KR102017433B1 - Medical imaging apparatus and method for tensor estimation through motion correction - Google Patents

Abstract

According to an embodiment of the present invention, provided are a medical imaging apparatus and method for estimating tensor through motion compensation. According to an embodiment of the present invention, the medical imaging apparatus for estimating tensor through motion compensation comprises: a coil unit configured to form a diffusion-weighted gradient magnetic field corresponding to each of a plurality of directions in a target part of an object; an image processing unit corresponding to each direction by the diffusion-weighted gradient magnetic field formed through the coil unit, and obtaining a plurality of slice images of the target part; and a processor configured to generate a diffusion-weighted image corresponding to each direction by sequentially combining the obtained slice images. The processor generates a simulated diffusion-weighted image by matching a reference image, which is not diffusion weighted, with the diffusion-weighted image corresponding to each direction, matches a plurality of slice images constituting the simulated diffusion-weighted image with a plurality of slice images of the diffusion-weighted image corresponding to each direction, corrects a direction vector for each slice image of the diffusion-weighted image corresponding to the direction when each slice is measured by matching with the simulated diffusion-weighted image, estimates a diffusion tensor for each slice image by using the corrected direction vector for each slice image of the diffusion-weighted image corresponding to each direction, and generates a diffusion tensor image by using the corrected direction vector and an estimated diffusion tensor for each slice image of the diffusion-weighted image corresponding to each direction. The present invention can obtain an accurate diffusion tensor image, thereby being able to accurately and precisely diagnose disease.

Description

MEDICAL IMAGING APPARATUS AND METHOD FOR TENSOR ESTIMATION THROUGH MOTION CORRECTION

The present invention relates to a medical imaging apparatus and method for tensor estimation through motion compensation, and more particularly, to a medical imaging apparatus and method for tensor estimation through hierarchical motion correction in a diffusion-weighted image.

Diffusion weighted imaging (DWI) is an imaging technique using Magnetic Resonance Imaging (MRI) devices, which uses diffusion tensor imaging (DTI) to visualize molecular diffusion into the brain, spinal cord, and muscle. It is an imaging technique. In the MRI apparatus, a gradient magnetic field is formed corresponding to each of the plurality of directions, and a diffused emphasis image corresponding to each direction is obtained by the formed gradient magnetic field, and a diffuse tensor image is generated through the obtained diffused emphasis images. Can be.

KR 2014-0089103 A

In order to acquire a diffusion tensor image, the MRI apparatus sequentially acquires slice images corresponding to each of the plurality of slices constituting the target region of the object, and generates a diffusion-weighted image by stacking slice images according to the obtained order. . The MRI apparatus forms a gradient magnetic field that emphasizes diffusion along each of six or more directions, and generates six or more diffusion-weighted images measuring signals emitted by the formed diffusion-weighted gradient magnetic field. The MRI apparatus may construct a diffusion tensor image by a calculation based on the six or more diffusion-weighted images.

However, noises may be generated in each slice image by the movement of the object while acquiring six or more diffusion-weighted images, and an incorrect diffusion-weighted image may be generated when the slice images in which the noises are generated are stacked. have.

In addition, in the case of the diffusion-weighted image, even if motion correction is performed to solve a problem caused by motion, the direction vector of each voxel is not corrected, and thus a diffusion tensor image having a distorted tensor for each voxel may be generated.

Accordingly, there is a need for a method for simultaneously performing motion correction and direction vector correction on an image in which a motion is generated.

Accordingly, an object of the present invention is to provide a medical imaging apparatus and method for tensor estimation through hierarchical motion correction in a diffusion weighted image.

Specifically, the present invention is to provide a method and apparatus for accurately performing motion correction and direction vector correction for a motion-enhanced diffusion-weighted image.

The objects of the present invention are not limited to the above-mentioned objects, and other objects that are not mentioned will be clearly understood by those skilled in the art from the following description.

In order to solve the above problems, there is provided a medical imaging apparatus and method for tensor estimation through motion compensation according to an embodiment of the present invention. According to an embodiment of the present invention, a medical imaging apparatus for tensor estimation through motion compensation includes: a coil unit configured to form a diffusion-weighted gradient magnetic field corresponding to each of a plurality of directions on a target portion of an object; An image processor corresponding to each direction by a diffusion-weighted gradient magnetic field formed through the coil unit, and obtaining a plurality of slice images by measuring a signal emitted according to each of the plurality of slices constituting the target portion; And a processor configured to sequentially combine the obtained plurality of slice images to generate a diffusion-weighted image corresponding to each direction, wherein the processor includes a reference image without diffusion-weighting and a diffusion-weighting corresponding to each direction. Generate a simulated spread-weighted image by matching images, match the plurality of slice images constituting the simulated spread-weighted image with a plurality of slice images of the spread-weighted image corresponding to each direction, and simulate the spread-diffused image. Correcting the direction vector for each slice image of the diffusion-weighted image corresponding to the direction at the time of measuring each slice by matching with, and estimating a diffusion tensor for each slice image using the corrected direction vector, Diffusion Tense Using The Corrected Direction Vector And The Estimated Diffusion Tensor And it generates an image.

In the medical imaging apparatus according to an embodiment of the present invention, a method for estimating a tensor through motion compensation may include: forming a diffusion-weighted gradient magnetic field corresponding to each of a plurality of directions on a target portion of an object through a coil unit; Acquiring a plurality of slice images by measuring a signal emitted corresponding to each of the slices corresponding to each direction by the formed diffusion-weighted gradient magnetic field; Generating a diffusion-weighted image corresponding to each direction by sequentially combining the obtained plurality of slice images; Generating a simulated spread-weighted image by matching a reference image having no spread-weighted image with a spread-weighted image corresponding to each direction; Registering a plurality of slice images constituting the simulated diffusion-weighted image and a plurality of slice images of the diffusion-weighted image corresponding to each direction; Correcting a direction vector of each slice image of a diffusion-weighted image corresponding to a direction when the respective slices are measured by matching with the simulated diffusion-weighted image; Estimating a spreading tensor for each slice image using the corrected direction vector; And generating a diffusion tensor image by using the corrected direction vector and the estimated diffusion tensor.

Specific details of other embodiments are included in the detailed description and drawings.

According to the present invention, when the diffusion tensor image is generated, by correcting the direction vector together with the motion correction of each slice image according to the motion generation, an accurate diffusion tensor image can be obtained, and accurate and delicate disease diagnosis can be performed.

Effects according to the present invention are not limited by the contents exemplified above, and more various effects are included in the present invention.

1 is a schematic diagram illustrating a medical imaging apparatus according to an exemplary embodiment of the present invention.
2 is a schematic diagram illustrating a configuration of a medical imaging apparatus according to an exemplary embodiment of the present invention.
3 is a schematic flowchart illustrating a method for performing motion correction and direction vector correction of a medical image in a medical imaging apparatus according to an exemplary embodiment of the present invention.
4 to 7 are exemplary diagrams for describing a method for performing motion correction and direction vector correction of a medical image in a medical imaging apparatus according to an exemplary embodiment of the present invention.
FIG. 8 is a diagram for comparing a result of performing motion correction and direction vector correction on a medical image in which a motion is generated in a conventional medical imaging apparatus and a medical imaging apparatus according to an exemplary embodiment of the present invention.

Advantages and features of the present invention and methods for achieving them will be apparent with reference to the embodiments described below in detail with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, only the present embodiments to make the disclosure of the present invention complete, and common knowledge in the art to which the present invention pertains. It is provided to fully inform the person having the scope of the invention, which is defined only by the scope of the claims.

Although the first, second, etc. are used to describe various components, these components are of course not limited by these terms. These terms are only used to distinguish one component from another. Therefore, of course, the first component mentioned below may be the second component within the technical spirit of the present invention.

