JP3384897B2 - MRI equipment - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to MR (Magnetic Res
onance) imaging method and MRI (Magnetic Res
onance Imaging) device. More specifically, the present invention relates to an MR imaging method and an MRI apparatus capable of imaging a moving subject and suppressing an artifact.

[0002]

2. Description of the Related Art FIG. 14 shows a conceptual diagram of the configuration of an example of a conventional MRI apparatus. In this MRI apparatus 500, 50
Is an imaging device, 50a is a bore, and 60 is a cradle for moving the subject G. A movement stop control unit 51 stops the cradle 60 during image pickup. This MRI device 50
At 0, the movement of the subject G is stopped and an image is taken to obtain a first image having a length T in the movement direction. Next, after moving by the distance T, the image is stopped and imaged again to obtain a second image having a length T in the moving direction. This is repeated n times to obtain the first to n-th
Up to n images are stitched together to obtain an image of length nT. The MRI apparatus 500 described above is disclosed in, for example, Japanese Patent Laid-Open No. 63-2
No. 72335.

[0003]

In the above-mentioned conventional MRI apparatus 500, the cradle 60 is stopped when the image pickup apparatus 50 takes an image, and the image pickup apparatus 50 is stopped when the subject G is moved by the cradle 60. There is. Therefore, there is a problem that the imaging efficiency is not good. Therefore, it is an object of the present invention to provide an MR imaging method and an MRI apparatus capable of imaging a moving subject and suppressing artifacts.

[0004]

According to a first aspect of the present invention, the frequency axis data number of k-space is f (-N / 2≤).
f ≦ N / 2-1) and the phase axis data number is p (-M)
/ 2 ≦ p ≦ M / 2−1), an RF pulse is applied, phase encoding is applied and frequency encoding is applied, and imaging data D ′ is generated from the generated echo.
By repeatedly collecting (f, p), M views of imaging data D '(f, -M / 2) to D'.
In the MR imaging method of collecting (f, M / 2-1), at least two echoes are generated for one RF pulse, the phase encode amount is set to “0” for one echo, and the frequency axis While applying the read-out gradient of the sampling number N'to the data number f '(-N' /
2 ≦ f ′ ≦ N ′ / 2−1) navigation data SRp (f ′) in the frequency axis direction in the p view
, And phase encoding and frequency encoding according to the p-view of k-space are added to the remaining echoes, whereby the imaging data D ′ (f, p) of the p-view of k-space is added. By repeating the collection, the imaging data D ′ (f, −M / 2) for M views
D ′ (f, M / 2-1) is collected, and the phase correction amount Δφx (is obtained from the mutual relation of the navigation data SR _{−M / 2} (f ′) to SRM _{/ 2-1} (f ′) of each view. p) is calculated, and the phase correction amount Δφx (p) is used to phase-correct the imaging data D ′ (f, p) of each view.
An imaging method is provided.

According to a second aspect of the present invention, the frequency axis data number of k-space is f (-N / 2≤f≤N / 2-1).
And the phase axis data number is p (-M / 2 ≦ p ≦ M / 2
-1), RF pulse is applied, phase encoding is applied, frequency encoding is applied, and imaging data D ′ (f, p) is collected from the generated echo, and imaging for M views is repeated. In the MR imaging method of collecting data D '(f, -M / 2) to D' (f, M / 2-1) for use, at least two echoes are generated for one RF pulse, and one echo is generated. , The frequency encode amount is set to “0”, and the data number p ′ (− M ′ / 2 ≦ p ′ ≦ M ′ / 2− is applied while applying the reading gradient of the sampling number M ′ to the phase axis.
1) Collect navigation data SWp (p ′) in the phase axis direction in the p-view of 1), add phase encoding corresponding to the p-view of k-space and frequency encode to the remaining echoes, and The image data D ′ (f, p) for p-views in the k-space is repeatedly collected according to the above, and the imaging data D ′ (f, −M / 2) to D ′ (f, M / 2-1)
Data for each view navigation data SW _{-M / 2}
The phase correction amount Δφy (p) is calculated from the mutual relationship of (p ′) to SW _{M / 2-1} (p ′), and the imaging data D ′ (f, f, f of each view is calculated with the phase correction amount Δφy (p). There is provided an MR imaging method characterized in that p) is phase-corrected.

