JP3023831B2 - Nuclear magnetic resonance imaging equipment - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to nuclear magnetic resonance (nucl)
ear magnetic resonance; NM
R) The present invention relates to a nuclear magnetic resonance imaging apparatus that obtains morphological information such as a two-dimensional image or a three-dimensional image of a subject (living body) or functional information of spectroscopy using a phenomenon.

[0002]

2. Description of the Related Art In nuclear magnetic resonance imaging, shortening the imaging time has a great effect in terms of improving system throughput, reducing patient's pain, and reducing or eliminating artifacts due to pulsation. For this reason, an effective three-dimensional fast nuclear magnetic resonance imaging method has been conventionally proposed.

[0003] Among such high-speed nuclear magnetic resonance imaging methods, the following have been relatively recently attempted.

(1) FLASH method In this method, excitation is performed with a pulse having a flip angle of a small angle of 90 ° or less, and image data is obtained in a time of about 20 to 50 msec, which is much shorter than in the case of the spin echo (SE) method or the like. Can be collected repeatedly, and a magnetic resonance image can be obtained in an imaging time of about several seconds. For this reason,
It is widely used for diagnosis of parts with severe body movement such as the abdomen. (P. Van Der Meulen, J. et al.
P. Groen, and J.M. J. M. Cuppe
n, "Very fast MR imaging b
y field echoes and small
angle exitation, "Mag. R
eson. imag. , Vol. 3, pp297-
(299, 1985) (2) Echo Planar Method This method collects image data with one excitation using a gradient magnetic field that reverses at a high speed, and uses transverse relaxation (T ₂ ).
One image data can be collected in a time period of about several tens of msec when the attenuation of the image does not matter. For this reason, it is expected to provide an image with extremely large clinical significance, such as displaying the motion of the heart in real time.
(P. Mansfield, IL Pykett,
“Biological and medical i
mag. by NMR, "J. Magn.
Reson. , Vol. 29, pp. 355-373,
1978. )

[0005]

However, in the above method (1), the data collection time in the case of two-dimensional imaging requires several seconds, which is considered to be still too long. In the future, if imaging is performed at high speed with a high magnetic field device,
When performing two-dimensional imaging, the burden on the high-frequency circuit is increased, and the effect on the human body, in particular, the effect on the eyeball due to the high frequency, and the effect on the fetus may be a problem.

In the above method (2), although the problem of high frequency does not occur, it is necessary to reverse the gradient magnetic field at a very high speed. In particular, the transverse relaxation time (T ₂ ^* ) is required to obtain all data. Therefore, there are many technical difficulties because the time must be several tens of msec or more.

Further, according to either of the above methods (1) and (2), the NMR signal is collected in a pulsed manner.
Correspondingly, the reversal of the gradient magnetic field must be performed in a pulsed manner. The accuracy of the time and magnitude of the pulsed gradient magnetic field has a subtle effect on the image quality. Since this pulse-like gradient magnetic field induces eddy currents in surrounding structures, the reversal of the gradient magnetic field may be delayed or the spatial distribution of the gradient magnetic field may be distorted.
Therefore, the image may be deteriorated. In addition, when imaging is performed at high speed with a high magnetic field device, the generation of atrial and ventricular fibrillation due to nerve stimulation with a magnetic field that changes at high speed may cause a problem.

[0008] Currently Anyway, the proposed method of obtaining a magnetic resonance imaging to help clinical diagnosis in such a high-speed magnetic resonance imaging does not reach the practical stage.

[0009] Therefore, a nuclear magnetic resonance imaging method using phase encoding was studied and presented at an academic meeting (Japan Society for M.E., May 13, 1994).

[0010] However, subsequent studies have revealed that this method cannot detect the proton density distribution.

[0011] The nuclear magnetic resonance imaging apparatus of the present invention comprises:
It was done to eliminate the disadvantages of the above conventional example,
The purpose is as follows.

[0012] It is possible to perform high-speed shooting with a scan time of only tens of ms to tens of ms, and to prevent artifacts from occurring due to movement of a human body and deteriorating image quality.
It also reduces the physical and mental distress of the patient.