Like reference numerals refer to like elements throughout.

Each of the features of the various embodiments of the present invention may be combined or combined with each other in part or in whole, various technically interlocking and driving as can be understood by those skilled in the art, each of the embodiments may be implemented independently of each other It may be possible to carry out together in an association.

Hereinafter, various embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a schematic diagram illustrating a medical imaging apparatus according to an exemplary embodiment of the present invention.

Referring to FIG. 1, the medical imaging apparatus 1000 may include a cylindrical bore in which the object 10 is carried in order to acquire a medical image of a target portion of the object (eg, a person, an animal, etc.) 10. 100), the object 10 may be placed, and may include a transfer unit 150 for carrying in the bore 100 and a data processor 200 for controlling at least one medical image by controlling the bore 100. For example, the target site may include the spine, chest, upper abdomen, lower abdomen, lungs, brain, liver, varicose veins, uterus, prostate, testes, musculoskeletal system, thyroid or breast. However, the target site is not limited thereto and may be various sites that can be acquired as an image in the medical imaging apparatus 1000. The medical imaging apparatus 1000 may be a magnetic resonance image (MRI) device, a functional MRI device, or the like. The medical image of the target region may be a 2D image, a 3D volume image, a still image of one cut, a video composed of a plurality of cuts, or a plurality of images having various cross-sectional images. In the presented embodiment, an example for acquiring a 3D volume image in a functional magnetic resonance imaging apparatus will be described.

The medical imaging apparatus 1000 applies the first signal to the bore 100 under the control of the data processor 200 to form a constant magnetic field inside the bore 100, and resonates the atomic nucleus of the object 10 located in the magnetic field. At least a second signal for irradiating the second signal to form a magnetic field inside the bore 100, irradiating the second signal to the object 10 having the magnetic field formed therein, and receiving a third signal generated by the second signal irradiation. One coil may be provided. Here, the first signal may be a current signal, the second signal may be an RF signal, and the third signal may be a magnetic resonance signal.

When the magnetic field is formed inside the bore 100 by at least one coil, and the second signal is irradiated to the object 10, a resonance phenomenon may be generated for the atomic nucleus of the object to generate a third signal from the atomic nucleus. The third signal generated as described above may be transmitted to the data processor 200 through at least one coil. The data processor 200 generates a medical image by receiving a third signal, and the generated medical image may be a 3D volume image. The data processing unit 200 includes a display unit 250 and an input unit 260, displays a medical image through the display unit 250, and receives a user input regarding image acquisition or data processing through the input unit 260. Can be.

The 3D volume image is an image in which slice images corresponding to each of the slices constituting the target region of the object are sequentially combined. For example, a 3D volume image is a diffusion modeling a molecular diffusion of the brain, spinal cord, muscle, etc. of the object 10. Tif imaging (diffusion tensor imaging, DTI). The diffusion tensor image is formed by the data processor 200 in the bore 100 in accordance with a plurality of directions (for example, six or more directions), and has a diffused gradient magnetic field, and is obtained for each diffusion by the diffused gradient magnetic field. It can be configured using the highlighted image. Each voxel of the diffusion-weighted image may have a different direction vector for each slice by a diffusion-weighted gradient magnetic field applied in a plurality of directions.

As described above, the spread-weighted image acquired according to each of the plurality of directions may be expressed as shown in Equation 1 below based on the reference image without the spread-weighted gradient magnetic field.

Here, S _i represents the i-th diffusion-weighted image, S ₀ represents the reference image, b _i represents the direction vector (b-vector) of the applied diffused gradient magnetic field (for example, diffused gradient magnetic field vector), D Represents a diffusion tensor. I may also be a natural number.

In Equation 1, the direction vector b may be expressed as Equation 2 below.

Where γ represents a constant representing the rotational magnetic ratio, G represents the intensity of the applied gradient field, δ represents the duration of the applied gradient field, and Δ represents the time interval when two diffusely stressed gradient fields are applied. .

In Equation 1, the diffusion tensor D may be expressed as Equation 3 below.

According to the above, the diffusion tensor has three coefficients of diffusion (D _xx , D _yy , D _zz ) and additionally covariance diffusion coefficients (D _xy , D _xz , D _yz ) to represent the diffusion occurring in the three-dimensional space. It can be expressed as a symmetric matrix using.

On the other hand, at least one coil for irradiating the second signal corresponding to each slice may be provided to be moved to a position corresponding to each slice. When a second signal corresponding to each slice is sequentially irradiated through at least one coil, the data processor 200 sequentially receives a slice signal corresponding to each slice by receiving a third signal for each slice generated by the second signal irradiation. One volume image may be generated by stacking the obtained slice images. The volume image generated as described above may be displayed through the display unit 250.

The second signal corresponding to each slice is irradiated at a time of repetition (TR) through at least one coil. The order and time interval of the second signal to be irradiated may be variously set. For example, when a plurality of slices of the object 10 are made up of ten, at least one coil may radiate a second signal corresponding to each of the ten slices through an x-axis, a y-axis, or a z-axis movement. In various embodiments, the at least one coil irradiates a second signal corresponding to odd ones of ten slices through x, y, or z axis movements, and corresponds to slices corresponding to even numbers. The second signal may be examined.

In order to acquire a diffusion tensor image of the target region of the object 10, the medical imaging apparatus 1000 scans the head region of the object 10 located inside the bore 100 in six or more directions to emphasize six or more diffusion highlights. An image can be obtained. In this case, the number of images acquired in each direction and the time at which the images are acquired may be set by the user. As described above, the scanning method may be performed by irradiating the first signal and the second signal to the object 10 through the bore 100 and receiving a third signal by magnetic field generation and atomic resonance.

Each of the acquired diffusion-enhanced images may be generated by sequentially stacking slice images corresponding to each of the slices constituting the head of the object 10. The stacking order as described above may be determined by the order in which the slice images are acquired. In detail, the medical imaging apparatus 1000 acquires a diffusion-weighted image for each specific time point, and acquires each slice image of the object 10 at a specific time point in a preset order, and acquires each obtained slice image in the acquisition order. By stacking accordingly, a diffusion weighted image may be generated. As described above, when the slice image is acquired, the movement of the object 10 may be generated. As a result, each slice image obtained in time series may be stacked in different directions or angles to generate an incorrect diffusion emphasis image.

In order to solve this problem, the medical imaging apparatus 1000 may perform motion compensation for each obtained slice image. However, in the case of the diffusion-weighted image, since each voxel constituting each slice image has a direction vector, it is necessary to correct the direction vector of each slice image together with motion correction. Thus, in the present invention, an embodiment in which the medical imaging apparatus 1000 corrects the direction vector together with the motion correction for each slice image of the first diffusion weighted image among the plurality of diffusion weighted images will be described.

First, the medical imaging apparatus 1000 acquires a reference image without diffusion enhancement gradient magnetic field and a first diffusion enhancement image to which the diffusion enhancement gradient magnetic field is applied in one of a plurality of directions. Each of the reference image and the first diffusion-weighted image may be registered. The medical imaging apparatus 1000 may use a rigid body transformation registration method to match the reference image and the first diffusion weighted image. However, the present invention is not limited thereto, and various methods for matching may be used.

In detail, the medical imaging apparatus 1000 may acquire an image of the diffusion-weighted gradient magnetic field applied thereto, and acquire the first reference registration data by matching the reference image with the first diffusion-weighted image. Here, the first reference matching data may include a matching matrix function used to convert a position of each voxel of the first spread data image into a position of each voxel of the reference image. Since such image registration is a translational and / or rotational transformation, the registration function may comprise a motional motion component and / or a rotational motion component. For example, the medical imaging apparatus 1000 may match the reference image with a first diffusion weighted image corresponding to one direction and acquire first reference registration data indicating a degree to which the first diffusion weighted image is matched with the reference image. Can be. The first reference registration data includes transform data indicating a degree of shifting and / or rotation-converting each voxel position of the first diffusion weighted image to each voxel position of the reference image so that the first diffusion weighted image is matched with the reference image. can do. Such conversion data may consist of moving and / or rotating motion components.