According to a third aspect of the present invention, the frequency axis data number of k-space is f (-N / 2≤f≤N / 2-1).
And the phase axis data number is p (-M / 2 ≦ p ≦ M / 2
-1), RF pulse is applied, phase encoding is applied, frequency encoding is applied, and imaging data D ′ (f, p) is collected from the generated echo, and imaging for M views is repeated. In an MRI apparatus that collects data D '(f, -M / 2) to D' (f, M / 2-1) for use, at least two echoes are generated for one RF pulse, and one echo is generated. On the other hand, the phase encoding amount is set to “0”, and the reading gradient of the sampling number N ′ is applied to the frequency axis, and the p of the data number f ′ (−N ′ / 2 ≦ f ′ ≦ N ′ / 2−1) is set. Navigation data SRp (f ') in the frequency axis direction in the view is collected, and k is applied to the remaining echoes.
-Phase-encoding and frequency-encoding according to the p-view of the space are added, whereby the collection of the imaging data D '(f, p) of the p-view of the k-space is repeated, and for the M-view imaging. Data D '
Data collecting means for collecting (f, -M / 2) to D '(f, M / 2-1), and navigation data SR- _{M / 2} (f') to SRM _{/ 2-1} for each view. The phase correction amount Δφx (p) is calculated from the mutual relationship of (f ′), and the phase correction amount Δφx is calculated.
In (p), the imaging data D ′ (f, f,
An MRI apparatus comprising: a phase correction unit that corrects the phase p).

According to a fourth aspect of the present invention, the frequency axis data number of the k-space is f (-N / 2≤f≤N / 2-1).
And the phase axis data number is p (-M / 2 ≦ p ≦ M / 2
-1), RF pulse is applied, phase encoding is applied, frequency encoding is applied, and imaging data D ′ (f, p) is collected from the generated echo, and imaging for M views is repeated. In an MRI apparatus that collects data D '(f, -M / 2) to D' (f, M / 2-1) for use, at least two echoes are generated for one RF pulse, and one echo is generated. On the other hand, the frequency encode amount is set to “0”, and the data number p ′ (−M ′ / 2 ≦ p ′ ≦ M ′ / 2−1) p is applied while applying the read gradient of the sampling number M ′ to the phase axis. The navigation data SWp (p ′) in the phase axis direction in the view is collected, and k− for the remaining echoes.
Phase encoding according to the p-view of the space and frequency encoding are added, whereby the collection of the imaging data D ′ (f, p) of the p-view of the k-space is repeated, and the imaging data of M views is repeated. D '
Data collecting means for collecting (f, -M / 2) to D '(f, M / 2-1) and navigation data SW _{-M / 2} (p') to SW _{M / 2-1} for each view. The phase correction amount Δφy (p) is calculated from the mutual relation of (p ′), and the phase correction amount Δφy is calculated.
In (p), the imaging data D ′ (f, f,
An MRI apparatus comprising: a phase correction unit that corrects the phase p).

In a fifth aspect, the present invention sets the encoding amount of the first encoding axis of k-space to "0", and k-
Data collecting means for collecting the navigation data S1 while applying the read gradient to the second encode axis of the space, and then collecting the navigation data S2 in the same manner, and a one-dimensional Fourier transform of the navigation data S1. From the cross-correlation between the projection profile obtained from the absolute value and the projection profile obtained from the absolute value of the one-dimensional Fourier transform of the navigation data S2, the movement amount Δ of the signal source in the direction of the real space corresponding to the second encode axis is calculated. An MR imaging method is provided, which comprises a moving amount calculating means for calculating.

[0009]

As shown in FIG. 12, -Fw / 2≤x≤Fw /
When there is a signal source g (x, y) in the real space (x, y) of 2, −Pw / 2 ≦ y ≦ Pw / 2, the acquired data D (f, p) in k-space is It becomes like a formula. However, the x-axis direction is the frequency axis direction, and the y-axis direction is the phase axis direction (note that the y-axis and the x-axis are axes of the real space defined corresponding to either the phase axis or the frequency axis). On the other hand, the Z-axis, Y-axis, and X-axis of the gradient magnetic field coil, which will be described later, are the axes of the real space defined as the structure of the magnet assembly 1).

[0010]

[Equation 1]

Next, as shown in FIG. 13, the signal source g
Assuming that (x, y) moves in the x-axis direction by Δx, the collected data D ′ (f, p) in k-space is
It becomes like the following formula.

[0012]

[Equation 2]

Here, if X = x-Δx, (Equation 2)
The formula is as follows.

[0014]

[Equation 3]

From (Equation 3), the phase correction is performed on the data D ′ (f, p) as follows: D (f, p) = exp {j2πfΔx / Fw} × D ′ (f, p) By performing the two-dimensional Fourier transform, as shown in FIG. 13, it can be seen that imaging corresponding to the case where the signal source g (x, y) does not move can be obtained. Here, if the movement amount Δx is different for each view, using the movement amount Δxp for each view, all of D (f, p) = exp {j2πfΔxp / Fw} × D '(f, p) (4) If the signal source g (x, y) is not moving, if D '(f, p) of the view is phase-corrected and the obtained D (f, p) is subjected to two-dimensional Fourier transform. A corresponding image is obtained. That is, a moving subject can be imaged and artifacts can be suppressed and imaged.

