KR101773617B1 - Method and MRI device for correcting distortion in an EPI image - Google Patents

Abstract

A method of correcting distortion of an echo planar imaging (EPI) image according to the present invention includes: forming an EPI image from a target object; Forming three-dimensional point spread function (PSF) data from the object; Integrating the three-dimensional PSF data into a predetermined plane to obtain two-dimensional PSF data of a line pattern composed of a plurality of pixels on each of a plurality of planes formed by axes in which an image is distorted and non-distorted axes; A step of selectively obtaining a PSF movement mapping map using the slope of the two-dimensional PSF data of the line pattern in each of the plurality of planes, and a step of correcting the distortion of the EPI image using the PSF movement mapping map.

Description

[0001] The present invention relates to a method of correcting distortion of an EPI image and an MRI apparatus using the same,

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to image correction in a magnetic resonance imaging (UHF) magnetic resonance imaging (MRI) apparatus, and more particularly, to image distortion generated in an EPI image acquired by an echo planar imaging (EPI) Function) mapping.

Magnetic Resonance Imaging (MRI) is an imaging device that images the magnetic characteristics of a material constituting a living body according to magnetic field and high frequency application, and is one of the most widely used imaging devices. Recently, an echo planar imaging (EPI) technique, which is effective in data processing time to acquire an image, has been used as a magnetic resonance imaging technique.

When the EPI technique is applied to an MRI apparatus, functional MRI (fMRI), diffusion weighted imaging (DTI), perfusion MRI, and cardiac imaging can be obtained. However, there is a problem that geometric distortion and intensity distortion occur due to off-resonance effect, that is, inhomogeneity and susceptibility of the main magnetic field (B ₀ ) in the image using the EPI technique. The distortion in the EPI image becomes more severe as the intensity of the magnetic field becomes larger, and the degree of distortion of the EPI image is further increased in the ultra high magnetic field MRI apparatus. Therefore, it is necessary to correct the distortion of the EPI image in order to obtain more accurate clinical results.

Patent Document 1: Korean Patent Laid-Open No. 10-2010-0010481 (published on February 21, 2010)

Accordingly, the present invention provides a method for effectively correcting image distortion using an optional PSF (Point Spread Function) mapping method in an EPI image acquired by an EPI technique in an MRI apparatus.

An MRI apparatus according to the present invention includes: a driving unit operable to generate a predetermined EPI (Echo Planar imaging) pulse sequence; A coil portion including a plurality of coils, wherein the coil portion generates a first magnetic field in the object in response to a pulse sequence from the driving portion when the object is located near the coil portion, Operating to cause a first magnetic resonance signal to be induced; And a data processing unit operable to form EPI image data based on a first magnetic resonance signal induced in the coil part, wherein the driving unit is further operable to generate a predetermined pulse sequence of PSF (Point Spread Function) The coil unit generates a second magnetic field in response to the PSF pulse sequence from the driving unit when the object is positioned near the coil part and generates a second magnetic field in response to the generated second magnetic field so that a second magnetic resonance signal Wherein the data processing unit is further operative to form three-dimensional PSF image data based on the second magnetic resonance signal, the three-dimensional PSF image data being defined on a three-dimensional space having axes, The data processing unit integrates the three-dimensional PSF data in the direction of one of the axes, A first processing unit for obtaining a plurality of two-dimensional PSF data sets arranged perpendicular to an axis and corresponding to a plurality of planes of an image-distorted axis and an image-unstrained axis, A second processing unit for selectively obtaining a PSF movement mapping map in accordance with a slope of a line pattern composed of a plurality of pixels in the first processing unit and a third processing unit for correcting distortion of the EPI image data using the PSF movement mapping map .

A method for correcting distortion of an echo planar imaging (EPI) image in a magnetic resonance imaging apparatus according to another embodiment of the present invention includes the steps of: a) repeatedly applying a predetermined EPI pulse sequence to form EPI image data from a target object; b) Point Spread Function (PSF) data from the object by repeatedly applying a predetermined PSF pulse sequence, wherein the 3D PSF image data is defined in a three-dimensional space having axes, ; c) a step of integrating the three-dimensional PSF data in the direction of any one of the axes to form a plurality of axes, each of which corresponds to a plurality of planes of an axis in which an image is distorted and an axis in which an image is not distorted Obtaining a two-dimensional PSF data set; d) selectively obtaining a PSF movement mapping map according to a slope of a line pattern composed of a plurality of pixels in each of the plurality of two-dimensional PSF data sets; And e) correcting distortion of the EPI image data using the PSF motion mapping map.

