and cause overlapping pseudo-shadow.Overlapping pseudo-shadow can pass through sensitivity coefficient, and (Sensitivity Encoding, SENSE) technology is removed.
Yet; in traditional sensitivity encoding technology; it is too poor the sensitivity coefficient homogeneity to occur; the defective that the sensitivity map air spots that obtains according to sensitivity coefficient is sliding; and in the process of data acquisition; because it is unusual that motion usually can make data take place, thereby image is produced very large pseudo-shadow, and picture quality is produced significant impact.
[summary of the invention]
Based on this, be necessary to provide a kind of inhomogeneity MR imaging method of sensitivity that improves.
In addition, also be necessary to provide a kind of inhomogeneity MR imaging apparatus of sensitivity that improves.
A kind of MR imaging method comprises the steps: visual field calibration scan is obtained a plurality of combination pictures, a plurality of combination pictures is carried out the absolute value summation obtain the image of suing for peace, and obtain sensitivity coefficient according to described a plurality of combination pictures and summation image calculation; Proofread and correct described sensitivity coefficient, and generate reconstructed image.
Preferably, state and proofread and correct described sensitivity coefficient, and also comprise the step of proofreading and correct described combination picture deviation field before the step of generation reconstructed image.
Preferably, the step of the deviation field in the described combination picture of described correction is: to the combination picture pre-service, obtain mask; Described mask is acted in the combination picture extracting tissue part's image, and described tissue part image is carried out log-transformation, obtain adding the sexual deviation field; According to adding sexual deviation field and mask, obtain according to a preliminary estimate deviation field by gauss low frequency filter and normalization convolution; The deviation field that the B spline-fitting is accurately estimated is carried out in deviation field according to a preliminary estimate, and the combination picture that obtains proofreading and correct of the ratio by the gradation of image that measures and the deviation field of described accurate estimation.
Preferably, described to the combination picture pre-service, the step that obtains mask is: measure gray level image in combination picture, obtain the brightness maximal value of described gray level image mid point by described gray level image, setting threshold is the number percent of high-high brightness, is lower than threshold value and partly is made as 0, and tissue part is made as 1, form bianry image, obtain mask.
Preferably, the described sensitivity coefficient of described correction, and the step that generates reconstructed image is: the match by reference picture and the combination picture that obtains in the scanning of the visual field obtains residual error; Introduce annealing parameter, described residual error is imported in the strength of joint function, obtain the value of described strength of joint function, the value of described strength of joint function is as the element on the diagonal matrix diagonal line; Obtain not having pseudo-shadow combination picture by complex image, diagonal matrix and sensitivity coefficient.
A kind of MR imaging apparatus comprises at least: the phased array coil is used for visual field scanning is obtained a plurality of combination pictures and corresponding complex image; Calculation element is used for obtaining the image of suing for peace according to described a plurality of combination pictures, and obtains sensitivity coefficient according to described a plurality of combination pictures and summation image calculation; Equipment for reconstructing image is used to proofread and correct described sensitivity coefficient, and generates reconstructed image.
Preferably, also comprise: luminance correction device, the deviation field that is used for proofreading and correct described combination picture.
Preferably, described luminance correction device comprises at least: the mask process module is used for the combination picture pre-service is obtained mask; Organize conversion module, be used for described mask is acted on combination picture with extraction tissue part image, and described tissue part image is carried out log-transformation, obtain adding the sexual deviation field; The precorrection module is used for according to adding sexual deviation field and mask, obtains according to a preliminary estimate deviation field by gauss low frequency filter and normalization convolution; Level and smooth module is used for the deviation field that the level and smooth match of B batten is accurately estimated is carried out in deviation field according to a preliminary estimate, and the combination picture that obtains proofreading and correct of the ratio by the gradation of image that measures and the deviation field of described accurate estimation.
Preferably, described mask process module also is used for measuring gray level image at combination picture, obtain the brightness maximal value of described gray level image mid point by described gray level image, setting threshold is the number percent of high-high brightness, be lower than threshold value and partly be made as 0, tissue part is made as 1, forms bianry image, to obtain mask.
Preferably, described equipment for reconstructing image comprises at least: fitting module is used for obtaining residual error by the match of reference picture and described combination picture; Estimation module is used to introduce annealing parameter, and described residual error is imported in the strength of joint function, obtains the value of described strength of joint function, and the value of described strength of joint function is as the element on the diagonal line in the diagonal matrix; The image generation module is used for obtaining not having pseudo-shadow combination picture by complex image, diagonal matrix and sensitivity coefficient.Above-mentioned MR imaging method and device are sued for peace to the absolute value of combination picture and are obtained the image of suing for peace, and obtain sensitivity coefficient by this summation image calculation, thereby improved the homogeneity of sensitivity coefficient effectively, sensitivity coefficient is proofreaied and correct the back and is generated reconstructed image, effectively eliminated the pseudo-shadow that factor such as move in the scanning process causes image, improve the resolution of the pseudo-shadow combination picture of nothing of rebuilding the back generation, strengthened the robustness of reconstructed image.