The signal source g (x, y) is Δy in the y-axis direction.
The same applies to the case of moving by p, and D (f, p) = exp {j2πpΔyp / Pw} × D '(f, p) (5) By performing phase correction and performing two-dimensional Fourier transform on the obtained D (f, p), an image corresponding to the case where the signal source g (x, y) does not move can be obtained.

Further, when the signal source g (x, y) moves by Δxp in the x-axis direction and by Δyp in the y-axis direction, the same operation as above is performed, and D (f, p) = exp {j2π (FΔxp / Fw + pΔyp / Pw)} × D ′ (f, p) (6) Phase-corrects D ′ (f, p) of all views by (6), and the obtained D (f, p) is two-dimensional. By performing the Fourier transform, an image corresponding to the case where the signal source g (x, y) does not move can be obtained.

Now, in the MR imaging method according to the first aspect and the MRI apparatus according to the third aspect,
At least two echoes are generated for one RF pulse. For one of the echoes, the phase encode amount is set to “0” and the sampling number N ′ is set on the frequency axis.
While applying the readout gradient of, the data number f '(-
N ′ / 2 ≦ f ′ ≦ N ′ / 2−1) p-view navigation data SRp in the frequency axis direction
Collect (f '). And for the rest of the echoes,
Phase encoding and frequency encoding according to the p-view of the k-space are added, whereby the imaging data D ′ (f, p) of the p-view of the k-space is collected in order. Then, the navigation data SRp of each view
The phase correction amount Δφx (p) is calculated from the mutual relation of (f ′), and the imaging data D ′ (f, p) of each view is phase corrected by the phase correction amount Δφx (p). Since the navigation data SRp (f ′) is always collected with the phase encode amount “0”, the movement amount Δxp of the signal source in each view in each view can be obtained from the mutual relationship. Here, if 2πfΔxp / Fw = Δφx (p), the above equation (4) is obtained, and if the two-dimensional Fourier transform is performed on D (f, p) obtained by phase correction, the signal source g ( An image corresponding to the case where (x, y) is not moving is obtained. That is, the subject moving in the x-axis direction can be imaged, and the artifact can be suppressed and imaged.

In the MR imaging method according to the second aspect and the MRI apparatus according to the fourth aspect, at least two echoes are generated for one RF pulse. For one of the echoes, the frequency encoding amount is set to “0”, and the data number p ′ (− M ′ / 2) is applied while applying the read gradient of the sampling number M ′ to the phase axis.
The navigation data SWp (p ′) in the phase axis direction in the p view of ≦ p ′ ≦ M ′ / 2−1) is collected. Also, for the remaining echoes, p in k-space
Phase encoding and frequency encoding according to the view are added, whereby the p-view imaging data D ′ (f, p) of the k-space is collected in order. And
The phase correction amount Δφy (p) is calculated from the mutual relationship of the navigation data SWp (p ′) of each view, and the imaging data D ′ (f, p) of each view is phased by the phase correction amount Δφy (p). to correct. Since the navigation data SWp (p ′) is always collected with the frequency encoding amount “0”, the movement amount Δyp of the signal source in each view in the y-axis direction can be obtained from the mutual relationship. Here, if 2πpΔyp / Pw = Δφy (p), the above equation (5) is obtained, and if two-dimensional Fourier transform is performed on D (f, p) obtained by phase correction, the signal source g ( An image corresponding to the case where (x, y) is not moving is obtained. That is, the subject moving in the y-axis direction can be imaged, and the artifact can be suppressed and imaged.