There is provided a computer readable recording medium storing a program for performing a distortion correction method of an EPI (Echo Planar imaging) image in a magnetic resonance imaging apparatus according to another embodiment of the present invention, the method comprising the steps of: a) Repeatedly to form EPI image data from a target object; b) Point Spread Function (PSF) data from the object by repeatedly applying a predetermined PSF pulse sequence, wherein the 3D PSF image data is defined in a three-dimensional space having axes, ; c) a step of integrating the three-dimensional PSF data in the direction of any one of the axes to form a plurality of axes, each of which corresponds to a plurality of planes of an axis in which an image is distorted and an axis in which an image is not distorted Obtaining a two-dimensional PSF data set; d) selectively obtaining a PSF movement mapping map according to a slope of a line pattern composed of a plurality of pixels in each of the plurality of two-dimensional PSF data sets; And e) correcting distortion of the EPI image data using the PSF motion mapping map.

According to the present invention, more accurate clinical results can be obtained by correcting distortion in an EPI image using an optional PSF (Point Spread Function) mapping method for an EPI image acquired by an MRI apparatus.

1 is a block diagram showing an MRI apparatus according to the present invention.
2 is an exemplary diagram of an EPI pulse sequence according to an embodiment of the present invention.
3 is an illustration of a PSF pulse sequence for acquiring 3D PSF data according to an embodiment of the present invention.
FIG. 4A is a diagram schematically illustrating three-dimensional spatial data formed by filling plane data acquired according to a PSF sequence according to an embodiment of the present invention into a three-dimensional space.
4B is a diagram schematically showing three-dimensional image data (a plurality of EPI images) composed of a series of distorted xy plane images obtained by integrating three-dimensional spatial data in the s direction.
4C is a diagram schematically showing three-dimensional image data (a plurality of gradient echo images) composed of a series of undistorted xs plane images obtained by integrating three-dimensional spatial data in the y direction.
4D is a diagram schematically showing two-dimensional PSF data having a line pattern obtained by integrating three-dimensional spatial data in the x direction.
5 is a flowchart illustrating a method for removing PSF ghost artifacts from two-dimensional PSF data in accordance with an embodiment of the present invention.
6 is an exemplary diagram illustrating a method for removing PSF ghost artifacts from two-dimensional PSF data in accordance with an embodiment of the present invention.
7 is an exemplary diagram showing an example in which two-dimensional PSF data is blurred.
8 is a diagram showing an example in which the two-dimensional PSF data is out of the reference line.
9 is a flowchart illustrating a method for selectively obtaining a PSF movement map according to an embodiment of the present invention.
FIGS. 10A to 10E are views illustrating a method of selectively obtaining a PSF movement map according to an embodiment of the present invention.

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

1 is a block diagram showing an MRI apparatus according to the present invention. As shown in FIG. 1, the MRI apparatus 100 of the present invention includes a coil unit 110. The coil section 110 includes a cylindrical housing formed with a central bore. In the housing, a main magnetic field coil, a gradient coil, and a radio frequency (RF) coil may be provided. The coil portion 110 may further include a conveying means 112 for supporting the object 10 and capable of conveying the object 10 to the hole of the coil portion 110. [

The main magnetic field coil generates a magnetostatic field in a direction parallel to the axial direction of the object in the inner space of the coil part 110. That is, a horizontal magnetic field is generated in the inner space of the coil part 110. The gradient coil generates a gradient magnetic field to give a gradient to the intensity of the static field. In an embodiment, a gradient magnetic field, a phase encoding gradient magnetic field, a phase shift gradient magnetic field, a readout gradient magnetic field, and the like are included in the gradient magnetic field for forming an echo planar imaging (EPI) image in an MRI apparatus have. The gradient coil part 110 may be provided with gradient coils for generating such gradient magnetic fields.

The RF coil excites cells inside the object 10 in a static magnetic field space to generate a high frequency magnetic field for generating a spin. In addition, the RF coil receives electromagnetic waves generated by excited spins, that is, magnetic resonance signals.