There is no need to reverse the gradient magnetic field, thereby preventing image degradation due to eddy currents and without affecting the human body such as the occurrence of atrial or ventricular fibrillation due to nerve stimulation.

[0014] Since the waveform of the magnetic field is simple, reduce the cost of the coil is simplified.

[0015]

The principle of the present invention will be described below by taking the case of three-dimensional imaging as an example. Here, the proton ( ¹ H) is used as an example, but the same applies to other nuclear magnetic resonance nuclei.

(A) Conversion of spatially continuous proton density distribution to discrete lattice-like proton density distribution The x, y, and z directions of the image space v are respectively represented by N _x , Δy, and Δz in units of Δx, Δy, and Δz. N _y and N _{z are} equally divided, and the video space v is divided into N
divided into _x × N _y × N _z number of unit image element v _ijk. The resonance signal s (t) collected at this time is the sum of the signals s _ijk (t) generated from each unit image body v _ijk . That is,

[0017]

(Equation 1)

## EQU1 ##

According to the nuclear magnetic resonance theory, a signal s _ijk (t) generated from a unit image body subjected to a gradient magnetic field is

[0020]

(Equation 2)

It can be written that

Here, p _ijk (x, y, z): a proton density distribution function G _x , G _y , G _z in the unit image body v _ijk : a gradient magnetic field in the x, y, z directions ι: an imaginary unit √ (-1).

In this equation, the effects of the vertical relaxation time T ₁ , the horizontal relaxation time T ₂ , and the body motion are omitted because the measurement time is short and the effects can be ignored.

As will be described later, for a continuous proton density distribution, a proton density distribution function p _ijk (x, y, z) cannot be determined. However, when the volume of each unit image is sufficiently small, the proton density distribution function of each unit image can be regarded as a constant. Therefore, the proton density distribution of each unit image can be approximated as p _ijk (x, y, z) = p _ijk, and the signal generated from each unit image is

[0025]

(Equation 3)

The collected resonance signal is

[0027]

(Equation 4)

Can be written as Note that v ₀ in the equation is the volume of each unit image body _vijk .

Here,

[0030]

(Equation 5)

[0031]

(Equation 6)

[0032]

[0033]

(Equation 7)

Is obtained. This signal w (t) is at the position (i
Δx, jΔy, kΔz) proton density distribution p
It can be regarded as the sum of resonance signals generated from _ijk .

According to the principle described above, a signal generated by a spatially continuous proton density distribution p (x, y, z) is converted into a signal generated by a discrete lattice-like proton density distribution p _ijk. You can do it.

(B) Selection of a gradient magnetic field such that the resonance lines corresponding to the lattice points of the lattice-like proton density distribution converted as described above do not overlap, according to the conventional nuclear magnetic resonance imaging theory, 2
As shown in FIG. 1, in a proton density distribution 95 that is spatially continuous like a human body, the resonance lines 96 of each unit image always overlap with an arbitrary gradient magnetic field G.

On the other hand, as shown in FIG. 20, for the discrete lattice-like proton density distribution 91, discrete resonance lines are observed in the frequency spectra 93 and 94 of the resonance signal. By selecting the gradient magnetic field G2, the resonance line 92 corresponding to each lattice point can be prevented from overlapping. Then, the position of the lattice point can be known from the resonance frequency.

That is, in the discrete lattice-like proton density distribution converted as in the above (A), a gradient magnetic field can be selected such that the frequency spectrum lines of the proton density distribution at each lattice point do not overlap. is there.

Here, there are a plurality of methods for selecting the gradient magnetic field. When Δx = Δy = Δz = ΔN _x = N _y = N _z = N, as shown in FIG. In order not to be too large, it is good to do as follows.

G _y = NG _x G _z = NG _y = N ² G _x or

[0041]

(Equation 8)

[0042]

(Equation 9)

Here, the gradient magnetic field was selected as follows in order to facilitate image reconstruction.