It should be noted that as a combination of the above structures, at least three echoes are generated for one RF pulse. For one of the echoes, the phase encode amount is set to “0”, and the data number f ′ is applied while applying the readout gradient of the sampling number N ′ on the frequency axis.
Navigation data SR in the frequency axis direction in the p-view of (-N '/ 2≤f'≤N' / 2-1)
Collect p (f '). For the other echo, the data number p ′ (− M ′ / 2 ≦ p ′ ≦ is set while setting the frequency encoding amount to “0” and applying the read gradient of the sampling number M ′ to the phase axis. The navigation data SWp (p ′) in the phase axis direction in the p view of M ′ / 2-1) is collected. Further, for the remaining echoes, phase encoding and frequency encoding according to the p-view of k-space are added, whereby the imaging data D ′ (f, p) of the p-view of k-space is sequentially added. collect. Then, the phase correction amount Δφx is calculated from the mutual relation of the navigation data SRp (f ′) of each view.
(P) is calculated, and the phase correction amount Δφy is calculated from the mutual relation of the navigation data SWp (p ′) of each view.
(P) is calculated, and the phase correction amounts Δφx (p) and Δφy are calculated.
In (p), the imaging data D ′ (f, f,
p) is phase-corrected. Navigation data SR
Since p (f ') is always collected with the phase encode amount "0", the movement amount Δxp of the signal source in each view in the x-axis direction can be obtained from the mutual relationship. Further, since the navigation data SWp (p ′) is always collected with the frequency encoding amount “0”, the movement amount Δyp in the y-axis direction of the signal source in each view can be obtained by the mutual relationship. . Here, if 2πfΔxp / Fw = Δφx (p) 2πpΔyp / Pw = Δφy (p), the above equation (6) is obtained, and the two-dimensional Fourier transform is performed on D (f, p) obtained by phase correction. By doing so, an image corresponding to the case where the signal source g (x, y) does not move is obtained. That is, the subject moving in the x-axis direction and the y-axis direction can be imaged, and the artifact can be suppressed and imaged.

In the MRI apparatus according to the fifth aspect, k
-Set the encoding amount of the first encode axis of the space to "0" and collect the navigation data S1 while applying the read gradient to the second encode axis of the k-space. Then, the navigation data S2 is similarly collected. The cross-correlation between the projection profile obtained from the absolute value of the one-dimensional Fourier transform of the navigation data S1 and the projection profile obtained from the absolute value of the one-dimensional Fourier transform of the navigation data S2 corresponds to the second encode axis. The movement amount Δ of the signal source in the direction of the real space is calculated. Since the projection profile reflects the position of the signal source in real space,
The variable value that gives the maximum value of those cross-correlation functions represents the position shift. Therefore, from this, the amount of movement of the signal source Δ
I understand.

[0022]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in more detail based on the embodiments shown in the drawings. The present invention is not limited to this.

First Embodiment In the first embodiment, the axis of the real space corresponding to the frequency axis direction of the k-space is the x axis, and the axis of the real space corresponding to the phase axis direction of the k-space is y. This is an example corresponding to the case where the movement direction of the subject is the x-axis direction when the z-axis is the axis of the real space orthogonal to the x-axis and the y-axis.

FIG. 1 is a block diagram of a first embodiment of the MRI apparatus of the present invention. In this MRI apparatus 100, the magnet assembly 1 has a bore (hole) for inserting a subject (not shown) therein, and a constant static magnetic field is applied to the subject by surrounding the bore. A static magnetic field coil, a gradient magnetic field coil for generating a gradient magnetic field (the gradient magnetic field coil includes orthogonal Z-axis, Y-axis, and X-axis coils), and excites spins of nuclei in a subject. For this purpose, a transmission coil for applying an RF pulse, a reception coil for detecting an NMR (Nuclear Magnetic Resonance) signal from the subject, and the like are arranged. Static magnetic field coil,
The gradient magnetic field coil, the transmitting coil and the receiving coil are connected to the main magnetic field power source 2, the gradient magnetic field driving circuit 3, the RF power amplifier 4 and the preamplifier 5, respectively.

The sequence storage circuit 8 operates the gradient magnetic field drive circuit 3 based on the stored pulse sequence in accordance with a command from the computer 7 to generate a gradient magnetic field from the gradient magnetic field coil of the magnet assembly 1, and The gate modulation circuit 9 is operated to modulate the carrier wave output signal of the RF oscillating circuit 10 into a pulsed signal having a predetermined timing and a predetermined envelope shape, which is applied as an RF pulse to the RF power amplifier 4, and the power is supplied by the RF power amplifier 4. After amplification
It is applied to the transmission coil of the magnet assembly 1 to excite the target region of the subject.

The preamplifier 5 includes a magnet assembly 1
The NMR signal from the subject detected by the receiving coil is amplified and input to the phase detector 12. The phase detector 12 is
The carrier wave output signal of the RF oscillation circuit 10 is used as a reference signal, and the NMR signal from the preamplifier 5 is phase-detected and given to the A / D converter 11. The A / D converter 11 converts the analog signal after phase detection into a digital signal and inputs it to the computer 7. The computer 7 performs an image reconstruction operation on the digital signal from the A / D converter 11 to generate an image of the target area of the subject. This image shows the display device 6
Is displayed at. Further, the computer 7 is responsible for overall control such as receiving information input from the console 13.

The moving device 14 carries the subject and carries the subject into the bore of the magnet assembly 1.