The MRI apparatus 100 includes an RF driving unit 120 and a gradient driving unit 130 connected to the coil unit 110. In detail, the gradient driving unit 130 is connected to a plurality of gradient coils of the coil unit 110, and the RF coil is connected to the RF driving unit 120.

The RF driving unit 120 generates an RF pulse signal that is a driving signal of the RF coil. The generated RF pulse signal is applied to the RF coil to excite cells inside the target body 10 to generate a spin.

The gradient driving unit 130 generates a gradient pulse signal in which the gradient coil is a driving signal. The gradient pulse signal may be applied to the gradient coil of the coil section 110 to generate a gradient magnetic field. The gradient driving unit 130 may include driving circuits corresponding to the gradient coils provided in the coil unit 110, respectively.

The RF driving unit 120, the gradient driving unit 130, and the data obtaining unit 150 are connected to the controller 130. The control unit 130 controls the RF driving unit 120, the gradient driving unit 130, and the data obtaining unit 150 to obtain an MRI image.

The MRI apparatus 100 further includes a data acquiring unit 150 connected to the coil unit 110. The data acquisition unit 150 acquires the EPI image data by collecting the received signals received by the RF coil.

The MRI apparatus 100 further includes a data processing unit 160 connected to the data acquiring unit 150. The data processing unit 160 may include a memory to store a program for data processing. In addition, the data acquired by the data acquisition unit 150 may be stored in a memory to form a data space. The data space constitutes a three-dimensional Fourier space. The data processing unit 160 performs three-dimensional inverse Fourier transform on the data in the three-dimensional Fourier space to form EPI image data. In the following, the three-dimensional Fourier space is referred to as k-space. In an embodiment, the data processing unit 160 performs a function of correcting distortion in an EPI image, and a description thereof will be described later in more detail. The EPI image data whose distortion has been corrected by the data processing unit 160 is transmitted to the display unit 170 and displayed.

FIG. 2 shows an EPI sequence for obtaining an EPI image in an MRI apparatus according to an embodiment. Pulses in the EPI sequence 200 shown in FIG. 2 are generated by the RF driver 120 and the gradient driver 130 under the control of the controller 140 and supplied to the coil unit 110. 2, the EPI sequence 200 includes an RF excitation pulse 202 applied to the RF coil in the RF driver 120 and a plurality of gradient pulses 202 applied to the plurality of gradient coils in the gradient driver 130, . In one embodiment, the RF excitation pulse 202 is applied to the RF coil of the coil section 110 to be a magnetic field in a direction perpendicular to a plane (hereinafter, image plane) from which an image is to be obtained at the object, and a plurality of gradient pulses So that a magnetic field is generated along each axis on a coordinate system formed by three axes orthogonal to each other with respect to the image plane. Here, the three axes may include a first axis Gz perpendicular to the image plane, and a second axis Gy and a third axis Gx parallel to the image plane and orthogonal to each other.

Referring to FIG. 2, a slice selection magnetic field gradient pulse 204 for selectively exciting a slice along a first axis Gz direction in an object to which an EPI image is to be formed according to an EPI sequence 200 And is applied to the coil portion 110. In one embodiment, the slice selection magnetic field gradient pulse 204 may be applied to the coil section 110 simultaneously with the RF excitation pulse 202.

In addition, phase encoding gradient pulses 206 and 210 for generating a phase encoding gradient magnetic field along the second axis Gy may be applied to the coil portion 110 according to the EPI sequence 200. The cells can be excited by the magnetic field generated in response to the application of the RF excitation pulse 202 to spatially encode the generated spin. As shown in FIG. 2, the first phase encoding gradient pulse 206 may be applied at certain time intervals after the RF excitation pulse 202 and the slice select magnetic field gradient pulse 204 are applied. Also, the second phase encoding gradient pulse 210 may be applied at certain time intervals after the first phase encoding gradient 206 is applied. In one embodiment, the second phase encoding gradient 210 may comprise an EPI blip with a constant periodicity.