G _y = NG _x G _z = NG _y = N ² G _x Note that the spatially continuous proton density distribution p (x, y,
The conversion of the signal generated by z) into the signal generated by the discrete lattice-like proton density distribution p _ijk is determined by the method of dividing the image space v into unit image bodies v _ijk. The gradient magnetic field can be determined when v is divided into unit video _objects v _ijk .

(C) Acquisition of Image Data from Nuclear Magnetic Resonance Signals in the Gradient Magnetic Field The frequency ω _ijk of the frequency spectrum of the signal w (t) described in the above (A) is the position (iΔx, jΔy, kΔz ), And the height of the frequency spectrum line at the frequency ω _ijk becomes the proton density distribution p _{ijk at the} corresponding position.

Then, from the signal w (t), a proton density distribution p _{ijk in} each unit image is obtained by Fourier transform.

Here, sampling at the time of collecting image data at this time is divided into unit image bodies as assumed in the above (B), and when a gradient magnetic field is applied, the sampling time (L is a positive integer) ):

[0048]

(Equation 10)

Number of sampling points: N _s = N ³ Sampling interval: Δt = T _s / N _s Sampling start time: t ₀ = Δt / 2

(D) Reconstruction of Proton Density Distribution Image from Collected Image Data The proton density distribution p _{ijk in} each unit video obtained in (C) above is obtained by superposing the position information on the frequency. Has become. Therefore, the position of each proton density distribution p _ijk is specified from this frequency, and the image is reconstructed accordingly.

[0051]

The nuclear magnetic resonance imaging apparatus of the present invention based on the above principle has the following structure A or B.

A. A nuclear magnetic resonance imaging apparatus for obtaining an image from a nuclear magnetic resonance signal in a three-dimensional image space in which a temporally constant gradient magnetic field is superimposed on a static magnetic field, comprising : (a) a plurality of three-dimensional image spaces in each direction; Split
Divided into a plurality of unit video bodies Te, (b) in accordance with the way divided in the unit video material, each unit video bodies
Mutually different strength of the magnetic field determines a gradient magnetic field to be applied to the three-dimensional image space each, by applying a radio wave pulses (c) the three-dimensional image space, (d) O to the falling edge of the radio wave pulse At the same time, the gradient magnetic field determined in (b) is applied to the three-dimensional image space. (E) Nuclear magnetic resonance signals generated from each unit image are converted from a grid-like point having a uniform density of nuclear magnetic resonance nuclei. (F) A nuclear magnetic resonance imaging apparatus that collects image data by regarding the generated signal, and (f) reconstructs an image using the image data collected in (e).

B. A nuclear magnetic resonance imaging apparatus for obtaining an image from a nuclear magnetic resonance signal in a three-dimensional image space in which a temporally constant gradient magnetic field is superimposed on a static magnetic field, comprising : (a) a plurality of three-dimensional image spaces in each direction; Split
Divided into a plurality of unit video bodies Te, (b) in accordance with the way divided in the unit video material, each unit body
Mutually different strength of the magnetic field determines a gradient magnetic field to be applied to the three-dimensional image space, (c) gives the gradient magnetic field in the three-dimensional image space, (d) the 3-dimensional this gradient magnetic field is applied (E) Nuclear magnetic resonance signals generated from each unit image body by applying the radio wave pulses to a signal generated from a lattice-like point having a uniform density of nuclear magnetic resonance nuclei. (F) A nuclear magnetic resonance imaging apparatus for reconstructing an image using the image data collected in (e) above.

In the above-mentioned apparatus A or B,
The gradient magnetic field is determined by selecting and applying a radio frequency signal of each lattice point so that the frequency spectrum of each lattice point does not overlap when a nuclear magnetic resonance signal generated from each unit image is regarded as a signal generated from the lattice point. The wave pulse sets the excitation angle to 90
°.

The method of dividing the unit image bodies and the determination of the gradient magnetic field may be set in advance.

[0056]

According to the nuclear magnetic resonance imaging apparatus of the present invention having the above-described configuration , a nuclear magnetic resonance imaging image can be obtained at high speed within a certain error range using only a constant gradient magnetic field.

[0057]

Embodiments of the present invention will be described below with reference to the drawings.