Above-mentioned MR imaging method and device had at first been proofreaied and correct the deviation field in the combination picture before corrected sensitivity coefficient, thereby obtained the even brightness image, and had improved signal to noise ratio (S/N ratio), improved the accuracy that the observer analyzes visual image.
Above-mentioned MR imaging method and device gauss low frequency filter and normalization involve in capable filtering, have obtained deviation field according to a preliminary estimate, have removed the artefact between background area and the tissue image.
[description of drawings]
Fig. 1 is the process flow diagram of MR imaging method among the embodiment;
Fig. 2 is the process flow diagram of the deviation field in the positive combination picture of an embodiment lieutenant colonel;
Fig. 3 A is the brain original image;
Fig. 3 B be with the corresponding correction of brain original image after image;
Fig. 4 A is the line chart of brain original image;
Fig. 4 B be with the corresponding correction of brain original image after the line chart of image;
Fig. 5 is a corrected sensitivity coefficient and generate the process flow diagram of reconstructed image among the embodiment;
Fig. 6 is the synoptic diagram that generates reconstructed image among the embodiment;
Fig. 7 is the detailed block diagram of MR imaging apparatus among the embodiment;
Fig. 8 A is the reference picture of phantom;
Fig. 8 B is the phantom image by the classic method gained;
Fig. 8 C is the phantom image that the present invention generated;
Fig. 9 A is the reference picture of brain;
Fig. 9 B is the brain image by the classic method gained;
Fig. 9 C is the brain image that the present invention generated;
Figure 10 A is the reference picture of belly;
Figure 10 B is the abdomen images by the classic method gained;
Figure 10 C is the abdomen images that the present invention generated.
[embodiment]
Fig. 1 shows the method flow of magnetic resonance imaging among the embodiment, comprises the steps:
In step S10, visual field calibration scan is obtained a plurality of combination pictures, a plurality of combination pictures are carried out the absolute value summation obtain the image of suing for peace, and obtain sensitivity coefficient according to a plurality of combination pictures and summation image calculation.Among one embodiment, in the K space, calibration scan sampling in the visual field is obtained combination picture f l(l=1,2 ..., L), by will be according to combination picture f lThe absolute value images addition that generates and obtain the image of suing for peace
Figure BSA00000232502100041
Thereby utilize following formula to calculate sensitivity coefficient C l:
C l＝f l/Am
Wherein, l by in a plurality of phased array coils a certain phased array coil of correspondence.
By combination picture f 1The absolute value images addition and obtain sensitivity coefficient C l, improved homogeneity effectively, make sensitivity coefficient C thus lThe sensitivity image that produces has level and smooth surface.
In step S20, proofread and correct the deviation field in the described combination picture.In one embodiment, the deviation field is also referred to as the property the taken advantage of field of low frequency, is meant image in same physiological tissue or structure, the phenomenon that brightness slowly changes.When image data,, can cause the luminance non of gained image owing to the difference of susceptibility coefficient.Brightness image heterogeneous provides false contrast, has seriously hindered the observer this image is carried out correlativity and analysis of the accuracy.Based on this, as shown in Figure 2, in one embodiment, the detailed process of step S20 is:
In step S201,, obtain mask to the combination picture pre-service.The pre-service of combination picture is a mask process, refers to selected image, figure or object pending image (whole or local) is blocked, with the processing procedure of control image-region.The selected digital image, figure or the object that are covered on the pending image are called mask.The process of mask process is specifically: at combination picture f 1In measure gray level image, generate brightness histogram, find out the brightness maximal value of pixel in the gray level image by brightness histogram, setting threshold is the number percent of high-high brightness then, the part that is lower than threshold value can be regarded background area or noise as, be set to 0, tissue part is made as 1, thereby obtains a width of cloth bianry image as mask t.
In step S202, mask is acted in the combination picture extracting tissue part's image, and tissue part's image is carried out log-transformation, obtain adding the sexual deviation field.Among one embodiment, mask t is blocked in combination picture f 1On, extract the image v ' of tissue part l, shown in the following formula, to the image v ' of tissue part lCarry out log-transformation, add the sexual deviation field taking advantage of the sexual deviation field to be converted into.
V l＝logv′ l＝logu′+logf′ l
Wherein, u ' is the true picture gray scale behind the mask, f ' lIt is the smooth variation deviation field behind the mask.
In step S203, according to adding sexual deviation field and mask, obtain according to a preliminary estimate deviation field by gauss low frequency filter and normalization convolution.Among one embodiment, carry out filtering, between background area and tissue image, produce in various degree artefact in meeting after the filtering, seriously influenced the quality of image by gauss low frequency filter.Therefore, also carry out filtering once more, remove the artefact between background area and the tissue image, obtain deviation field d according to a preliminary estimate by the method for normalization convolution l
d l＝exp(LPF[V l]/LPF[t l])
In step S204, the deviation field that the B spline-fitting is accurately estimated is carried out in deviation field according to a preliminary estimate, and the combination picture that obtains proofreading and correct of the ratio by the gradation of image that measures and the deviation field of accurate estimation.Among one embodiment, owing in the filtering of normalization convolution, be to think that the data of tissue image are known, the loss of data of background area to appearing at the not effect of artefact in the tissue image, is therefore carried out the B spline-fitting to deviation field according to a preliminary estimate.To carrying out the B spline-fitting and the process of the deviation field accurately estimated is in deviation field according to a preliminary estimate: in deviation field according to a preliminary estimate, select more smooth zone to carry out sub-sampling, the sample point of gained is as fitting nodes, use the resulting fitting nodes of B spline-fitting then, by interpolation, extrapolation image, obtain the deviation field f ' of level and smooth accurate estimation at last l, B spline-fitting function can be expressed as:
f l ( x , y ) = Σ i Σ j θ ij B i ( x ) B j ( y )
Wherein, B i, B jBe one dimension B-spline function, θ IjBe fitting parameter.
The B spline-fitting is the optimizing process that makes J=E (θ)+ω R (θ) minimum, and in the formula, E (θ) is consistent degree, reflected B spline-fitting function and raw data s (x, degree of closeness y), promptly
E ( θ ) = Σ x , y | | s ( x , y ) - f ( x , y ) | | 2
ω is a weighting factor of rule of thumb selecting; R (θ) is a smoothness, the smooth degree of reflection B-spline function, promptly
R ( θ ) = Σ x , y [ ( ∂ 2 f ∂ x ∂ x ) 2 + 2 ( ∂ 2 f ∂ x ∂ y ) 2 + ( ∂ 2 f ∂ y ∂ y ) 2 ]
According to the gradation of image v that measures lAnd the deviation field f ' that accurately estimates lThe combination picture that obtains proofreading and correct by following formula, promptly
u＝v l/f′ l
Its brightness nonuniformity correction of the combination picture result who proofreaies and correct is shown in Fig. 3 A to Fig. 3 B, and the line chart of the brain original image of Fig. 3 A is shown in Fig. 4 A, and Fig. 3 B proofreaies and correct the correction result of the 60th row in the image of back shown in Fig. 4 B.In Fig. 4 A to Fig. 4 B, to compare with the line chart of brain original image, the line chart of proofreading and correct the back image is more level and smooth, and burr is less.
In step S30, corrected sensitivity coefficient, and generate reconstructed image.In one embodiment, when gathering raw data, be positioned over the influence that the coil at some position of human body is very easy to be moved or suffer the destruction of noise and take place unusual when image data, the destruction data that contain pseudo-shadow become the exceptional value that raw data is concentrated, moved or the pseudo-image data of noise corrupted for counting, to sensitivity coefficient C lCarry out the AM Robust Estimation.As shown in Figure 5, in one embodiment, the process of step S30 specifically:
In step S301, the match by reference picture and the combination picture that obtains in visual field scanning obtains residual error.Be defined in the combination picture f that obtains in the scanning of the visual field lWith the difference of reference picture be residual error r lReference picture obtains fast by only scanning low-frequency data, and this reference picture resolution is higher.
In step S302, introduce annealing parameter, residual error is imported in the strength of joint function, obtain the value of strength of joint function, the value of strength of joint function is as the element on the diagonal matrix diagonal line.According to the discontinuous markov prior model of self-adaptation, the strength of joint function is
Figure BSA00000232502100071
Wherein t is the annealing parameter that successively decreases gradually in the iterative process, and its cooling strategy passes through t n=0.1 * t N-1Finish, the initial value of annealing parameter t is 2.Obtain the element on the diagonal line, i.e. D=diag (d among the diagonal matrix row D by the strength of joint function calculation 1, d 2..., d l), wherein l is that residual error is counted out.
In step S303, obtain not having pseudo-shadow combination picture by complex image, diagonal matrix and the sensitivity coefficient that in the scanning of the visual field, obtains.Carry out low coverage scanning and obtain complex image S lThereby, with complex image S l, diagonal matrix D and sensitivity coefficient C lImport in the following image reconstruction formula
ρ AM＝(C l HDC l) -1C l HDS l
Wherein, ρ AMBe the pseudo-shadow combination picture of the nothing that obtains after rebuilding, C HBe the special transposition of the Hull rice of Matrix C.
As shown in Figure 6, by phased array coil Coil_l (l=1,2 ... L) carry out quick Fu Ye (fast fourier transform, FFT) conversion obtain corresponding complex image S after the scanning lThereby, with complex image S lWith susceptibility coefficient C lAfter carrying out the AM Robust Estimation, obtain artifact-free combination picture.
Fig. 7 shows the MR imaging apparatus among the embodiment, and this MR imaging apparatus comprises phased array coil 20, calculation element 40, luminance correction device 60 and equipment for reconstructing image 80, wherein:
Phased array coil 20 is used for visual field scanning is obtained a plurality of combination pictures and corresponding complex image.Phased array coil 20 is owed sampling and is obtained combination picture f in the K space l, and the process Fast Fourier Transform (FFT) obtains complex image S l
Calculation element 40 is used for calculating the summation image according to a plurality of combination pictures, and obtains sensitivity coefficient according to a plurality of combination pictures and summation image calculation.Among one embodiment, by will be according to a plurality of combination picture f lThe absolute value images addition that generates and obtain the image of suing for peace
Figure BSA00000232502100072
And calculate sensitivity coefficient C according to following formula l:
C l＝f l/Am
Luminance correction device 60, the deviation field that is used for proofreading and correct combination picture.As previously mentioned, in one embodiment, luminance correction device 60 comprises:
Mask process module 601 is used for the combination picture pre-service is obtained mask.601 pairs of combination pictures of this mask process module carry out mask process, at combination picture f lIn measure gray level image, generate brightness histogram, find out the brightness maximal value of gray level image mid point by this brightness histogram, setting threshold is the number percent of high-high brightness then, the part that is lower than threshold value can be regarded background area or noise as, be set to 0, tissue part is made as 1, thereby obtains a width of cloth bianry image as mask t.
Organize conversion module 602, be used for mask is acted on combination picture with extraction tissue part image, and tissue part's image is carried out log-transformation, obtain adding the sexual deviation field.This tissue conversion module 602 blocks mask t in combination picture f lOn, extract the image v ' of tissue part l, by following formula with the image v ' of tissue part lCarry out log-transformation, add the sexual deviation field taking advantage of the sexual deviation field to be converted into, promptly
v l＝logv′ l＝logu′+logf′ l