2, a phase shift gradient pulse 208 and a lead-out gradient pulse 212 are generated to generate a gradient magnetic field along the third axis Gx direction in accordance with the EPI sequence 200 . A phase shift gradient pulse 208 is applied to shift the phase of the phase encoded spin and a readout gradient pulse 212 is applied to generate a gradient echo magnetic resonance by repositioning the spins. Phase shift gradient pulse 208 and lead out gradient pulse 212 may be applied at regular time intervals after the RF excitation pulse 202 and slice selection magnetic field gradient pulse 204 are applied. In an embodiment, the lead-out gradient pulse 212 may be repeatedly applied for a period of time in the form of a pulse waveform having a periodicity after the phase shift gradient pulse 208 is applied.

In an embodiment, a phase encoding gradient 206 pulse and a phase shift gradient pulse 208 may be applied simultaneously. In addition, the second phase encoding gradient pulse 210 and the readout gradient pulse 212 may be applied simultaneously in the same period. The slice data obtained by applying the second phase encoding gradient pulse 210 and the readout gradient pulse 212 according to the EPI sequence 200 is filled in k-space so that the EPI image can be acquired in the data processing unit 160 have.

2, a first axis Gz perpendicular to an imaging plane of a target object, and a second axis Gy parallel to the image plane and orthogonal to each other, and a third axis Gx, The spatially separated gradient magnetic field applied through the gradient coil of the MRI apparatus is not limited to such a coordinate system but may be modified into various embodiments having different coordinate systems. In addition, the waveform of each gradient in the EPI sequence and the time point at which the gradient pulse is applied are not limited to those shown in FIG. 2, and can be variously modified.

According to an embodiment of the present invention, in order to obtain a PSF (Point Spread Function) mapping map for correcting the distortion of the EPI image, the controller 140 sets the RF excitation pulse and the plurality of gradient pulses to a predetermined PSF sequence . The data obtained by generating the magnetic field by the PSF sequence is collected in the data processing unit 160 to form the three-dimensional PSF data.

FIG. 3 shows a PSF pulse sequence for acquiring 3D PSF data according to an embodiment of the present invention. The PSF sequence 300 shown in FIG. 3 includes a phase encoding pre-phase gradient encoding (EPR) sequence for generating a gradient magnetic field in the phase encoding direction, i.e., the second axis Gy direction, to the EPI sequence 200 shown in FIG. and applying prewinder gradients: Gs pulses 302. [ Thus, the PSF sequence 300 includes an RF excitation pulse 202, similar to the EPI sequence 200 shown in FIG. 2, and includes a slice selection magnetic field gradient applied to generate a gradient magnetic field along the first axis Gz Phase shift gradient pulses 206 and 210 applied to generate a gradient magnetic field along the second axis Gy and a phase shift applied to generate a gradient magnetic field along the third axis Gx. A gradient pulse 208 and a lead-out gradient pulse 212.

 In an embodiment, an encoding pre-wind gradient pulse 302 may be applied before applying a phase encoding gradient pulse 206, 210. In this manner, the PSF pulse sequence 300 can be repeatedly performed a predetermined number of times while changing the intensity of the encoding pre-wind gradient pulse 302. For example, it is possible to repeatedly execute the PSF pulse sequence 300 while changing the intensity of the pulse of the encoding pre-and-gradient pulse 302 constantly from -Vy to + Vy. By doing so, the data corresponding to the slice selected by the magnetic field generated in the first axis Gz in response to the slice selection magnetic field gradient pulse 204, i.e., the kx-ky plane data in the k space, Can be obtained.

It should be noted that the PSF sequence 300 shown in FIG. 3 is only one example, and the PSF sequence 300 may be modified in other forms. For example, the PSF sequence 300 may be constructed based on an EPI sequence of a different type than the EPI sequence 200 shown in FIG.

As shown in FIG. 4A, it is possible to repeatedly execute the PSF pulse sequence 300 to acquire a plurality of kx-ky plane data 410_1 to 410 - n. Dimensional k-space data 400 can be obtained by accumulating a plurality of kx-ky plane data 410_1 to 410_n in the k space in the ks direction. As described above, each kx-ky plane data can be obtained by sequentially changing the size of the phase encoding pre-winder gradient pulse Gs of the PSF sequence 300. [ The three-dimensional PSF data can be formed by three-dimensionally Fourier transforming the three-dimensional k-space data 420.