The medical imaging apparatus 1000 may calculate first error data between the reference image and the first diffusion weighted image by comparing the first diffusion weighted image converted by matching with the reference image. The medical imaging apparatus 1000 may obtain first correction matching data by correcting the first reference matching data using the calculated first error data.

The medical imaging apparatus 1000 may obtain first transformed data for the first diffusion-weighted image converted by matching with the reference image, using the obtained first corrected matching data. The obtained first transformed data may include a rotation transform function (or matrix) indicating a degree in which each voxel of the first diffusion-weighted image is rotated to each voxel of the reference image. The first diffusion-weighted image converted by the matching may be expressed as in Equation 4 below.

Here, y _α may represent the transformed first diffusion-weighted image, R may represent a rotation transformation matrix, and a may represent a displacement vector.

The medical imaging apparatus 1000 may obtain a first correction direction vector by correcting the direction vector of each voxel of the first diffusion-weighted image converted using the obtained first transformed data. The first correction direction vector thus obtained may be expressed as in Equation 5 below.

Here, new B represents a corrected direction vector (b-matrix) for each voxel of the transformed first diffusion-weighted image, B represents a corrected direction vector, and R represents a rotation transform function used for matching. T may represent a matching matrix function.

The medical imaging apparatus 1000 may estimate the first diffusion tensor for each voxel of the converted first diffusion-weighted image by using the obtained first correction direction vector. In order to estimate the first diffusion tensor, the medical imaging apparatus 1000 may use Equation 1 described above.

The above-described operations may be performed with respect to each of six or more diffusion-weighted images. By performing the above-described operations, the medical imaging apparatus 1000 obtains a correction direction vector corresponding to each of the six or more diffusion-weighted images, Using the obtained correction direction vectors, the diffusion tensors for each of six or more diffusion-weighted images may be estimated.

Motion is generated during the generation of one diffusion-weighted image by sequentially stacking each slice image of the target region of the object, and matching is performed when matching between the reference image and the motion-diffused diffusion-weighted image is performed through motion compensation. The same problem may occur as a different gradient magnetic field vector is applied to each slice image of the highlighted image. As a result, the distribution of the image intensity for each slice image may be changed, so it is necessary to accurately correct the distribution of the image intensity.

To this end, the medical imaging apparatus 1000 may generate a simulated DWI by using a corrected direction vector and an estimated diffusion tensor obtained for each of six or more diffusion-weighted images. For example, the medical imaging apparatus 1000 may generate a simulated diffusion weighted image by using Equation 1 described above. The simulated diffusion-weighted image generated as described above may be represented by Equation 6 below.

는 모의 확산 강조 영상을 나타내고, S₀는 기준 영상을 나타내고, b_i는 i번째 확산 강조 영상의 변환된 확산 강조 경사자계 방향 벡터를 나타내며,

는 추정된 확산 텐서를 나타낼 수 있다.Where sDWI and

Denotes a simulated diffuse-weighted image, S ₀ denotes a reference image, b _i denotes a transformed diffuse-weighted gradient magnetic field direction vector of the i-th diffused-weighted image,

Can represent the estimated diffusion tensor.

The medical imaging apparatus 1000 may match each slice image of the simulated diffusion-weighted image generated in this manner with each slice image of the first diffusion-weighted image corresponding to one direction. For example, the medical imaging apparatus 1000 may match the simulated diffusion weighted image and the first diffusion weighted image by an odd upper slice image, an odd lower slice image, an even upper slice image, and an even lower slice image.

As such, the medical imaging apparatus 1000 may obtain the second reference matching data by matching between each slice image of the simulated diffusion-weighted image and each slice image of the first diffusion-weighted image. For example, the second reference matching data may include respective matching data indicating a degree of matching the simulated spread image and the first spread image by the odd upper slice image, the odd lower slice image, the even upper slice image, and the even lower slice image. It may include.

The medical imaging apparatus 1000 may obtain second error data between each slice image of the first diffusion weighted image and each slice image of the simulated diffusion weighted image converted by matching with each slice image of the simulated diffusion weighted image. . For example, the medical imaging apparatus 1000 calculates a first error value between odd upper slice images of the simulated diffusion-weighted image and odd upper slices of the transformed first diffusion-weighted image, and odd-numbered lower portion of the simulated diffusion-weighted image. Computing a second error value between odd lower slices of the slice images and the transformed first diffusion-enhanced image, and generating a second error value between even upper slice images of the simulated diffusion-enhanced image and even upper slices of the transformed first diffusion-enhanced image. The third error value is calculated, and a fourth error value between the even lower slice images of the simulated diffusion-weighted image and the even lower slices of the transformed first diffusion-weighted image is calculated to be the first, second, third, and fourth errors. Second error data including values may be obtained.

The medical imaging apparatus 1000 may generate second correction matching data by correcting the second reference matching data using the obtained second error data. The second correction matching data generated as described above may include a linear or nonlinear matching matrix function in which the sum of squared errors between each slice image of the simulated spread-weighted image and each slice image of the first spread-weighted image is minimized. In order to obtain a matching matrix function that minimizes the sum of error squares, the medical imaging apparatus 1000 generates corrected matching data by correcting the matching data for each slice image using the error value calculated for each slice image, and calculates an error square sum. The process of estimating whether the calculated error sum of squares has a minimum value may be repeatedly performed a specific number of times. In this case, the specific number may be any number and may be the number of times that the sum of squared errors between each slice image of the simulated spread image and each slice image of the first spread image is set to have a minimum value. As described above, the correction matching data obtained for each slice image through the iterative estimation process may be expressed as in Equation 7 below.

는 i번째 확산 경사자계 방향 벡터를 가지는 경사자계가 가해진 경우 j번째 횟수에서 획득될 수 있는 특정 복셀에서의 모의 확산 강조 영상 신호를 나타내고,

는 i번째 방향 벡터를 가지는 경사자계가 가해진 경우 j번째 횟수에서 측정된 특정 복셀에서의 확산 강조 영상 신호를 나타낼 수 있다. 또한, i 및 j는 자연수일 수 있다.Where v represents a particular voxel, T _j (v) represents a matching matrix function in the particular voxel estimated at the j th number,

Denotes a simulated spread-weighted image signal in a particular voxel that can be obtained at a j th number when an inclined magnetic field having an i th diffuse gradient field direction vector is applied,

When the gradient magnetic field having the i-th direction vector is applied, may represent a diffusion-weighted image signal in a particular voxel measured at the j-th number. In addition, i and j may be natural numbers.

)의 각 슬라이스 영상을 이루는 복셀들의 집합(V)과 제1 확산 강조 영상(예:

)의 각 슬라이스 영상을 이루는 복셀들의 집합의 오차 제곱합이 최소가 되는 j번째 정합 데이터(

)를 추정할 수 있다. 여기서, j는 오차 제곱합이 최소가 되는 횟수를 의미한다.Using the above Equation 7, the medical imaging apparatus 1000 simulates a diffusion-weighted image (eg,

Set of voxels constituting each slice of the image (V) and the first diffusion-weighted image (eg

J-th registration data (the minimum sum of squared errors of a set of voxels constituting each slice image)

) Can be estimated. Here, j means the number of times that the sum of error squares is the minimum.