Wherein, u ' is the true picture gray scale behind the mask, f ' lIt is the smooth variation deviation field behind the mask.
Precorrection module 603 is used for according to adding sexual deviation field and mask, obtains according to a preliminary estimate deviation field by gauss low frequency filter and normalization convolution.After this precorrection module 603 is carried out filtering by gauss low frequency filter, by the normalization convolution carry out filtering once more, obtain deviation field d according to a preliminary estimate l, promptly
d l＝exp(LPF[V l]/LPF[t l])
Level and smooth module 604 is used for the deviation field that the B spline-fitting is accurately estimated is carried out in deviation field according to a preliminary estimate, and the combination picture that obtains proofreading and correct of the ratio by the gradation of image that measures and the deviation field of accurate estimation.This level and smooth module 604 selects more smooth zone to carry out sub-sampling in deviation field according to a preliminary estimate, with resulting sample point as fitting nodes, use the resulting fitting nodes of B spline-fitting then,, obtain the deviation field f ' of level and smooth accurate estimation at last by interpolation, extrapolation image l, B spline-fitting function can be expressed as:
f l ( x , y ) = Σ i Σ j θ ij B i ( x ) B j ( y )
Wherein, B i, B jBe one dimension B-spline function, θ IjBe fitting parameter.
The B spline-fitting is the optimizing process that makes J=E (θ)+ω R (θ) minimum, and in the formula, E (θ) is consistent degree, reflected B spline-fitting function and raw data s (x, degree of closeness y), promptly
E ( θ ) = Σ x , y | | s ( x , y ) - f ( x , y ) | | 2
ω is a weighting factor of rule of thumb selecting; R (θ) is a smoothness, the smooth degree of reflection B-spline function, promptly
R ( θ ) = Σ x , y [ ( ∂ 2 f ∂ x ∂ x ) 2 + 2 ( ∂ 2 f ∂ x ∂ y ) 2 + ( ∂ 2 f ∂ y ∂ y ) 2 ]
According to the gradation of image v that measures lAnd the deviation field f ' that accurately estimates lThe combination picture that obtains proofreading and correct by following formula, promptly
u＝v l/f′ l
Equipment for reconstructing image 80 is used for corrected sensitivity coefficient, and generates reconstructed image.Among one embodiment, 80 couples of sensitivity coefficient C of equipment for reconstructing image lCarry out the AM Robust Estimation, comprising:
Fitting module 801 is used for obtaining residual error r by the match of reference picture and combination picture lReference picture obtains fast by scanning low frequency biogas, and resolution is higher.
Estimation module 802 is used to introduce annealing parameter, and residual error is imported in the strength of joint function, obtains the value of strength of joint function, and the value of strength of joint function is as the element on the diagonal line in the diagonal matrix.As previously mentioned, estimation module 802 is selected for use
Figure BSA00000232502100093
The strength of joint function, wherein t is the annealing parameter that successively decreases gradually in the iterative process, its cooling strategy pass through t n=0.1 * t N-1Finish, the initial value of annealing parameter t is 2.Obtain the element on the diagonal line, i.e. D=diag (d among the diagonal matrix row D by the strength of joint function calculation 1, d 2..., d l).
Image generation module 803 is used for obtaining not having pseudo-shadow combination picture by complex image, diagonal matrix and sensitivity coefficient.As previously mentioned, image generation module 803 is with complex image S l, diagonal matrix D and sensitivity coefficient C lImport in the following image reconstruction formula, to obtain not having pseudo-shadow combination picture
ρ AM＝(C l HDC l) -1C l HDS l
Wherein, ρ AMBe the pseudo-shadow combination picture of the nothing that obtains after rebuilding, C HBe the special transposition of the Hull rice of Matrix C.
Shown in Fig. 8 A to Fig. 8 C, by the comparison that resulting no pseudo-shadow combination picture was sought and installed in reference picture, classic method gained image and above-mentioned magnetic resonance imaging in phantom data reconstructed results, it is more clear not have pseudo-shadow combination picture as can be seen.
Shown in Fig. 9 A to Fig. 9 C, by in the reconstructed results of true brain data, resulting no pseudo-shadow combination picture is sought and is installed in contrast reference picture, classic method gained image and above-mentioned magnetic resonance imaging, in the indicated position of the arrow of classic method, it is clear and do not have a pseudo-shadow not have pseudo-shadow combination picture.
Shown in Figure 10 A to Figure 10 C, by in the reconstructed results of belly data, resulting no pseudo-shadow combination picture is sought and is installed in contrast reference picture, classic method gained image and above-mentioned magnetic resonance imaging, in the indicated position of the arrow of classic method, it is clear and do not have a pseudo-shadow not have pseudo-shadow combination picture.
Above-mentioned MR imaging method and device are sued for peace to the absolute value of combination picture and are obtained the image of suing for peace, and obtain sensitivity coefficient by this summation image calculation, thereby improved the homogeneity of sensitivity coefficient effectively, sensitivity coefficient is proofreaied and correct the back and is generated reconstructed image, effectively eliminated the pseudo-shadow that factor such as move in the scanning process causes image, improve the resolution of the pseudo-shadow combination picture of nothing of rebuilding the back generation, strengthened the robustness of image reconstruction.
Above-mentioned MR imaging method and device had at first been proofreaied and correct the deviation field in the combination picture before corrected sensitivity coefficient, thereby obtained the even brightness image, and had improved signal to noise ratio (S/N ratio), improved the accuracy that the observer analyzes visual image.
Above-mentioned MR imaging method and device gauss low frequency filter and normalization involve in capable filtering, have obtained deviation field according to a preliminary estimate, have removed the artefact between background area and the tissue image.
The above embodiment has only expressed several embodiment of the present invention, and it describes comparatively concrete and detailed, but can not therefore be interpreted as the restriction to claim of the present invention.Should be pointed out that for the person of ordinary skill of the art without departing from the inventive concept of the premise, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be as the criterion with claims.