Hereinafter, a method of correcting distortion in the EPI image using the three-dimensional PSF data in the data processing unit 160 according to an embodiment of the present invention will be described in detail.

The data processing unit 160 integrates the three-dimensional PSF data formed on the three-dimensional space (x, y, s) in the s direction to generate three-dimensional PSF data consisting of a series of distorted xy plane images, Image data (a plurality of EPI images) 440 can be obtained. In addition, the data processing unit 160 integrates the three-dimensional PSF data in the y direction to generate three-dimensional image data (a plurality of gradient echoes) composed of a series of undistorted xs plane images as shown in Fig. GE) image) 460 can be obtained. GE images have little distortion, but have problems that take too long to acquire clinically available resolution data (for example, about 5 minutes on a 7T MRI). On the other hand, in the case of an EPI image, it can be acquired in about 2 to 3 seconds.

The data processing unit 160 may integrate the three-dimensional PSF data in the x-axis direction to obtain two-dimensional PSF data corresponding to the s-y plane. 4D shows two-dimensional PSF data corresponding to a plurality of s-y planes obtained by integrating the three-dimensional PSF data 480 along the x-axis. As shown in FIG. 4D, when the three-dimensional PSF data is integrated in the x direction, the two-dimensional PSF data appears as a correlated line pattern on the s-y plane.

On the other hand, the data processor 160 performs three-dimensional Fourier transform on the three-dimensional k-space data 420. As a result, the three-dimensional PSF data may include PSF ghost artifacts appearing in parallel in the s direction and the main image (two-dimensional PSF data of the line pattern) in the s-y plane image. If the slice data is acquired by gradually shifting in the s direction in order to acquire the three-dimensional PSF data, and the k space is filled, the center of the k space also moves, and the phase slightly differs. When the acquired k spatial data is subjected to three-dimensional Fourier transform, a PSF ghost artifact is generated. These PSF ghost artifacts may be strongly affected by pulsatile flow or strong local magnetic field non-uniformity, and may cause errors in obtaining the PSF motion mapping map, thus eliminating PSF ghost artifacts.

Generally, PSF ghost artifacts may have a signal strength smaller than the PSF data of the line pattern, which is always the main signal in the s direction, but may be larger than the signal strength of the PSF data in the y direction. In this case, when a pixel shift map is obtained in the y direction perpendicular to the PSF data signal for image correction by the PSF movement mapping method, an error may occur in the pixel movement map due to the PSF ghost artifact . Therefore, in order to correct the correct EPI image, PSF ghost artifacts should be removed from the two-dimensional PSF data. Hereinafter, a method of removing PSF ghost artifacts according to an embodiment will be described in detail with reference to FIGS. 5 and 6. FIG.

5 is a flowchart illustrating a method for removing PSF ghost artifacts from two-dimensional PSF data in a data processing unit 160 according to an embodiment. First, the two-dimensional PSF data is acquired from the three-dimensional PSF data (S510). Since the two-dimensional PSF data may include noise as shown in FIG. 6A, the two-dimensional PSF data is filtered using a three-dimensional anisotropic diffusion filter (S520). Three-dimensional anisotropic diffusion filtering removes noise from the two-dimensional PSF data, as well as signal enhancement, data smoothing, and the like. Through the 3D anisotropic diffusion filtering, two-dimensional PSF data with noise removed and smoothed can be obtained as shown in the image of FIG. 6 (b).

As shown in the image (c) of FIG. 6, when the first derivative is performed in the s direction with respect to each y value 670 with respect to the filtered two-dimensional PSF data, the first derivative Value graph can be obtained (S530). In the embodiment, the position (maximum point) having the maximum value and the position (minimum point) having the minimum value in the first differential value graph shown in the image (c) of FIG. 6 are selected and the width between the maximum point and the minimum point is And sets it as a selected band (S540). The reason why the maximum and minimum points are selected in the first differential value graph is that the first differential value graph shows a profile in which the maximum value and the minimum value appear on the left and right sides based on the two-dimensional PSF data as the main signal. Thus, PSF ghost artifacts can be eliminated by setting the selection band for all y values and selecting only the data contained in the set selection band. The d image in Fig. 6 shows an example in which PSF ghost artifacts are removed from the two-dimensional PSF data. Thus, ghost artifacts can be effectively removed from the two-dimensional PSF data by performing PSF ghost artifact removal repeatedly for each frequency encoding step (x direction) along the y direction. In one embodiment, the PSF ghost artifact can generate the final pixel shift map using the eliminated two-dimensional SF data.