The sequence storage circuit 8 stores a pulse sequence for carrying out the MR imaging method of the present invention. Further, the computer 7 collects imaging data and navigation data based on a pulse sequence for implementing the MR imaging method of the present invention, and phase-corrects the imaging data based on the navigation data, An image reconstruction calculation is performed on the corrected imaging data to generate an image of the target area of the subject.

FIG. 2 shows a pulse sequence for imaging a subject while moving it in the x-axis direction. In this pulse sequence A, an RF pulse Rα having a flip angle α ° is applied and a slice gradient ss is applied to the z axis. Next, the phase encode gradient pe is applied to the y axis. next,
While applying the frequency encode gradient ri to the x-axis, the image echo IE generated after the time TE1 from the RF pulse Rα is sampled to collect p-view imaging data D ′ (f, p). Next, a rewind gradient pr is applied to return the phase encode amount to "0". Next, the readout gradient rn of the sampling number N'in which the polarity is inverted with respect to the frequency encoding gradient ri.
While applying 1, the time TE2 from the RF pulse Rα
The navigation echo NE1 generated later is sampled to obtain p-view imaging data D ′ (f,
navigation data SRp (f ') corresponding to p)
To collect. This is repeated M times with the repetition time TR while changing the phase encode gradient pe and the rewind gradient pr, and M views of imaging data D ′ are obtained.
(F, -M / 2) to D '(f, M / 2-1) and M views worth of navigation data SR corresponding thereto
_{-M / 2} (f ') to SR _{M / 2-1} (f') are collected. In addition,
In FIG. 2, dp is a Dephase gradient and cr is a Crush gradient.

FIG. 3A shows k when p = -M / 2.
-Data for imaging in space D '(f, -M / 2)
3 is a conceptual diagram showing the relationship between navigation data SR _{-M / 2} (f '). In FIG. 3B, p = −M / 2 + 1
Imaging data D '(f,
-M / 2 + 1) and navigation data SR _{-M / 2 + 1}
It is a conceptual diagram which shows the relationship of (f ').

FIG. 4 shows the navigation data SR.
Based on p (f '), the imaging data D' (f,
It is a flowchart of the process which carries out the phase correction of p). At step V1, the view number counter p is initialized to p = -M / 2. In step V2, the conjugate SRp of the navigation data SRo (f ') corresponding to the 0 view and the navigation data SRp (f') in the p view.
^* Multiply (f '). That is, ZRp (f ′) = SRp ^* (f ′) · SRo (f ′) is calculated.

In step V3, the product ZRp (f ') is set to 1
The cross-correlation function zrp (x) is obtained by performing a four-dimensional Fourier transform. FIG. 5 shows the cross-correlation function. Cross-correlation function zrp
(X) is 1 of the navigation data SRo (f ')
Projection profile sro (x) obtained from the absolute value of the three-dimensional Fourier transform and navigation data SRp
It is a cross-correlation function of the projection profile srp (x) obtained from the absolute value of the one-dimensional Fourier transform of (f ') (correlation theorem).

Returning to FIG. 4, in step V4, the cross-correlation function zrp (x) is obtained by interpolation approximation using the least square method.
The variable value Δxp that gives the maximum value of is calculated. Step V5
Then, the phase correction amount Δφx (p) is calculated by 2πfΔxp / Fw = Δφx (p), and the imaging data D ′ is obtained.
(F, p) is multiplied by exp {jΔφx (p)}, and this is used as imaging data D (f, p) after phase correction of the p view. In steps V6 and V7, the above steps V2 to V5 are repeatedly executed from p = -M / 2 + 1 to p = M / 2-1 to acquire the imaging data D (f, p) of all the views forming the k space. To do.

Imaging data D with the above phase correction
If the image is reconstructed from (f, p) and the image is generated,
Even if the subject is imaged while continuously moving in the x-axis direction,
High-quality images with reduced motion artifacts can be obtained.

Second Embodiment In the second embodiment, the axis of the real space corresponding to the frequency axis direction of the k-space is the x axis, and the axis of the real space corresponding to the phase axis direction of the k-space is y. This is an example corresponding to the case where the movement direction of the subject is the y-axis direction when the z-axis is the axis of the real space orthogonal to the x-axis and the y-axis. The block diagram of the MRI apparatus is the same as that of FIG.