Claims (10)

1.一种磁共振成像方法，包括如下步骤：1. A magnetic resonance imaging method, comprising the steps of: 对视野校准扫描得到多个复合图像，对多个复合图像进行绝对值求和得到求和图像，并根据所述多个复合图像及求和图像计算得到灵敏度系数；Scanning the field of view calibration to obtain multiple composite images, summing the absolute values of the multiple composite images to obtain a sum image, and calculating a sensitivity coefficient according to the multiple composite images and the sum image; 校正所述灵敏度系数，并生成重建图像。The sensitivity coefficients are corrected, and a reconstructed image is generated. 2.根据权利要求1所述的磁共振成像方法，其特征在于，所述校正所述灵敏度系数，并生成重建图像的步骤之前还包括校正所述复合图像偏差场的步骤。2. The magnetic resonance imaging method according to claim 1, characterized in that, before the step of correcting the sensitivity coefficient and generating the reconstructed image, the step of correcting the bias field of the composite image is also included. 3.根据权利要求2所述的磁共振成像方法，其特征在于，所述校正所述复合图像中的偏差场的步骤是：3. The magnetic resonance imaging method according to claim 2, wherein the step of correcting the bias field in the composite image is: 对复合图像预处理，得到掩模；Preprocessing the composite image to obtain a mask; 将所述掩模作用于复合图像中以抽取组织部分图像，并对所述组织部分图像进行对数变换，得到加性偏差场；applying the mask to the composite image to extract a tissue portion image, and logarithmically transforming the tissue portion image to obtain an additive bias field; 根据加性偏差场和掩模，通过高斯低通滤波器和归一化卷积得到初步估计的偏差场；According to the additive bias field and the mask, the bias field is initially estimated by Gaussian low-pass filter and normalized convolution; 对初步估计的偏差场进行B样条拟合得到精确估计的偏差场，并通过测量得到的图像灰度与所述精确估计的偏差场之比得到校正的复合图像。Perform B-spline fitting on the preliminarily estimated deviation field to obtain an accurately estimated deviation field, and obtain a corrected composite image by measuring the ratio of the grayscale of the image obtained to the accurately estimated deviation field. 4.根据权利要求3所述的磁共振成像方法，其特征在于，所述对复合图像预处理，得到掩膜的步骤是：4. magnetic resonance imaging method according to claim 3, is characterized in that, described compound image preprocessing, the step of obtaining mask is: 在复合图像中测量得到灰度图像，通过所述灰度图像得到所述灰度图像中点的亮度最大值，设定阈值为最大亮度的百分比，低于阈值部分设为0，组织部分设为1，形成二值图像，得到掩膜。Measure the grayscale image in the composite image, obtain the maximum brightness value of the midpoint of the grayscale image through the grayscale image, set the threshold as the percentage of the maximum brightness, set the part below the threshold to 0, and set the tissue part to 1. Form a binary image and get a mask. 5.根据权利要求3所述的磁共振成像方法，其特征在于，所述校正所述灵敏度系数，并生成重建图像的步骤是：5. The magnetic resonance imaging method according to claim 3, wherein the step of correcting the sensitivity coefficient and generating a reconstructed image is: 通过参考图像与在视野扫描中获得的复合图像的拟合得到残差；Residuals obtained by fitting the reference image to the composite image obtained in the field of view scan; 引入退火参数，将所述残差导入连接强度函数中，得到所述连接强度函数的值，所述连接强度函数的值作为对角矩阵对角线上的元素；Introducing an annealing parameter, importing the residual into a connection strength function to obtain a value of the connection strength function, and using the value of the connection strength function as an element on the diagonal of a diagonal matrix; 通过复数图像、对角矩阵以及灵敏度系数得到无伪影复合图像。Artifact-free composite images are obtained through complex images, diagonal matrices, and sensitivity coefficients. 6.一种磁共振成像装置，其特征在于，至少包括：6. A magnetic resonance imaging device, characterized in that it at least comprises: 相控阵列线圈，用于对视野扫描得到多个复合图像、以及相对应的复数图像；The phased array coil is used to scan the field of view to obtain multiple composite images and corresponding complex images; 计算装置，用于根据所述多个复合图像得到求和图像，并根据所述多个复合图像及求和图像计算得到灵敏度系数；A computing device, configured to obtain a summation image according to the plurality of composite images, and calculate a sensitivity coefficient according to the plurality of composite images and the summation image; 图像重建装置，用于校正所述灵敏度系数，并生成重建图像。