On the other hand, as described above, the three-dimensional PSF data is obtained by repeatedly performing the PSF sequence while changing the phase encoding pre-wind gradient pulse 302 to a predetermined magnitude. Accordingly, when distortion does not occur in the three-dimensional PSF data, ideally, the two-dimensional PSF data pixels are arranged along an arbitrary straight line (hereinafter referred to as a reference line) on the s-y plane. Here, the reference line may be a diagonal line passing through the origin of the s-y coordinate system with a slope "1 " in the s-y plane.

However, when distortion occurs due to non-uniformity of the magnetic field, local magnetic susceptibility, or the like during 3D PSF data acquisition, the 2D PSF data pixels may be arranged out of the reference line. For example, as shown in FIG. 7, it may be composed of pixels arranged along the reference line 710 in the s-y plane among the two-dimensional PSF data pixels and pixels arranged out of the reference line 710.

The pixels out of the reference line 710 are due to image distortion. The distorted area in the EPI image is contracted due to the non-uniformity of the strong magnetic field and the local magnetic susceptibility, and the distorted area (hereinafter referred to as the shrink area) (Hereinafter referred to as an extension region). In the contraction region, the slope of the line in which the two-dimensional PSF data pixels are arranged is relatively large, and in the extension region, the slope of the line in which the PSF data pixels are arranged is comparatively gentle.

As shown in FIG. 8, the pixels out of the reference line can be divided into pixels included in the contraction region 810 and pixels included in the extension region 820. In one embodiment, the PSF pixel shift map for the contraction region and the PSF pixel shift map for the extension region can be obtained in different directions. Here, the pixel shift map can be calculated by a deviation from a two-dimensional PSF data pixel and a deviation from a reference line, that is, a distance between a PSF data pixel and a reference line. Specifically, the PSF pixel shift map can be obtained in the s direction, which is the distorted coordinate, or the y direction, which is the undistorted coordinate, and the direction for obtaining the PSF pixel shift map can be selectively determined according to the distortion type of the image.

In the contraction region, it is possible to obtain more information than the undistorted coordinates, that is, obtaining the pixel shift map in the y direction is more distorted than the pixel shift map in the s direction. On the other hand, in the extension region, more information can be obtained by obtaining the pixel shift map in the s direction. Therefore, it is possible to correct the geometric distortion more accurately for the EPI image by obtaining the pixel shift map in the y direction and the s direction according to the shrinkage region and the extension region, respectively.

9 is a flowchart illustrating a process of selectively computing a PSF movement map according to an embodiment of the present invention. Referring to FIG. 9, pixels having the maximum signal intensity are selected from s-values along the reference line in the two-dimensional PSF data (S910). In the case of the two-dimensional PSF data from which the PSF ghost artifact is removed, as shown in the image shown in FIG. 10A, a plurality of two-dimensional PSF data pixels may be distributed and distributed for each s value due to the blurring phenomenon. In order to obtain an accurate PSF mapping map from the blurred PSF data pixels, in the embodiment of the present invention, among the blurred PSF data pixels, only the signal with the strongest signal intensity in the vertical direction (y direction) Thereby removing blurred pixels. FIG. 10B shows the PSF data of the sharp line pattern from which the blurred pixels have been removed. This process is performed for all x values, and then two-dimensional polynomial fitting is performed (S920) to obtain continuous PSF data in a three-dimensional space as shown in FIG. 10C. PSF data can be visually classified into shrinkage area, extension area, and the like through continuous PSF data on a three-dimensional space.

Subsequently, data corresponding to an arbitrary x value 1010 in the PSF data on the three-dimensional space shown in Fig. 10C can be displayed as a PSF data line as shown in Fig. 10D. When the first derivative is applied to the PSF data line thus selected (S930), the first derivative value graph shown in FIG. 10E can be obtained. The primary differential value in FIG. 10E represents the slope of the PSF data line. In an embodiment of the present invention, if the absolute value of the first order differential value is greater than 1, a pixel shift map is obtained in the y direction (1010, 1030) corresponding to the extension region, 1, the pixel shift map is obtained in the s-direction by the shrinkage area 1020 (S940). In this way, a pixel shift map can be obtained for all x values, and a y direction pixel shift map and an s direction pixel shift map can be combined to form a PSF pixel shift map. The data processing unit 160 may correct the distortion of the EPI image acquired according to the EPI sequence using the PSF pixel shift map thus formed, and finally form the EPI image with the distortion corrected.