FIG. 6 shows a pulse sequence when the subject is imaged while moving in the y-axis direction. In this pulse sequence B, an RF pulse Rα having a flip angle α ° is applied and a slice gradient ss is applied to the z axis. Next, the phase encode gradient pe is applied to the y axis. next,
While applying the frequency encode gradient ri to the x-axis, the image echo IE generated after the time TE1 from the RF pulse Rα is sampled to collect p-view imaging data D ′ (f, p). Next, the rewind gradient pr is applied to return the phase encode amount to "0".
Further, the frequency encoding amount is returned to “0” by applying a rephase gradient rp whose polarity is inverted from that of the frequency encoding gradient ri. Next, while applying the dephasing gradient dp and the readout gradient rn2 of the sampling number M ′ to the phase axis (y-axis), the navigation echo NE2 generated after the time TE3 from the RF pulse Rα is sampled to perform p-view imaging. Data D '
Navigation data SWp corresponding to (f, p)
Collect (p '). While changing the phase encode gradient pe and the rewind gradient pr, this is repeated time T
Repeated M times in R, imaging data D '(f, -M / 2) to D' (f, M / 2-1) for M views and corresponding navigation data SW _-M for M views. Collect _{/ 2} (f ') to SWM _{/ 2-1} (f'). FIG. 7A shows imaging data D ′ (f, −M / 2) in k-space when p = −M / 2 and navigation data SW _{−M / 2} (f ′). It is a conceptual diagram which shows a relationship. FIG. 7B shows imaging data D ′ (f, −M / 2 + 1) in k-space when p = −M / 2 + 1.
3 is a conceptual diagram showing the relationship between the navigation data SW _{-M / 2 + 1} (f ').

FIG. 8 shows the navigation data SW.
Based on p (p '), the imaging data D' (f,
It is a flow chart of processing which carries out phase correction of p). At step U1, the view number counter p is initialized to p = -M / 2. In step U2, the navigation data SWo (p ′) corresponding to the 0 view and the conjugate SWp of the navigation data SWp (p ′) in the p view are
^* Multiply (p '). That is, ZWp (p ′) = SWp ^* (p ′) · SWo (p ′) is obtained.

In step U3, the product ZWp (p ') is set to 1
A two-dimensional Fourier transform is performed to obtain a cross-correlation function zwp (y). FIG. 9 shows the cross-correlation function zwp (y). The cross-correlation function zwp (y) is the navigation data SWo.
A cross-correlation function of the projection profile swo (y) obtained from the absolute value of the one-dimensional Fourier transform of (p ′) and the projection profile srp (y) obtained from the absolute value of the one-dimensional Fourier transform of the navigation data SWp (p ′). (Correlation theorem).

Returning to FIG. 8, in step U4, the cross-correlation function zwp (y) is obtained by interpolation approximation using the least square method.
The variable value Δyp that gives the maximum value of is calculated. Step U5
Then, the phase correction amount Δφy (p) is calculated by 2πpΔyp / Pw = Δφy (p), and the imaging data D ′ is obtained.
(F, p) is multiplied by exp {jΔφy (p)} to obtain imaging data D (f, p) after phase correction of the p view. In steps U6 and U7, the above steps U2 to U5 are repeatedly executed from p = -M / 2 + 1 to p = M / 2-1, and the imaging data D (f, p) of all views forming the k space are acquired. To do.

The phase-corrected imaging data D
If the image is reconstructed from (f, p) and the image is generated,
Even if the subject is imaged while continuously moving in the y-axis direction,
High-quality images with reduced motion artifacts can be obtained.

Third Embodiment In the third embodiment, the axis of the real space corresponding to the frequency axis direction of the k-space is the x axis, and the axis of the real space corresponding to the phase axis direction of the k-space is y. This is an example corresponding to the case where the moving direction of the subject is the x-axis and the y-axis direction when the z-axis is the axis of the real space orthogonal to the x-axis and the y-axis. MRI
The block diagram of the device is similar to FIG.