An image reconstruction device is used for correcting the sensitivity coefficient and generating a reconstructed image. 7.根据权利要求6所述的磁共振成像装置，其特征在于，还包括：7. The magnetic resonance imaging apparatus according to claim 6, further comprising: 亮度校正装置，用于校正所述复合图像中的偏差场。Brightness correction means for correcting the bias field in the composite image. 8.根据权利要求7所述的磁共振成像装置，其特征在于，所述亮度校正装置至少包括：8. The magnetic resonance imaging apparatus according to claim 7, wherein the brightness correction device at least comprises: 掩模处理模块，用于对复合图像预处理，得到掩模；A mask processing module is used to preprocess the composite image to obtain a mask; 组织变换模块，用于将所述掩模作用于复合图像中以抽取组织部分图像，并对所述组织部分图像进行对数变换，得到加性偏差场；A tissue transformation module, configured to apply the mask to the composite image to extract a tissue part image, and logarithmically transform the tissue part image to obtain an additive deviation field; 预校正模块，用于根据加性偏差场和掩模，通过高斯低通滤波器和归一化卷积得到初步估计的偏差场；A pre-correction module for deriving an initial estimate of the bias field via a Gaussian low-pass filter and normalized convolution from the additive bias field and the mask; 平滑模块，用于对初步估计的偏差场进行B样条平滑拟合得到精确估计的偏差场，并通过测量得到的图像灰度与所述精确估计的偏差场之比得到校正的复合图像。The smoothing module is used to perform B-spline smooth fitting on the preliminarily estimated deviation field to obtain an accurately estimated deviation field, and obtain a corrected composite image by measuring the ratio of the grayscale of the image obtained to the accurately estimated deviation field. 9.根据权利要求8所述的磁共振成像装置，其特征在于，所述掩模处理模块还用于在复合图像中测量得到灰度图像，通过所述灰度图像得到所述灰度图像中点的亮度最大值，设定阈值为最大亮度的百分比，低于阈值部分设为0，组织部分设为1，形成二值图像，以得到掩模。9. The magnetic resonance imaging apparatus according to claim 8, wherein the mask processing module is also used to measure and obtain a grayscale image in the composite image, and obtain the grayscale image in the grayscale image through the grayscale image The maximum brightness of the point, the threshold is set as the percentage of the maximum brightness, the part below the threshold is set to 0, and the tissue part is set to 1 to form a binary image to obtain a mask. 10.根据权利要求8所述的磁共振成像装置，其特征在于，所述图像重建装置至少包括：10. The magnetic resonance imaging device according to claim 8, wherein the image reconstruction device at least comprises: 拟合模块，用于通过参考图像与所述复合图像的拟合得到残差；A fitting module, used to obtain a residual error by fitting the reference image and the composite image; 估计模块，用于引入退火参数，将所述残差导入连接强度函数中，得到所述连接强度函数的值，所述连接强度函数的值作为对角矩阵中对角线上的元素；The estimation module is used to introduce annealing parameters, import the residual into the connection strength function, obtain the value of the connection strength function, and use the value of the connection strength function as an element on the diagonal in the diagonal matrix; 图像生成模块，用于通过复数图像、对角矩阵以及灵敏度系数得到无伪影复合图像。The image generation module is used to obtain a composite image without artifacts through the complex image, the diagonal matrix and the sensitivity coefficient.
CN 201010255677 2010-08-17 2010-08-17 Magnetic resonance imaging method and device Pending CN101915901A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 201010255677 CN101915901A (en) 2010-08-17 2010-08-17 Magnetic resonance imaging method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 201010255677 CN101915901A (en) 2010-08-17 2010-08-17 Magnetic resonance imaging method and device