In detail, the medical imaging apparatus 1000 may match the simulated diffusion weighted image with the first diffusion weighted image, and obtain second reference registration data indicating a degree to which the first diffusion weighted image is matched with the simulation diffusion weighted image. Can be. The second reference matching data is a position for each voxel of each slice image of the simulated diffusion-weighted image, the position of each voxel of each slice image of the first diffusion-weighted image so that the first diffusion-weighted image is matched with the simulated diffusion-weighted image. And second conversion data indicating a degree of movement and / or rotation conversion to a position.

The medical imaging apparatus 1000 converts the first type slice images of the first diffusion-weighted image by using the second reference matching data, and converts the first type of the transformed first type slice images and the simulated diffusion-weighted image. The first type of error data between the slice images may be calculated. The medical imaging apparatus 1000 may obtain second reference matching data of the first type by correcting the second reference matching data using the calculated first type of error data. The medical imaging apparatus 1000 converts the second type slice images of the first diffusion-weighted image by using the second reference matching data, and converts the second type of the converted second type slice images and the simulated diffusion-weighted image. The second type of error data between the slice images may be calculated. The medical imaging apparatus 1000 may obtain second reference matching data of the second type by correcting the second reference matching data using the calculated second type of error data.

The medical imaging apparatus 1000 converts the first partial slice images of the first type slice images of the first diffusion-weighted image by using the second type of reference matching data of the first type, and converts the first partial slice images. A first error value between the first partial slice images among the first type slice images of the simulated spread-weighted image may be calculated. Here, the first partial slice images of the first type slice images may be upper partial slice images of odd numbered slice images. The medical imaging apparatus 1000 may generate the first matching data by correcting the second reference matching data of the first type by using the calculated first error value.

The medical imaging apparatus 1000 converts the second partial slice images of the first type slice images of the first diffusion-weighted image by using the second type of reference matching data of the first type, and converts the second partial slice images. A second error value between the second partial slice images among the first type slice images of the simulated spread-weighted image may be calculated. Here, the second partial slice images of the first type slice images may be lower partial slice images of odd-numbered slice images. The medical imaging apparatus 1000 may generate second matching data by correcting the second reference matching data of the first type by using the calculated second error value.

The medical imaging apparatus 1000 converts the first partial slice images of the second type slice images of the first diffusion-weighted image by using the second type of second reference registration data, and converts the first partial slice images. A third error value between the first partial slice images of the second type slice images of the simulated spread-weighted image may be calculated. Here, the first partial slice images of the second type slice images may be upper partial slice images of even-numbered slice images. The medical imaging apparatus 1000 may generate third matching data by correcting the second reference matching data of the second type by using the calculated third error value.

The medical imaging apparatus 1000 converts the second partial slice images of the second type slice images of the first diffusion-weighted image by using the second type of second reference registration data, and converts the second partial slice images. A fourth error value between the second partial slice images among the second type slice images of the simulated spread-weighted image may be calculated. Here, the second partial slice images of the second type slice images may be lower partial slice images among even-numbered slice images. The medical imaging apparatus 1000 may generate fourth matching data by correcting the second reference matching data of the second type by using the calculated fourth error value.

The medical imaging apparatus 1000 may obtain second converted data corresponding to each slice image of the converted first diffusion-weighted image by using the matched data (eg, first to fourth matched data) obtained as described above. . For example, the medical imaging apparatus 1000 may obtain a second transform vector using Equation 4.

The medical imaging apparatus 1000 may obtain a second correction direction vector by correcting a direction vector corresponding to each slice image of the first diffusion-weighted image converted using the second transformed data acquired corresponding to each slice image. have. Here, the direction vector may be a direction vector when measuring a signal emitted according to each of the plurality of slices of the target region. The second correction direction vector obtained corresponding to each slice image may be expressed as in Equation 8 below.

는 j+1번째 횟수에서 추정된 특정 복셀(v)에서의 확산 경사자계 벡터를 나타내고,

는 j번째 횟수에서 사용된 특정 복셀에서의 회전 변환 함수를 나타내며,

는 j번째 횟수에서 사용된 특정 복셀에서의 정합 행렬 함수를 나타낼 수 있다.here,

Denotes the diffusion gradient magnetic field vector in the particular voxel v estimated at the j + 1 th number,

Represents the rotation transform function in the particular voxel used in the j th count,

May represent a matching matrix function in a particular voxel used in the j th number.

The medical imaging apparatus 1000 may estimate a second spreading tensor corresponding to each slice image of the converted first spreading weighted image by using the obtained second correction direction vector. In order to estimate the second diffusion tensor, the medical imaging apparatus 1000 may use Equation 1 described above.

The above-described operations may be performed with respect to each of six or more diffusion-weighted images. By performing the above-described operations, the medical imaging apparatus 1000 obtains a correction direction vector corresponding to each of the six or more diffusion-weighted images, The diffusion tensor can be estimated. The processor 270 may generate a diffusion tensor image by using the correction direction vector and the estimated diffusion tensor obtained corresponding to each of six or more diffusion-weighted images.

Through this, the medical imaging apparatus can obtain accurate diffuse tensor image by correcting the direction vector together with the motion correction of each slice image according to the motion when generating the diffuse tensor image, thereby making it possible to accurately and delicately diagnose the disease. .

2 is a schematic diagram illustrating a configuration of a medical imaging apparatus according to an exemplary embodiment of the present invention.

1 and 2, the medical imaging apparatus 1000 may include a bore 100 and a data processor 200. The bore 100 includes a coil unit 110, and the data processor 200 includes a signal generator 210, a receiver 220, an image processor 230, a storage unit 240, a display unit 250, and an input unit. 260 and processor 270.

First, referring to the bore 100, the coil unit 110 of the bore 100 may include a first signal for forming a magnetic field from the data processor 200 and a second signal for resonating an atomic nucleus of the object 10. When is received, forms a magnetic field based on the first signal, irradiates a second signal to the object 10 in the formed magnetic field to resonate the atomic nucleus of the object 10, and the third signal generated from the atomic nucleus to the data processor 200 ) Can be delivered. For example, the first signal may be a ramp current signal, the second signal may be an RF signal, and the third signal may be a magnetic resonance signal.

The coil unit 110 may include at least one coil disposed in the bore 100 along a predetermined direction. For example, the predetermined direction may be a cylindrical direction of the bore 100. At least one coil is a static magnetic field coil to form a static magnetic field, a gradient coil to apply a gradient to the static magnetic field to form a gradient field (gradient field) and irradiating the RF signal to the object 10 At least one of the RF coil for resonating the atomic nucleus of the object 100 and receiving a magnetic resonance signal generated from the atomic nucleus. These coils are not limited to the embodiments shown, and various types of coils may be used to perform each operation.

In detail, the coil unit 110 may be configured to form a diffusion-weighted gradient magnetic field corresponding to each of the plurality of directions on the target portion of the object 10.

In various embodiments, the RF coil of the at least one coil may be formed to be movable to a position corresponding to each slice in order to obtain an image of each of the plurality of slices of the object 10.

Next, referring to the data processor 200, the signal generator 210 of the data processor 200 may control the diffused gradient magnetic field corresponding to each of the plurality of directions in the bore 100 under the control of the processor 270. The first signal and the second signal may be transmitted to the bore 100 to form a.

The transceiver 220 may receive a third signal from the bore 100 and transmit a control signal for controlling the coil unit 110 of the bore 100 to the bore 100.

The image processor 230 receives the third signal through the receiver 220 and corresponds to each direction by the diffusion-weighted gradient magnetic field formed through the coil unit 110, according to each of the plurality of slices constituting the target portion. A plurality of slice images may be obtained by measuring the emitted signal.

The storage unit 240 may store various data used in the medical imaging apparatus 1000. In detail, the plurality of slice images of the target region obtained through the image processor 230 may be stored, and the medical image to which the plurality of slice images are combined may be stored. The medical image may be a diffusion weighted image.