FIG. 10 shows a pulse sequence in the case of imaging an object while moving it in the x-axis direction and the y-axis direction. In this pulse sequence C, the flip angle R is R.
An F pulse Rα is applied and a slice gradient ss is applied to the z axis.
Is applied. Next, the phase encode gradient pe is applied to the y axis. Next, while applying the frequency encode gradient ri to the x-axis, the image echo IE generated after the time TE1 from the RF pulse Rα is sampled to collect the p-view imaging data D ′ (f, p). Next, the rewind gradient pr is applied to return the phase encode amount to "0". Next, the readout gradient rn of the sampling number N'in which the polarity is inverted with respect to the frequency encoding gradient ri.
While applying 1, the time TE2 from the RF pulse Rα
The navigation echo NE1 generated later is sampled to obtain p-view imaging data D ′ (f,
navigation data SRp (f ') corresponding to p)
To collect. Next, a rephase gradient rp whose polarity is inverted from that of the frequency encode gradient rn1 is applied to return the frequency encode amount to "0". Further, while applying the dephasing gradient dp and the reading gradient rn2 of the sampling number M ′ to the phase axis (y-axis), the navigation echo NE2 generated after the time TE3 from the RF pulse Rα is sampled and used for p-view imaging. Navigation data SW corresponding to data D '(f, p)
Collect p (p '). This is the phase encode gradient pe
While changing the rewind gradient pr and M times for the repetition time TR, M views worth of imaging data D ′ (f, −M / 2) to D ′ (f, M / 2-1) and corresponding to them. Navigation data SR _{-M / 2} (f ') to SR _{M /} 2 _-1 (f') for M views and navigation data SW _{-M / 2} (f ') to SW
Collect _{M / 2-1} (f '). FIG. 11 is a flowchart of a process of phase-correcting the imaging data D ′ (f, p) based on the navigation data SRp (f ′) and SWp (p ′). In step S1,
The view number counter p is initialized to p = -M / 2. In step S2, the navigation data SRo (f ') corresponding to the 0 view is multiplied by the conjugate SRp ^* (f') of the navigation data SRp (f ') in the p view. That is, ZRp (f ′) = SRp ^* (f ′) · SRo (f ′) is calculated. In step S3, the product ZRp (f ') is subjected to one-dimensional Fourier transform to obtain a cross-correlation function zrp (x). In step S4, the variable value Δxp that gives the maximum value of the cross-correlation function zrp (x) is obtained by interpolation approximation using the method of least squares.

In step S5, the conjugate SWp ^{* of} the navigation data SWo (p ') corresponding to the 0 view and the navigation data SWp (p') in the p view ^.
Multiply (p '). That is, ZWp (p ′) = SWp ^* (p ′) · SWo (p ′) is obtained. In step S6, the product ZWp (p ′) is subjected to one-dimensional Fourier transform to obtain a cross-correlation function zwp (y). In step S7, a variable value Δyp that gives the maximum value of the cross-correlation function zwp (y) is obtained by interpolation approximation using the least square method. In step S8, 2πfΔxp /
The phase correction amount Δφx (p) is calculated by Fw = Δφx (p), the phase correction amount Δφy (p) is calculated by 2πpΔyp / Pw = Δφy (p), and the imaging data D ′ is calculated.
(F, p) is multiplied by exp {j (Δφx (p) + Δφy (p))}, and this is used as imaging data D (f, p) after phase correction of the p view. In steps S9 and S10, the above steps S2 to S8 are repeatedly executed from p = -M / 2 + 1 to p = M / 2-1, and the imaging data D (f, p) of all the views forming the k space are acquired. To do.

Imaging data D with the above phase correction
If the image is reconstructed from (f, p) and the image is generated,
Even if the subject is imaged while moving in the x-axis direction and the y-axis direction, a high-quality image with reduced motion artifacts can be obtained.

-Other Embodiments-In the first to third embodiments, the pulse sequence of the gradient echo method has been described, but the present invention is not limited to this.
For example, a pulse sequence in which a rewind gradient and a read gradient for reading the navigation echo NE are added to the pulse sequence of the spin echo method may be used.

Further, in the first to third embodiments, the variable value which gives the maximum value of the cross-correlation function is obtained by the least square method, but the Lagrange interpolation method, the Chebyshev interpolation method, the Hanning interpolation method and the Cubic interpolation method are used. The variable value that gives the maximum value of the cross-correlation function may be obtained using, for example.

Further, in the first to third embodiments, the correlation correction theorem is used to calculate the phase correction amount. However, the change of the center of gravity position of the projection profile obtained from the absolute value of the one-dimensional Fourier transform of the navigation data is changed. Alternatively, the phase correction amount may be calculated from the change in the rising / falling position.

[0048]

According to the MR imaging method and the MRI apparatus of the present invention, it is possible to image a moving subject and suppress artifacts. Therefore, it is possible to obtain a high-quality image by continuously imaging the subject while moving it, and it is possible to improve the imaging efficiency.

[Brief description of drawings]

FIG. 1 is a configuration diagram of an embodiment of an MRI apparatus of the present invention.

FIG. 2 is a pulse sequence diagram for imaging a subject while moving in the x-axis direction.

FIG. 3 is data for imaging D ′ (f, f in k-space,
It is a conceptual diagram which shows the relationship between p) and navigation data SRp (f ').

FIG. 4 is a flowchart of phase correction processing when an image of a subject is taken while moving in the x-axis direction.

FIG. 5 is a conceptual diagram of a cross-correlation function when a subject is imaged while moving in the x-axis direction.

FIG. 6 is a pulse sequence diagram for imaging a subject while moving in the y-axis direction.

FIG. 7 is data D ′ (f, for imaging in k-space
It is a conceptual diagram which shows the relationship between p) and navigation data SWp (p ').

FIG. 8 is a flowchart of a phase correction process when an image of a subject is taken while moving in the y-axis direction.