In another embodiment of the present invention, a computer-readable recording medium storing a program for performing a distortion correction method of an EPI (Echo Planar imaging) image in a magnetic resonance imaging apparatus can be provided. Here, the method of correcting the distortion of the EPI image includes the steps of repeatedly applying a predetermined EPI pulse sequence to form EPI image data from a target object; (PSF) image data is defined on a three-dimensional space having axes by repeatedly applying a predetermined PSF (Pulse Sequence) pulse sequence to form three-dimensional PSF (Point Spread Function) data from the object ; c) a step of integrating the three-dimensional PSF data in the direction of any one of the axes to form a plurality of axes, each of which corresponds to a plurality of planes of an axis in which an image is distorted and an axis in which an image is not distorted Obtaining a two-dimensional PSF data set; d) selectively obtaining a PSF movement mapping map according to a slope of a line pattern composed of a plurality of pixels in each of the plurality of two-dimensional PSF data sets; And e) correcting distortion of the EPI image data using the PSF motion mapping map.

It should be understood that the above-described embodiments are merely illustrative of some of the various embodiments that have applied the principles of the present invention. It will be apparent to those skilled in the art that various modifications may be made without departing from the spirit of the present invention.

200 EPI pulse sequence
202 RF pulse pulse
204 Slice selection magnetic field gradient pulse
206 First Phase Encoding Gradient Pulse
210 Second Phase Encoding Gradient Pulse
208 Phase Shift Gradient Pulse
212 Lead Out Gradient Pulse
300 PSF pulse sequence
302 Phase Encoding Pre-Wind Gradient Pulse
810 kidney area
820 Compression Area

Claims (18)