[0058] (Embodiment 1) FIG. 1 is a block diagram of a nuclear magnetic resonance Imaging GuSo location of the present invention.

In this figure, reference numeral 1 denotes an excitation power supply, which supplies power to the electromagnet C1, thereby generating a static magnetic field in the image space. Reference numeral 2 denotes a high-frequency pulse generation unit, and an irradiation coil C 3
It generates more radio wave pulses. 3 is a linear magnetic field gradient generator and gives the gradient magnetic field in the image space from the magnetic field gradient coil C 4. 5 is a detection and amplification unit, in which further amplifies the nuclear magnetic resonance signals detected through the receiving coils C 2. Reference numeral 4 denotes an arithmetic and processing unit that controls each unit, determines a gradient magnetic field, and calculates and reconstructs image data from a received nuclear magnetic resonance signal. Reference numeral 6 denotes a display / recording unit for displaying the most composed image on a display or recording the image on a recording medium.

FIG. 2 is a flowchart of the proton image imaging of this embodiment. Hereinafter, the nuclear magnetic resonance imaging apparatus according to the present embodiment will be described with reference to FIG.

First, the image space is divided into unit image bodies, and the processing in the arithmetic / processing unit 4 is performed (S11). Here, an arbitrary division method such as 64 × 64 × 64, 128 × 128 × 128, etc. can be selected, and the division method can be manually input or automatically determined. It is. Next, the calculation / processing unit 4 determines a gradient magnetic field to be applied to the image space according to the method of dividing the unit image bodies, and sends this data to the linear magnetic field gradient generation unit 3 (S1).
2). If the method of dividing the unit video body is set in the apparatus in advance, it is not necessary to perform these processes each time.

Next, the arithmetic-processing unit 4 controls the high-frequency pulse generator 2, the excitation angle 9 to the image space than the irradiation coil C 3
A radio wave pulse of 0 ° is applied (S13). As a result, the magnetization is tilted laterally, and the nuclear spin is tilted. Next, at the same time as the fall of the radio wave pulse,
4 controls the linear magnetic field gradient generator 3 and generates a magnetic field gradient coil C
From step 4, the gradient magnetic field is applied to the image space (S14).

Next, receives nuclear magnetic resonance signals generated in this case the receiving coils C 2 and detection and amplified by the detection and amplification unit 5, sends them to the arithmetic-processing unit 4, the arithmetic-processing unit 4 As described in the principle of the invention, nuclear magnetic resonance signals generated from each unit image are calculated as described above, and image data is collected by regarding these as signals generated from grid points having a uniform proton density. (S15). In the processing so far, a static magnetic field has been given to the image space by the electromagnet C1.

The calculation / processing section 4 reconstructs the proton density distribution from the collected image data and displays and reconstructs it.
The reconstructed image is sent to the recording section 6, and the reconstructed image is displayed on a display or recorded on a recording medium (S16).

[0065] Hereinafter, will be described with reference to the drawings results of a simulation performed on a computer for nuclear magnetic resonance imaging apparatus according to our method described above.

The conditions in this simulation are as follows: the static magnetic field strength is 0.2 T, and the pixels are 64 × 64 two-dimensional pixels.

FIG. 3 is a pulse diagram of the proton image imaging of the present embodiment.

Here, the pulse diagram is given as a three-dimensional display, but if it is considered that there is no component in the z direction, it will be a two-dimensional one. The acquisition of image data is described in the “Acquis
It is performed at the time of “Ition”. These are the same in the pulse diagrams of the subsequent simulations.

FIG. 4 shows the proton density distribution in the imaging space to be simulated here, and the object to be measured is a head model. In this figure, the image space 10
Is a square with a side of 0.2 m, the outermost part 11 is the space outside the skull, the proton density is 0, the inside of the large ellipse 12 is the brain, the proton density is 1 and the small ellipse 13 is Inside is a tumor, assuming a proton density of 2.