The display unit 250 may display the generated medical image. The display unit 250 may display the medical image generated by the motion correction and the correction of the direction vector.

The input unit 260 is not limited to a keyboard, a mouse, a touch screen panel, and the like. The input unit 260 may set the medical imaging apparatus 1000 and instruct an operation of the medical imaging apparatus 1000. For example, the medical person may input a request or an instruction for moving the transfer unit 150 through the input unit 260, obtaining a medical image, or performing an operation such as displaying or selecting the obtained medical image.

The processor 270 is operatively connected to the coil unit 110, the signal generator 210, the receiver 220, the image processor 230, the storage unit 240, the display unit 250, and the input unit 260. The controller may perform various commands on the medical imaging apparatus 1000.

Specifically, the processor 270 transmits and receives a control signal for controlling the coil unit 110 of the bore 100 when the input unit 260 receives an input for obtaining a medical image of the object 10. Through the bore 100 can be delivered. The processor 270 transmits the first signal and the second signal to the bore 100 so that the diffuse stress gradient field corresponding to each of the plurality of directions is formed in the bore 100 through the signal generator 210. The third signal may be received from the bore 100 through the transceiver 230. The processor 270 corresponds to each direction by a diffusion-weighted gradient magnetic field formed through the coil unit 110 through the image processor 230, and measures a signal emitted according to each of the plurality of slices of the target region. The slice images may be obtained, and the plurality of slice images may be sequentially combined to generate a medical image corresponding to each direction. When motion is detected in the generated medical image, the processor 270 may simultaneously perform motion correction and direction vector correction on the corresponding medical image.

In detail, the processor 270 acquires a reference image without diffusion-weighted gradient magnetic field and a diffusion-weighted image to which diffusion-weighted gradient magnetic field is applied according to each of six or more directions. Diffusion-weighted images acquired along each of the above directions may be matched.

The processor 270 may acquire matching data of the spread-weighted image corresponding to each direction converted by the matching, and may obtain first converted data of each of the spread-weighted images converted using the acquired matching data. The processor 270 may obtain a first corrected direction vector by correcting the direction vector of each transformed weighted image by using the obtained first transformed data of each transformed weighted image. The processor 270 may estimate the first spreading tensor of each of the spread-weighted images by using the obtained first correction direction vector. As described above, the processor 270 may generate a simulated spread-weighted image by using the first correction direction vector and the estimated first spreading tensor obtained corresponding to each spread-weighted image.

The processor 270 matches each slice image of the generated simulated diffusion-weighted image and each slice image of the diffusion-weighted image corresponding to each direction, and matches each slice image of the diffusion-weighted image corresponding to each direction converted by matching. Matching data may be obtained.

In detail, the processor 270 may match the simulated spread-weighted image with the spread-weighted image corresponding to each direction to obtain second reference matching data indicating a degree to which the spread-weighted image corresponding to each direction is matched to the simulated spread-weighted image. can do. The processor 270 converts the first type slice images of the spread-weighted image corresponding to each direction by using the second reference matching data, and the first type of the transformed first type slice images and the simulated spread-weighted image. Error data of a first type between slice images of may be calculated. The processor 270 converts the second type slice images of the diffusion-weighted image corresponding to each direction by using the second reference matching data, and the second type of the converted second type slice images and the simulated diffusion-weighted image. The second type of error data between the slice images may be calculated. The processor 270 may obtain second reference matching data of the second type by correcting the second reference matching data using the calculated second type of error data.

The processor 270 converts the first partial slice images among the slice images of the first type of the diffusion-weighted image corresponding to each direction by using the second type of reference matching data of the first type, and converts the first partial slice images. And a first error value between the first partial slice images among the first type slice images of the mock spread-weighted image. The processor 270 may generate the first matching data by correcting the second reference matching data of the first type by using the calculated first error value.

The processor 270 converts the second partial slice images of the first type slice images of the diffusion-weighted image corresponding to each direction by using the second type of reference matching data of the first type, and converts the second partial slice images. And a second error value between the second partial slice images of the first type slice images of the mock spread-weighted image. The processor 270 may generate second matching data by correcting the second reference matching data of the first type by using the calculated second error value.

The processor 270 converts the first partial slice images of the second type slice images of the diffusion-weighted image corresponding to each direction by using the second reference matching data of the second type, and converts the first partial slice images. And a third error value between the first partial slice images among the second type slice images of the mock spread-weighted image. The processor 270 may generate third matching data by correcting the second reference matching data of the second type by using the calculated third error value.

The processor 270 converts the second partial slice images of the second type slice images of the diffusion-weighted image corresponding to each direction by using the second reference matching data of the second type, and converts the second partial slice images. And a fourth error value between the second partial slice images among the second type slice images of the mock spread-weighted image. The processor 270 may generate fourth matching data by correcting the second reference matching data of the second type by using the calculated fourth error value.

The processor 270 may obtain transformed data for each slice image of the diffusion-weighted image corresponding to each direction by using matching data for each slice image of the diffusion-weighted image corresponding to each direction. In detail, the processor 270 may acquire second converted data corresponding to each slice image of the diffusion-weighted image corresponding to each direction converted using the acquired matching data (eg, first to fourth matching data). Can be.

The processor 270 may obtain a correction direction vector by correcting the direction vector of each slice image of the diffusion-weighted image corresponding to each direction by using the transform data of each slice image of the diffusion-weighted image corresponding to each direction. have. In detail, the processor 270 may correspond to each slice image of the diffusion emphasis image corresponding to each direction converted by using the second transformed data obtained corresponding to each slice image of the diffusion emphasis image corresponding to each converted direction. The second correction direction vector may be obtained by correcting the direction vector.

The processor 270 may estimate a diffusion tensor for each slice image of the diffusion-weighted image corresponding to each direction by using the obtained correction direction vector for each slice image of the diffusion-weighted image corresponding to each direction. In detail, the processor 270 may estimate a second diffusion tensor corresponding to each slice image of the diffusion-weighted image corresponding to each of the converted directions by using the obtained second correction direction vector.

In this way, the processor 270 may generate a diffusion tensor image using the correction direction vector and the estimated diffusion tensor obtained corresponding to each slice image of the diffusion-weighted image corresponding to each direction.

3 is a schematic flowchart illustrating a method for performing motion correction and direction vector correction of a medical image in a medical imaging apparatus according to an exemplary embodiment of the present invention.

1 and 3, the medical imaging apparatus 1000 may generate a simulated spread-weighted image by matching a reference image without diffusion emphasis and a diffusion-weighted image corresponding to each of the plurality of directions and having diffusion emphasis. It may be (S300). In detail, the medical imaging apparatus 1000 obtains first reference registration data by matching the reference image with the diffusion-weighted image corresponding to each direction, and compares the reference image with the matched diffusion-weighted image corresponding to each direction to obtain a first image. 1 error data can be calculated. The medical imaging apparatus 1000 corrects the first reference matching data by using the calculated first error data to obtain first correction matching data, and uses the obtained first correction matching data to perform diffusion emphasis corresponding to each direction. First transformed data about an image may be obtained. The medical imaging apparatus 1000 obtains a first correction direction vector by correcting the direction vector of the matched diffusion-weighted image corresponding to each direction by using the acquired first transformed data, and uses the obtained first correction direction vector. The first spreading tensor of the matched spread-weighted image corresponding to each direction may be estimated. The medical imaging apparatus 1000 may generate a simulated spread-weighted image by using the estimated first spreading tensor of the matched spread-weighted image corresponding to each direction.