Publications (1)

Publication Number Publication Date
CN101915901A true CN101915901A (en) 2010-12-15

Family

ID=43323461

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 201010255677 Pending CN101915901A (en) 2010-08-17 2010-08-17 Magnetic resonance imaging method and device

Country Status (1)

Country Link
CN (1) CN101915901A (en)

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102217934A (en) * 2011-04-08 2011-10-19 中国科学院深圳先进技术研究院 Magnetic resonance imaging method and system
CN102435966A (en) * 2011-09-02 2012-05-02 中国科学院深圳先进技术研究院 Three-dimensional magnetic resonance imaging method and system
CN102521809A (en) * 2011-12-08 2012-06-27 沈阳工业大学 Regularization correction method of magnetic resonance phase array coil image uniformity
CN103163497A (en) * 2011-12-16 2013-06-19 西门子公司 Magnetic resonance system and method to generate a magnetic resonance image of an examination subject
CN103222868A (en) * 2012-01-25 2013-07-31 株式会社东芝 Magnetic resonance imaging apparatus and magnetic resonance imaging method
CN103424727A (en) * 2012-05-23 2013-12-04 深圳市贝斯达医疗器械有限公司 Magnetic resonance image brightness non-uniformity modification algorithm
CN103608839A (en) * 2011-03-28 2014-02-26 皇家飞利浦有限公司 Contrast-dependent resolution image
CN103630860A (en) * 2012-08-28 2014-03-12 江苏晶裕探测器科技有限公司 Method for correcting gray distortion of nuclear magnetic resonance image
CN104520728A (en) * 2012-08-08 2015-04-15 皇家飞利浦有限公司 Multiple shot magnetic resonance imaging with ghosting stability correction
CN106772167A (en) * 2016-12-01 2017-05-31 中国科学院深圳先进技术研究院 Magnetic resonance imaging method employing and device
CN108037496A (en) * 2017-10-11 2018-05-15 中国船舶重工集团公司第七〇五研究所 A kind of free field hydrophone plural number sensitivity accurate measurement method
CN108287325A (en) * 2018-01-03 2018-07-17 上海东软医疗科技有限公司 A kind of image rebuilding method, device and equipment
CN108414462A (en) * 2018-02-10 2018-08-17 中国科学院国家天文台 A kind of low resolution fixed star continuous spectrum automatic Matching Method based on template matches
CN110074786A (en) * 2019-04-30 2019-08-02 上海东软医疗科技有限公司 Nuclear magnetic resonance method for shimming, calculates equipment and MRI system at device

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1799498A (en) * 2004-12-31 2006-07-12 西门子（中国）有限公司 Quick generalized partially parallel acquisition(GRAPA) image reconstruction algorithm for magnetic resonance imaging(MRI)
CN1939213A (en) * 2005-09-30 2007-04-04 Ge医疗系统环球技术有限公司 Mr scanning method and mri apparatus
CN101082659A (en) * 2006-05-29 2007-12-05 西门子公司 Method for improving sensitive coding magnetic resonance imaging by using receiver coil array and the device thereof

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1799498A (en) * 2004-12-31 2006-07-12 西门子（中国）有限公司 Quick generalized partially parallel acquisition(GRAPA) image reconstruction algorithm for magnetic resonance imaging(MRI)
CN1939213A (en) * 2005-09-30 2007-04-04 Ge医疗系统环球技术有限公司 Mr scanning method and mri apparatus
CN101082659A (en) * 2006-05-29 2007-12-05 西门子公司 Method for improving sensitive coding magnetic resonance imaging by using receiver coil array and the device thereof

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
《中国博士学位论文全文数据库 医药卫生科技辑》 20090515 黄鑫 基于多通道并行采集磁共振成像技术的伪影消除方法研究 , 第05期 *