The ellipses 12 and 13 indicating the boundaries between the skull and the tumor are respectively

[0071]

[Equation 11]

(Each parameter is: a1 = 0.08 m; b
1 = 0.1 m; x10 = 0.1 m; y10 = 0.1 m)

[0073]

(Equation 12)

(Each parameter is: a2 = 0.03 m; b
2 = 0.04 m; x10 = 0.12 m; y10 = 0.1
2m).

The parameters in this simulation are as follows.

Proton specific gyromagnetic ratio: γ = 42.5759 [MHz / T] Number of sampling points: N _s = 4096 Amount of gradient magnetic field in x-axis direction: G _x = 0.0002 [mT / cm] y-axis direction Of the gradient magnetic field: G _y = 0.0128 [mT / cm] Frequency interval of the unit image body: Δf = 4.235103 [kHz] Frequency width of spectrum: f _max = 17.334247 [Hz] Sampling interval: Δt = 0.002217 [ms] Sampling time: T _s 1 = 9.081606 [ms] Sampling frequency: f _s = 451.021539 [kHz] Sampling start time: t ₀ = 0.001108 [ms] And here FIG. 5 shows a proton image obtained by the simulation of FIG. The simulation was performed under such conditions, and the average value of the proton density of the part outside the skull obtained by the calculation was 3.5785 × 10 ⁻⁹ , the average value of the proton density of the brain part was 1.0055663, Is 2.011326.

FIG. 6 is a pulse diagram for T ₁ image imaging of the present embodiment.

Here, as shown in FIG.
The direction of magnetization falls sideways due to the 90th pulse. By sampling at this point, the first proton image can be obtained. Next, by applying a second 90 ° pulse, the direction of the magnetized sideways returns to vertical again. However, the direction is just opposite to the direction of the static magnetic field. And by the third 90 ° pulse,
The direction of magnetization falls to the side, and sampling at that time causes 2
A second proton image is obtained. Then, using the proton density of the second image and the proton density of the _first image, a T1 image is obtained by the following relational expression.

[0079]

(Equation 13)

Here, p ₁ (x, y, z) is the proton density of the first image, p ₂ (x, y, z) is the proton density of the second image, and T _1d Is the delay time.

FIGS. 7 and 8 show a proton density distribution and a longitudinal relaxation time distribution in an imaging space to be simulated here, respectively. In these figures, the image space 20
Is a square with a side of 0.2 m, the outermost part 21 is the space outside the skull, the proton density is 0, the large ellipse 22 is the brain, the small ellipse 23 is the tumor, and both protons It is assumed that the density is 1. The longitudinal relaxation time of the brain part is 50 msec, and the longitudinal relaxation time of the tumor part is 1 msec.
00 msec.

The other parameters are the same as those described in the simulation of the proton image, but the sampling time T _s 2 = T _s 1 and the delay time T _1d = 100 ms.

When the simulation is performed under such conditions, the average value of the longitudinal relaxation time of the brain portion calculated by the calculation is 52.4786 msec, and the average value of the longitudinal relaxation time of the tumor portion is 103.3784 msec.

9 and 10 show simulation results of the _first proton density distribution image and the T1 image, respectively.

FIG. 11 is a pulse diagram of T ₂ image imaging of the present embodiment.

Here, as shown in FIG. 11, first, the direction of magnetization falls sideways by the first 90 ° pulse.
By sampling at this point, the first proton image can be obtained. Next, after the transverse magnetization is attenuated in the section of _{T2d in} the figure, sampling is performed again to obtain a second proton image. And
Using the proton density of the second image and the proton density of the first image, a T ₂ image is obtained by the following relational expression.

[0087]

[Equation 14]

Here, p ₁ (x, y, z) is the proton density of the first image, p ₂ (x, y, z) is the proton density of the second image, and T _2d Is the delay time.

FIGS. 12 and 13 respectively show the proton density distribution in the imaging space simulated here,
The longitudinal relaxation time distribution is shown. In these figures, the image space 20 is a square having a side of 0.2 m, and the outermost part 3
Reference numeral 1 denotes a space outside the skull, where the proton density is 0, the brain is in the large ellipse 32, and the tumor is in the small ellipse 33, and it is assumed that the proton density is 1 in each case. The lateral relaxation time of the brain part is 20 msec, and the lateral relaxation time of the tumor part is 50 msec.