The medical imaging apparatus 1000 may match the plurality of slice images constituting the simulated diffusion weighted image and the plurality of slice images of the diffusion weighted image corresponding to each direction (S310). In detail, the medical imaging apparatus 1000 may obtain the second reference matching data by matching the simulated diffusion weighted image and the diffusion weighted image corresponding to each direction. The medical imaging apparatus 1000 may use the first type slice image among the plurality of slice images of the simulated diffusion-weighted image and the plurality of slice images of the diffusion-weighted image corresponding to each direction by using the acquired second reference matching data. Matching the slice image of the first type and the slice image of the second type among the plurality of slice images of the plurality of slice images of the mock diffusion-weighted image and the slice image of the second type among the plurality of slice images of the diffusion-weighted image corresponding to each direction. Can match. Through such matching, the medical imaging apparatus 1000 calculates a first type of error value between the slice image of the first type of the simulated diffusion-weighted image and the slice image of the first type of the diffusion-weighted image corresponding to each direction, and simulates. The second type error value between the slice type image of the diffusion type image and the slice type type image of the diffusion type image corresponding to each direction may be calculated. The medical imaging apparatus 1000 obtains the second reference matching data of the first type by correcting the second reference matching data using the first type of error value, and uses the second type of error value to match the second reference matching data. The second reference matching data of the second type may be obtained by correcting the data.

The medical imaging apparatus 1000 converts the first partial slice image of the first type slice image of the diffusion-weighted image corresponding to each direction by using the second type of reference matching data of the first type, and converts the first type slice image. The first error value between the first partial slice images of the transformed first type slice image and the simulated diffusion enhanced image may be calculated by comparing the first partial slice images and the simulated diffusion enhanced image. The medical imaging apparatus 1000 may generate the first matching data by correcting the second reference matching data of the first type by using the calculated first error value.

The medical imaging apparatus 1000 converts the second partial slice image of the first type slice image of the diffusion-weighted image corresponding to each direction by using the second type of reference matching data of the first type, and converts the converted first type image. The second error value between the second partial slice image of the transformed first type and the simulated diffusion emphasis image may be calculated by comparing the second partial slice image and the simulated diffusion emphasis image. The medical imaging apparatus 1000 may generate second matching data by correcting the second reference matching data of the first type by using the calculated second error value.

The medical imaging apparatus 1000 converts the first partial slice image of the second type slice image of the diffusion-weighted image corresponding to each direction by using the second type of second reference matching data, and converts the converted second type of image. A third error value between the first partial slice image of the converted second type slice image and the simulated diffusion-weighted image may be calculated by comparing the first partial slice image and the simulated diffusion-weighted image. The medical imaging apparatus 1000 may generate third matching data by correcting the second reference matching data of the second type by using the calculated third error value.

The medical imaging apparatus 1000 converts the second partial slice image of the second type slice image of the diffusion-weighted image corresponding to each direction by using the second type of second reference registration data, and converts the converted second type image. A fourth error value between the second partial slice image of the converted second type slice image and the simulated diffusion emphasis image may be calculated by comparing the second partial slice image and the simulated diffusion emphasis image. The medical imaging apparatus 1000 may generate fourth matching data by correcting the second reference matching data of the second type by using the calculated fourth error value.

The medical imaging apparatus 1000 may correct the direction vector of each slice image of the diffusion-weighted image corresponding to the direction when measuring each slice by matching the simulated diffusion-weighted image (S320). In detail, the medical imaging apparatus 1000 uses the matched data (eg, first to fourth matched data) obtained as described above, and the second transformed data for each slice image of the diffusion-weighted image converted in each direction. The second correction direction vector may be obtained by correcting a direction vector of each slice image of the diffusion-weighted image corresponding to each direction by using the obtained second transformed data.

The medical imaging apparatus 1000 may estimate the diffusion tensor of each slice image by using the corrected direction vector of each slice image of the diffusion-weighted image corresponding to each direction (S330). In detail, the medical imaging apparatus 1000 may estimate the second diffusion tensor for each slice image of the diffusion-weighted image converted in each direction by using the obtained second correction direction vector.

The medical imaging apparatus 1000 may generate a diffusion tensor image by using the corrected direction vector and the estimated diffusion tensor of each slice image of the diffusion-weighted image corresponding to each direction (S340). In detail, the medical imaging apparatus 1000 may generate a diffusion tensor image by using the estimated second diffusion tensor. These operations are applied to the diffusion-weighted image obtained corresponding to each of the plurality of directions, the direction vector is corrected for each direction, the diffusion tensor is estimated, and the diffusion tensor image is corrected through the corrected direction vector and the estimated diffusion tensor. Can be generated.

4 to 7 are exemplary diagrams for describing a method for performing motion correction and direction vector correction of a medical image in a medical imaging apparatus according to an exemplary embodiment of the present invention.

1 and 4, the medical imaging apparatus 1000 obtains a first volume image 402 composed of each of the plurality of voxels 400 as illustrated in FIG. 4A. can do. The first volume image 402 may be a diffusion-weighted image obtained by a diffusion-weighted gradient magnetic field corresponding to each of a plurality of directions formed in the core 100. Each voxel of the diffusion-weighted image may have a direction vector for each slice by a diffusion-weighted gradient magnetic field applied along a plurality of directions. In the illustrated embodiment of FIG. 4, the plurality of slice images 404 constituting the volume image 402 will be described as having the same direction vector. Each of the slice images 404 of the obtained volume image 402 may have different image intensities.

The medical imaging apparatus 1000 acquires a second volume image 406 as shown in FIG. 4B, and a motion may be detected in the acquired second volume image 406. For example, the second volume image 406 may be a volume image rotated and converted. In this case, the medical imaging apparatus 1000 may match the second volume image 406 with the reference volume image and rotate the voxel position of the second volume image 406 to move to each voxel position of the reference volume image. . As such, the medical imaging apparatus 1000 may generate the converted second volume image 408 as illustrated in FIG. 4C. However, even if the motion correction is performed, the direction vector of each voxel of the second volume image 406 may not be corrected as shown in FIG. 4C. Accordingly, the medical imaging apparatus 1000 obtains reference matching data by matching the reference volume image and the second volume image 406, and compares the second volume image 408 converted by the matching with the reference volume image. Error data between two volume images may be calculated. The medical imaging apparatus 1000 may correct the reference match data using the calculated error data, and obtain the converted data using the corrected reference match data. The medical imaging apparatus 1000 corrects the direction vector of the converted second volume image 408 by using the obtained converted data, and thus, the second volume image 410 having the corrected direction vector as shown in FIG. 4D. Can be generated.

1 and 5, when the medical imaging apparatus 1000 generates a motion while acquiring each of a plurality of slice images of a target portion of the object 10 at a specific time point, as shown in FIG. 5A. The first volume image 500 in which different movements are generated may be obtained. For example, the obtained first volume image 500 may include six slice images 502, 504, 506, 508, 510 that are differently generated and rotated to different sizes while obtaining each slice image. 512 may be formed by stacking. In this case, motion correction may be performed corresponding to each slice image, and thus correction of a direction vector corresponding to each slice image may be performed. As described above, in order to perform motion correction and direction vector correction according to each slice image, the medical imaging apparatus 1000 includes a plurality of slices constituting a reference volume image for each slice image constituting the first volume image 500. Match each image. This matching operation will be described below in detail with reference to FIG. 6.

1 and 6, the medical imaging apparatus 1000 matches a reference volume image V ₁ and a first volume image V ₂ , and the first volume image V ₂ is a reference volume image V _1. ) Can be calculated as a reference matching matrix M _f . Here, the reference matching matrix M _f is a reference volume based on the position values (eg, x-axis position value, y-axis position value, and z-axis position value) of each voxel with respect to the entire slice image of the first volume image V ₂ . It may be reference registration data indicating the degree of rotation and translation by transforming the position of each voxel with respect to the whole slice image of the image V ₁ .