Cited By (25)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103608839A (en) * 2011-03-28 2014-02-26 皇家飞利浦有限公司 Contrast-dependent resolution image
CN103608839B (en) * 2011-03-28 2017-04-12 皇家飞利浦有限公司 Contrast-dependent resolution image
CN102217934A (en) * 2011-04-08 2011-10-19 中国科学院深圳先进技术研究院 Magnetic resonance imaging method and system
CN102435966A (en) * 2011-09-02 2012-05-02 中国科学院深圳先进技术研究院 Three-dimensional magnetic resonance imaging method and system
CN102435966B (en) * 2011-09-02 2014-07-02 中国科学院深圳先进技术研究院 Three-dimensional magnetic resonance imaging method and three-dimensional magnetic resonance imaging system
CN102521809A (en) * 2011-12-08 2012-06-27 沈阳工业大学 Regularization correction method of magnetic resonance phase array coil image uniformity
CN103163497A (en) * 2011-12-16 2013-06-19 西门子公司 Magnetic resonance system and method to generate a magnetic resonance image of an examination subject
US9297871B2 (en) 2011-12-16 2016-03-29 Siemens Aktiengesellschaft Magnetic resonance system and method to generate a magnetic resonance image of an examination subject
CN103222868B (en) * 2012-01-25 2016-08-17 东芝医疗系统株式会社 MR imaging apparatus and MR imaging method
CN103222868A (en) * 2012-01-25 2013-07-31 株式会社东芝 Magnetic resonance imaging apparatus and magnetic resonance imaging method
CN103424727A (en) * 2012-05-23 2013-12-04 深圳市贝斯达医疗器械有限公司 Magnetic resonance image brightness non-uniformity modification algorithm
US9989606B2 (en) 2012-08-08 2018-06-05 Koninklijke Philips N.V. Multiple shot magnetic resonance imaging with ghosting stability correction
CN104520728A (en) * 2012-08-08 2015-04-15 皇家飞利浦有限公司 Multiple shot magnetic resonance imaging with ghosting stability correction
CN103630860B (en) * 2012-08-28 2016-12-21 江苏麦格思频仪器有限公司 The method that the tonal distortion of nuclear magnetic resonance image is modified
CN103630860A (en) * 2012-08-28 2014-03-12 江苏晶裕探测器科技有限公司 Method for correcting gray distortion of nuclear magnetic resonance image
CN106772167A (en) * 2016-12-01 2017-05-31 中国科学院深圳先进技术研究院 Magnetic resonance imaging method employing and device
CN106772167B (en) * 2016-12-01 2019-05-07 中国科学院深圳先进技术研究院 Magnetic resonance imaging method and device
CN108037496A (en) * 2017-10-11 2018-05-15 中国船舶重工集团公司第七〇五研究所 A kind of free field hydrophone plural number sensitivity accurate measurement method
CN108037496B (en) * 2017-10-11 2021-01-12 中国船舶重工集团公司第七一五研究所 Method for accurately measuring complex sensitivity of free-field hydrophone
CN108287325A (en) * 2018-01-03 2018-07-17 上海东软医疗科技有限公司 A kind of image rebuilding method, device and equipment
CN108287325B (en) * 2018-01-03 2020-08-11 上海东软医疗科技有限公司 Image reconstruction method, device and equipment
CN108414462A (en) * 2018-02-10 2018-08-17 中国科学院国家天文台 A kind of low resolution fixed star continuous spectrum automatic Matching Method based on template matches
CN108414462B (en) * 2018-02-10 2020-10-09 中国科学院国家天文台 Low-resolution fixed star continuous spectrum automatic fitting method based on template matching
CN110074786A (en) * 2019-04-30 2019-08-02 上海东软医疗科技有限公司 Nuclear magnetic resonance method for shimming, calculates equipment and MRI system at device
CN110074786B (en) * 2019-04-30 2022-12-06 上海东软医疗科技有限公司 Nuclear magnetic resonance shimming method and device, computing equipment and nuclear magnetic resonance imaging system

Similar Documents

Publication Publication Date Title
CN101915901A (en) Magnetic resonance imaging method and device
CN112368715B (en) Method and system for motion correction for magnetic resonance imaging
EP2660618B1 (en) Biomedical image reconstruction method
CN103632345B (en) A kind of MRI image inhomogeneity correction method based on regularization
CN111081354A (en) System and method for denoising medical images through deep learning network
CN110133556B (en) Magnetic resonance image processing method, device, equipment and storage medium
CN107240125B (en) Diffusion weighted imaging method
Qiu et al. Efficient medical image enhancement based on CNN‐FBB model
Bao et al. Structure-adaptive sparse denoising for diffusion-tensor MRI
CN108287324B (en) Reconstruction method and device of magnetic resonance multi-contrast image
US10698065B2 (en) System, method and computer accessible medium for noise estimation, noise removal and Gibbs ringing removal
CN114140341B (en) Magnetic resonance image non-uniform field correction method based on deep learning
CN102306378B (en) Image enhancement method
CN109544652B (en) Nuclear magnetic resonance multi-weighted imaging method based on depth generation antagonistic neural network
CN110118950B (en) A Phase Correction Method for Bipolar Readout Gradients in Abdominal Quantitative Susceptibility Imaging
CN102521809A (en) Regularization correction method of magnetic resonance phase array coil image uniformity
Patel et al. Medical image fusion based on Multi-Scaling (DRT) and Multi-Resolution (DWT) technique
CN103632367B (en) A kind of MRI coil sensitivity estimation method based on the matching of many tissue regions
Yan et al. A multi-modal medical image fusion method in spatial domain
CN115471580A (en) A physically intelligent high-definition magnetic resonance diffusion imaging method
CN115137347A (en) Myelin sheath quantitative imaging method for three-dimensional ultrashort echo time magnetic resonance fingerprint imaging
CN118628594B (en) Signal equalization method for MRI (magnetic resonance imaging)
CN112785540B (en) A system and method for generating a diffusion weighted image
CN119784598A (en) An underwater image enhancement method based on frequency domain analysis and visual Mamba
Gan et al. SS-JIRCS: Self-supervised joint image reconstruction and coil sensitivity calibration in parallel MRI without ground truth

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C12 Rejection of a patent application after its publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20101215