In the MRI apparatus,
A driver operative to generate a predetermined EPI (Echo Planar imaging) pulse sequence;
A coil portion including a plurality of coils, wherein the coil portion generates a first magnetic field in the object in response to a pulse sequence from the driving portion when the object is located near the coil portion, Operating to cause a first magnetic resonance signal to be induced; And
And a data processing section operable to form EPI image data based on a first magnetic resonance signal induced in the coil section,
The driving unit is operable to further generate a predetermined PSF (Point Spread Function) pulse sequence,
Wherein the coil unit generates a second magnetic field in response to the PSF pulse sequence from the driving unit when the object is located near the coil part and generates a second magnetic resonance signal in response to the generated second magnetic field, Further,
Wherein the data processing unit is further operative to form three-dimensional PSF image data based on the second magnetic resonance signal, the three-dimensional PSF image data being defined on a three-dimensional space having axes,
Wherein the data processing unit integrates the three-dimensional PSF data in a direction of any one of the axes and is arranged perpendicularly to the one axis and corresponds to a plurality of planes of an axis in which an image is distorted and an axis in which an image is not distorted A second processing unit for obtaining a PSF movement mapping map selectively according to a slope of a line pattern composed of a plurality of pixels in each of the plurality of two-dimensional PSF data sets, And a third processing unit for correcting the distortion of the EPI image data using the PSF movement mapping map. 2. The MRI apparatus of claim 1, wherein the data processing unit is further operative to remove PSF ghost artifacts from the two-dimensional PSF data. 3. The apparatus according to claim 2, wherein the data processing unit first differentiates the two-dimensional PSF data in the direction of the distorted axis, sets a selection band based on the first-differentiated result value, And to select the two-dimensional PSF data contained in the band to remove PSF ghost artifacts. 4. The MRI apparatus according to claim 3, wherein the selection band is determined as a maximum value and a minimum value in the first differentiated result value. 5. The MRI apparatus according to claim 4, wherein the data processing section is operative to perform three-dimensional anisotropic filtering on the two-dimensional PSF data before the first differentiation to remove noise. 6. The image processing apparatus according to any one of claims 1 to 5, wherein the data processing unit sets a reference line to two-dimensional PSF data of the line pattern, and calculates, based on the reference line, Selects a pixel having the maximum signal intensity, first differentiates the line data connecting the selected pixels with each other, and selectively obtains a PSF movement mapping map according to the first derivative result value. 7. The MRI apparatus according to claim 6, wherein the reference line is a diagonal line passing through the origin of the distorted axis and the undistorted axis. 7. The method according to claim 6, wherein the PSF movement mapping map obtains a PSF motion mapping map with a distorted axis when the first-order differential result value is equal to or greater than 1, and when the first-order differential result value is less than 0 to 1, MRI apparatus for obtaining a PSF movement mapping map. A method for correcting distortion of an echo planar imaging (EPI) image in an MRI apparatus,
a) repeatedly applying a predetermined EPI pulse sequence to form EPI image data from a subject;
b) Point Spread Function (PSF) data from the object by repeatedly applying a predetermined PSF pulse sequence, wherein the 3D PSF image data is defined in a three-dimensional space having axes, ;
c) a step of integrating the three-dimensional PSF data in the direction of any one of the axes to form a plurality of axes, each of which corresponds to a plurality of planes of an axis in which an image is distorted and an axis in which an image is not distorted Obtaining a two-dimensional PSF data set;
d) selectively obtaining a PSF movement mapping map according to a slope of a line pattern composed of a plurality of pixels in each of the plurality of two-dimensional PSF data sets; And
e) correcting the distortion of the EPI image data using the PSF motion mapping map
And correcting the distortion of the EPI image. 10. The method according to claim 9, wherein the three-dimensional Fourier transform is performed on the three-dimensional spatial data obtained by repeatedly executing the predetermined PSF sequence to obtain the three-dimensional PSF data. 11. The method of claim 10, further comprising removing PSF ghost artifacts from the two-dimensional PSF data. 12. The method of claim 11, wherein removing the PSF ghost artifact comprises:
First-differentiating the two-dimensional PSF data in a direction of a distorted axis;
Setting a selection band based on the first differentiated result value; And
Selecting two-dimensional PSF data included in the selected band from the two-dimensional PSF data and removing the PSF ghost artifact
And correcting the distortion of the EPI image. 13. The method of claim 12, wherein the selected band is determined as a maximum value and a minimum value in the first differentiated resultant value. 14. The method of claim 13, further comprising performing three-dimensional anisotropic filtering on the two-dimensional PSF data before the first differentiation to remove noise. 15. The method according to any one of claims 9 to 14, wherein step d)
Setting a reference line of a straight line to the two-dimensional PSF data of the line pattern;
Selecting pixels of the maximum signal intensity at each value of the distorted axis in the two-dimensional PSF data along the baseline;
Performing first differentiation on line data connecting the selected pixels; And
And selectively obtaining a PSF movement mapping map according to the first derivative result value
And correcting the distortion of the EPI image. 16. The method of claim 15, wherein the reference line is a diagonal line passing through the origin of the distorted axis and the undistorted axis. 17. The method according to claim 16, wherein the PSF movement mapping map is obtained by obtaining a PSF movement mapping map with a distorted axis when the first-order differential result value is equal to or greater than 1 and if the first-order differential result value is less than 0 to 1, A method for correcting distortion of an EPI image, which obtains a PSF motion mapping map. A computer readable recording medium storing a program for performing a distortion correction method of an EPI (Echo Planar imaging) image in a magnetic resonance imaging apparatus,
a) repeatedly applying a predetermined EPI pulse sequence to form EPI image data from a subject;
b) Point Spread Function (PSF) data from the object by repeatedly applying a predetermined PSF pulse sequence, wherein the 3D PSF image data is defined in a three-dimensional space having axes, ;
c) a step of integrating the three-dimensional PSF data in the direction of any one of the axes to form a plurality of axes, each of which corresponds to a plurality of planes of an axis in which an image is distorted and an axis in which an image is not distorted Obtaining a two-dimensional PSF data set;
d) selectively obtaining a PSF movement mapping map according to a slope of a line pattern composed of a plurality of pixels in each of the plurality of two-dimensional PSF data sets; And
e) correcting the distortion of the EPI image data using the PSF motion mapping map
And a computer-readable recording medium storing a program for performing the method.

Images