The medical imaging apparatus 1000 converts each of the odd slice images of the first volume image V ₂ using the reference matching matrix M _f , and converts each of the converted odd slice images and the reference volume image V ₁ . The odd values of the first type may be calculated by comparing each of the odd slice images. Using the error values of a first type calculated in this way medical imaging device 1000 may be corrected based on the matching matrix (M _f) generating a reference matching matrix (M _o) of the first type. For example, the medical imaging apparatus 1000 has a position value of each voxel for each of the converted odd slice images of the first volume image V ₂ , and a position of each voxel for each of the odd slice images of the reference volume image. by correcting the standard matching matrix (M _f) to match the value may be generated based on the matching matrix (M _o) of the first type.

The medical imaging apparatus 1000 converts each of the even slice images of the first volume image V ₂ using the reference matching matrix M _f , and converts each of the converted even slice images and the reference volume image V ₁ . Each of the even slice images of may be compared to calculate an error value of the second type. Using the error values of a second type calculated in this way medical imaging device 1000 may be corrected based on the matching matrix (M _f) generating a reference matching matrix (M _e) of the second type. For example, the medical imaging apparatus 1000 may determine a position value of each voxel for each of the converted even slice images of the first volume image V ₂ for each of the even slice images of the reference volume image V ₁ . The reference matching matrix M _f may be corrected to match the position value of each voxel to generate the second type reference matching matrix M _e .

Based on the medical imaging apparatus 1000 is each of the first volume, converting the odd-numbered top slice of the image in each image (V _2), and the transformed odd-numbered top slice image using a standard matching matrix (M _o) of the first type The first error value may be calculated by comparing each of the odd upper slice images of the volume image V ₁ . The medical imaging apparatus 1000 may generate the first matching matrix M _ou by correcting the reference matching matrix M _o of the first type by using the calculated first error value. For example, the medical imaging apparatus 1000 may determine a position of each voxel of each of the transformed odd upper slice images of the first volume image V ₂ using the calculated first error value. by correcting the standard matching matrix (M _o) of the first type to match the position of each voxel for each of the odd-numbered top slice image. ₁₎ may generate a first matching matrix (M _ou).

Based on the medical imaging apparatus 1000 is each of the standard matching matrix (M _o) of the first volume image (V ₂₎ the odd sub-slice images of the odd sub-slice image converted, respectively, and the conversion of using a first type The second error value may be calculated by comparing each of the odd lower slice images of the volume image V ₁ . The medical imaging apparatus 1000 may generate the second matching matrix M _ol by correcting the reference matching matrix M _o of the first type by using the calculated second error value. For example, the medical imaging apparatus 1000 uses the calculated second error value to determine the position value of each voxel for each of the converted odd lower slice images of the first volume image V ₂ . ₁₎ it is possible to generate a second matching matrix (M _ol) to correct the reference matching matrix (M _o) of the first type to match the position of each voxel for each of the odd sub-slice images.

The medical imaging apparatus 1000 converts each of the even upper slice images of the first volume image V ₂ using the second type of reference matching matrix M _e , and converts each of the converted even upper slice images and the reference. A third error value may be calculated by comparing each of the even upper slice images of the volume image V ₁ . The medical imaging apparatus 1000 may generate the third matching matrix M _eu by correcting the reference matching matrix M _e of the second type by using the calculated third error value. For example, the medical imaging apparatus 1000 uses the calculated third error value to determine the position value of each voxel for each of the even-numbered higher slice images of the first volume image V ₂ . The third matching matrix M _eu may be generated by correcting the reference matching matrix M _e of the second type to match the position value of each voxel for each of the even upper slice images of ₁ ).

The medical imaging apparatus 1000 converts each of the even lower slice images of the first volume image V ₂ by using the second type of reference matching matrix M _e , and converts each of the converted even lower slice images and the reference. A fourth error value may be calculated by comparing each of the even lower slice images of the volume image V ₁ . Imaging device 1000 may generate a fourth matching matrix (M _el) to correct the reference matching matrix (M _e) of the second type with a fourth error value calculated in this way. For example, the medical imaging apparatus 1000 uses the calculated fourth error value to determine the position value of each voxel for each of the even-numbered lower slice images of the first volume image V ₂ . The fourth matching matrix M _el may be generated by correcting the second type reference matching matrix M _e so as to match the position value of each voxel for each of the even lower slice images of ₁ ).

Through this, the medical imaging apparatus 1000 may generate the first volume image 514 whose motion is corrected as shown in FIG. However, the motion-corrected first volume image 514 may be corrected such that the direction vector corresponding to each slice image matches the direction vector corresponding to each slice of the reference volume image.

Accordingly, in order to correct the direction vector, the medical imaging apparatus 1000 performs a first matching matrix M _ou , a second matching matrix M _ol , a third matching matrix M _eu , and a first matching matrix generated by the above-described operation. A transformation function for each slice image of the first volume image V ₂ may be obtained using the four matching matrix _Mel . For example, the transform function may be calculated by Equation 4 described above.

Medical imaging apparatus 1000 of FIG. 5 by correcting the direction vector for each slice image of the first volume image (V ₂₎ a first volume image (V ₂₎ using a constant function obtained for each slice image of As shown in (c), the first volume image 516 with the corrected direction vector corresponding to each slice image may be obtained. For example, the direction vector may be corrected by Equation 5 described above. An operation of correcting the direction vector corresponding to each slice image will be described in detail with reference to FIG. 7.

Referring to FIGS. 1 and 7, the medical imaging apparatus 1000 may generate a movement as shown in FIG. 7A and acquire a diffusion-weighted image 700 corresponding to a specific direction. The particular voxel 702 of the diffusion-weighted image 700 may have a direction vector 704 corresponding to the first direction. In order to correct the motion, the medical imaging apparatus 1000 may generate the rotationally-converted diffusion-weighted image 710 as shown in FIG. 7B by matching the reference image and the diffusion-weighted image 700. In this case, the position of the specific voxel 702 of the diffusion-weighted image 700 may be rotated to the position of the specific voxel 712 of the diffusion-weighted image 710 rotated to match the position of the specific voxel of the reference image. . In this case, since the particular voxel 712 of the rotation-converted diffusion-weighted image 710 has the same direction vector 714 as the specific voxel 702 of the diffusion-weighted image 700, the medical imaging apparatus 1000 is diffuse-weighted. The particular voxel 712 of the diffusion-enhanced image 710 rotated according to the degree to which the position of the particular voxel 702 of the image 700 is rotationally transformed into the specific voxel 712 of the diffusion-enhanced image 710 rotated. The direction vector 714 of may be corrected to the second direction vector 716. As described above, the medical imaging apparatus 1000 may use Equation 5 or Equation 8 to correct the second direction vector 716.

FIG. 8 is a diagram for comparing a result of performing motion correction and direction vector correction on a medical image in which a motion is generated in a conventional medical imaging apparatus and a medical imaging apparatus according to an exemplary embodiment of the present invention.

Referring to FIG. 8, in the conventional medical imaging apparatus, when a diffusion tensor image is generated using a plurality of diffusion-weighted images, a motion occurs while sequentially stacking each of a plurality of slice images corresponding to a target region of an object. As shown in (a) of FIG. 8, different motions are generated in each slice image, and an incorrect diffusion tensor image is generated even when the motion is corrected.

The medical imaging apparatus 1000 of the present invention performs motion correction for each slice image and corrects the direction vector of each of the voxels of each slice image, thereby correcting the accurate and delicate diffusion tensor image as shown in FIG. Can be generated.

As described above, according to the present invention, when the diffusion tensor image is generated, by correcting the direction vector together with the motion correction of each slice image according to the motion generation, an accurate diffusion tensor image can be obtained, and accurate and delicate disease diagnosis can be performed.