FIG. 9 is a conceptual diagram of a cross-correlation function when a subject is imaged while moving in the y-axis direction.

FIG. 10 is a pulse sequence diagram for imaging a subject while moving in the x-axis direction and the y-axis direction.

FIG. 11 is a flowchart of a phase correction process when an image of a subject is captured while moving in the x-axis direction and the y-axis direction.

FIG. 12 is a conceptual diagram of a stationary signal source and k-space.

FIG. 13 is a conceptual diagram of a signal source moved in the x-axis direction and k-space.

FIG. 14 is a block diagram of an example of a conventional MRI apparatus.

[Explanation of symbols]

100 MRI device 1 Magnet assembly 3 Gradient magnetic field drive circuit 7 calculator 8 Sequence memory circuit 14 Mobile devices IE image echo NE navigation echo

─────────────────────────────────────────────────── ─── Continuation of the front page (56) Reference JP-A-63-164943 (JP, A) JP-A-6-47021 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB name) A61B 5/055

Claims (3)

(57) [Claims] 1. A frequency axis data number of k-space is represented by f (-
N / 2 ≦ f ≦ N / 2−1) and the phase axis data number is p (−M / 2 ≦ p ≦ M / 2−1), RF pulse is applied, phase encoding is applied, and frequency is applied. The encoding data is added and the imaging data D ′ (f, p) is collected from the generated echoes, and the imaging data D ′ (f, −M / 2) to D ′ for M views are repeated.
In an MRI apparatus that collects (f, M / 2-1), at least two echoes are generated for one RF pulse, the phase encode amount is set to "0" for one echo, and While applying the reading gradient of the sampling number N ′, the data number f ′ (− N ′ / 2 ≦
f ′ ≦ N ′ / 2−1) navigation data SRp (f ′) in the frequency axis direction in the p-view of f ′ ≦ N ′ / 2-1) is collected, and the remaining echoes are phase-encoded according to the p-view of k-space. In addition, frequency encoding is added, whereby the collection of imaging data D ′ (f, p) of p views in k-space is repeated, and imaging data D ′ (f, −M / 2) of M views is repeated. ~
Data collecting means for collecting D '(f, M / 2-1) and navigation data SR _{-M / 2} (f') for each view
The cross-correlation function is obtained by multiplying the navigation data of a certain view in SR _{M / 2-1} (f ') by the conjugate of the navigation data of a view different from that view, and the phase correction amount is calculated from the cross-correlation function. An MRI apparatus comprising: a phase correction unit that calculates Δφx (p) and phase-corrects the imaging data D ′ (f, p) of each view with the phase correction amount Δφx (p). 2. The frequency axis data number of k-space is represented by f (-
N / 2 ≦ f ≦ N / 2−1) and the phase axis data number is p (−M / 2 ≦ p ≦ M / 2−1), RF pulse is applied, phase encoding is applied, and frequency is applied. The encoding data is added and the imaging data D ′ (f, p) is collected from the generated echoes, and the imaging data D ′ (f, −M / 2) to D ′ for M views are repeated.
In an MRI apparatus that collects (f, M / 2-1), at least two echoes are generated for one RF pulse, the frequency encoding amount is set to "0" for one echo, and the phase axis is While applying the reading gradient of the sampling number M ′, the data number p ′ (− M ′ / 2 ≦ p ′
Collecting navigation data SWp (p ′) in the phase axis direction in the p view of ≦ M ′ / 2-1),
The remaining echoes are phase-encoded and frequency-encoded according to the p-view of k-space, whereby the imaging data D ′ of the p-view of k-space is added.
The collection of (f, p) is repeated, and the imaging data D ′ (f, −M / 2) to D ′ (f,
M / 2-1) for collecting data and navigation data SW _{-M / 2} (p ') for each view
The cross-correlation function is obtained by multiplying the navigation data of a view in SW _{M / 2-1} (p ') by the conjugate of the navigation data of a view different from that view, and the phase correction amount is calculated from the cross-correlation function. An MRI apparatus comprising: a phase correction unit that calculates Δφy (p) and phase-corrects the imaging data D ′ (f, p) of each view with the phase correction amount Δφy (p). 3. The navigation data S while setting the encode amount of the first encode axis of the k-space to "0" and applying a read gradient to the second encode axis of the k-space.
1 and then similarly collecting navigation data S2, a projection profile obtained from the absolute value of the one-dimensional Fourier transform of the navigation data S1, and a one-dimensional Fourier transform of the navigation data S2. And an amount of movement calculating means for calculating the amount of movement Δ of the signal source in the direction of the real space corresponding to the second encoding axis from the cross-correlation of the projection profile obtained from the absolute value of .