Apparatus and method according to an embodiment of the present invention can be implemented in the form of program instructions that can be executed by various computer means may be recorded on a computer readable medium. Computer-readable media may include, alone or in combination with the program instructions, data files, data structures, and the like.

The program instructions recorded on the computer readable medium may be those specially designed and constructed for the present invention, or may be known and available to those skilled in the computer software arts. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tape, optical media such as CD-ROMs, DVDs, and magnetic disks, such as floppy disks. Hardware devices specially configured to store and execute program instructions such as magneto-optical media and ROM, RAM, flash memory and the like. In addition, the above-described medium may be a transmission medium such as an optical or metal wire, a waveguide, or the like including a carrier wave for transmitting a signal specifying a program command, a data structure, and the like. Examples of program instructions include not only machine code generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like.

The hardware device described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

Although the embodiments of the present invention have been described in more detail with reference to the accompanying drawings, the present invention is not necessarily limited to these embodiments, and various modifications can be made without departing from the spirit of the present invention. . Therefore, the embodiments disclosed in the present invention are not intended to limit the technical idea of the present invention but to describe the present invention, and the scope of the technical idea of the present invention is not limited by these embodiments. Therefore, it should be understood that the embodiments described above are exemplary in all respects and not restrictive. The protection scope of the present invention should be interpreted by the following claims, and all technical ideas within the equivalent scope should be interpreted as being included in the scope of the present invention.

100: bore
110: coil part
150: transfer unit
200: data processing unit
210: signal generator
220: transceiver
230: the image processing unit
240: storage unit
250: display unit
260: input unit
270: processor
1000: medical imaging device

Claims (12)

A coil unit configured to form a diffusion-weighted gradient magnetic field corresponding to each of the plurality of directions in a target portion of the object;
An image processor corresponding to each direction by a diffusion-weighted gradient magnetic field formed through the coil unit, and obtaining a plurality of slice images by measuring a signal emitted according to each of the plurality of slices constituting the target portion; And
And a processor configured to sequentially combine the obtained plurality of slice images to generate a spread emphasis image corresponding to each direction.
The processor,
A simulated spread-weighted image is generated by matching a reference image having no spread-weighted image with a spread-weighted image corresponding to each direction.
Matching the plurality of slice images constituting the simulated diffusion emphasis image and the plurality of slice images of the diffusion emphasis image corresponding to each direction,
Correcting a direction vector for each slice image of a diffusion-weighted image corresponding to a direction when the respective slices are measured by matching with the simulated diffusion-weighted image,
Estimating a spreading tensor for each slice image using the corrected direction vector,
And generate a diffusion tensor image using the corrected direction vector and the estimated diffusion tensor. The method of claim 1, wherein the processor,
Matching the reference image with a diffusion-weighted image corresponding to each direction,
Obtaining first reference matching data indicating a degree of matching the diffusion-weighted image corresponding to each direction to the reference image,
Calculating first error data between the reference image and the converted spread-weighted image by comparing the spread-weighted image converted by matching with the reference image and the reference image,
Correcting the first reference matching data using the calculated first error data to obtain first correction matching data;
Obtaining first transformed data for the transformed spread-weighted image using the obtained first corrected match data,
Obtaining a first correction direction vector by correcting a direction vector for each voxel of the transformed spread-weighted image using the obtained first transformed data,
Estimate a first spreading tensor for each voxel of the transformed spread-weighted image using the first correction direction vector,
And generate the simulated spread-weighted image using the obtained first correction direction vector and the estimated first spreading tensor. The method of claim 1, wherein the processor,
Matching the simulated diffusion-weighted image and the diffusion-weighted image corresponding to each direction;
Obtaining second reference matching data indicating a degree to which the simulated spread-weighted image matches the spread-weighted image corresponding to each direction;
Obtaining second error data between each slice image of the diffusion-weighted image converted by matching with the simulated diffusion-weighted image and each slice image of the simulated diffusion-weighted image,
Correcting the second reference correction data using the obtained second error data to obtain second correction matching data;
Acquiring second transformed data for the transformed spread-weighted image by using the obtained second corrected matching data;
And correcting a direction vector for each voxel of the transformed spread-weighted image using the obtained second transformed data to obtain a second correction direction vector. The method of claim 3, wherein the second correction matching data,
And a linear or nonlinear matching matrix function configured to minimize the sum of squared errors between each slice image of the simulated spread-enhanced image and each slice image of the transformed spread-weighted image. . The processor of claim 3, wherein the processor comprises:
And estimating a second spreading tensor for each voxel of the transformed spread-weighted image using the obtained second corrected direction vector. The method of claim 4, wherein the processor,
And generate the diffusion tensor image using the obtained second correction direction vector and the estimated second diffusion tensor. Forming a diffusion-weighted gradient magnetic field corresponding to each of the plurality of directions in the target portion of the object through the coil unit;
Acquiring a plurality of slice images by measuring a signal emitted corresponding to each of the slices corresponding to each direction by the formed diffusion-weighted gradient magnetic field;
Generating a diffusion-weighted image corresponding to each direction by sequentially combining the obtained plurality of slice images;
Generating a simulated spread-weighted image by matching a reference image having no spread-weighted image with a spread-weighted image corresponding to each direction;
Registering a plurality of slice images constituting the simulated diffusion-weighted image and a plurality of slice images of the diffusion-weighted image corresponding to each direction;
Correcting a direction vector of each slice image of a diffusion-weighted image corresponding to a direction when the respective slices are measured by matching with the simulated diffusion-weighted image;
Estimating a spreading tensor for each slice image using the corrected direction vector; And
And generating a diffuse tensor image using the corrected direction vector and the estimated spread tensor. The method of claim 7, wherein the generating of the simulated diffusion weighted image comprises:
Matching the reference image with a diffusion-weighted image corresponding to each direction;
Acquiring first reference registration data indicating a degree to which the diffusion image corresponding to each direction is matched with the reference image;
Calculating first error data between the spread-weighted image and the reference image converted by matching with the reference image;
Obtaining first corrected match data by correcting the first reference match data using the calculated first error data;
Acquiring first transformed data of the transformed spread-weighted image using the obtained first corrected match data;
Obtaining a first correction direction vector by correcting a direction vector of each voxel of the transformed spread-weighted image using the obtained first transformed data;
Estimating a first spreading tensor for each voxel of the transformed spread-weighted image using the first correction direction vector; And
And generating the simulated spread-weighted image by using the obtained first correction direction vector and the estimated first spreading tensor. The method of claim 7, wherein the correcting of the direction vector of each slice image of the diffusion-weighted image corresponding to each direction comprises:
Registering the simulated spread-weighted image and the spread-weighted image corresponding to each direction;
Acquiring second reference registration data representing a degree to which the simulated spread-weighted image matches the spread-weighted image corresponding to each direction;
Acquiring second error data between each slice image of the spread-weighted image converted by matching with the simulated spread-weighted image and each slice image of the simulated spread-weighted image;
Correcting the second reference correction data using the obtained second error data to obtain second correction matching data;
Acquiring second transformed data for the transformed spread-weighted image using the obtained second corrected match data; And
Obtaining a second correction direction vector by correcting a direction vector of each voxel of the transformed spread-weighted image by using the obtained second transformed data. Way for you. The method of claim 9, wherein the second correction matching data,
And a linear or nonlinear matching matrix function that minimizes the sum of squared errors between each slice image of the simulated spread-weighted image and each slice image of the transformed-diffusion-weighted image. Way. The method of claim 9, wherein estimating a diffusion tensor for each slice image comprises:
Estimating a second spreading tensor for each voxel of the transformed spread-weighted image by using the obtained second corrected direction vector. The method of claim 11, wherein generating the diffusion tensor image comprises:
And generating the diffuse tensor image using the acquired second correction direction vector and the estimated second diffuse tensor.

Images