The other parameters are the same as those described in the simulation of the proton image, but the delay time is assumed to be T _2d = T _1d .

When a simulation is performed under such conditions, the average value of the lateral relaxation time of the brain portion calculated by the calculation is 19.93786 msec, and the average value of the lateral relaxation time of the tumor portion is 49.96371 msec.

14 and 15 respectively.
The simulation result of the _2nd proton image and T2 image is shown.

As a result of the above simulation, 9.574 was obtained in order to obtain one piece of proton image data.
It took only ms, and it was found that very high-speed shooting was possible. As a result, it is possible to prevent image quality from deteriorating due to artifacts caused by the movement of the human body. Also, the physical and mental distress of the patient can be reduced.

[0094] Further, since there is no need to invert the gradient magnetic field in the above proposed method, at all without that image degradation due to eddy currents, also is not affected by the neural stimulation atrium, to the human body such as the occurrence of ventricular fibrillation .

[0095] Further, in the above KiSo location, since the waveform of the magnetic field is easy, it can be expected that reducing the cost of the coil is simplified.

Embodiment 2 FIG. 16 is a flowchart of the proton image imaging of this embodiment.

In the first embodiment, the gradient magnetic field is applied at the same time as the fall of the radio pulse. However, in the present embodiment, the gradient magnetic field is applied in advance (S23), and the radio pulse is applied thereto. (S24), image data is collected from the nuclear magnetic resonance signal generated thereby (S2).
5) Things.

In the case of the apparatus according to the present embodiment, since the gradient magnetic field is applied before the application of the radio wave pulse,
Since a radio pulse having a frequency width enough to cover the resonance frequency width must be applied, compared with the case where a single frequency radio pulse is applied as in the case of the first embodiment. It is difficult to generate radio wave pulses.

However, as compared with the case where the rise and fall of the gradient magnetic field must be controlled as in the case of the first embodiment, time can be saved, the circuit can be simplified, and a stable magnetic field can be obtained. Can be generated, so that the image quality can be prevented from deteriorating.

FIGS. 17, 18 and 19 show pulse diagrams of the proton image, the T ₁ image, and the T ₂ image in the present embodiment, respectively.

When a simulation is performed under the same conditions as in the first embodiment, an image similar to that described in the first embodiment can be obtained.

[0102]

As described above, according to the nuclear magnetic resonance imaging apparatus of the present invention, it is possible to perform high-speed photographing with a very short scan, and an artefact is generated due to the movement of the human body, thereby deteriorating the image quality. That can be prevented. Also, the physical and mental distress of the patient can be reduced.

Further, since it is not necessary to reverse the gradient magnetic field, there is no deterioration of the image due to the eddy current, and there is no influence on the human body such as generation of atrial and ventricular fibrillation due to nerve stimulation.

[0104] In addition, nuclear magnetic resonance image in g of the present invention
In equipment, since the waveform of the magnetic field is easy, it can be expected that reducing the cost of the coil is simplified.

In addition to the above effects, the invention of each claim has the following effects.

According to the first aspect of the present invention, generation of a radio wave pulse is facilitated.

According to the second aspect of the present invention, fine control of the gradient magnetic field becomes unnecessary.

[0108] In the present invention of claim 5, since the classification and the gradient magnetic field of the unit video member are predetermined, it is possible to simplify the equipment.

[Brief description of the drawings]

FIG. 1 is a block diagram of a nuclear magnetic resonance imaging apparatus for implementing the present invention.

FIG. 2 is a flowchart of proton image imaging according to the first embodiment.

FIG. 3 is a pulse diagram of proton image imaging according to the first embodiment.

FIG. 4 is a diagram showing a proton density distribution to be simulated in the proton image imaging according to the first embodiment.

FIG. 5 is a diagram showing a simulation result of a proton image according to the first embodiment.

FIG. 6 is a pulse diagram of T ₁ image imaging according to the _first embodiment.

FIG. 7 is a diagram showing a proton density distribution to be simulated in T ₁ image imaging according to the first embodiment.

FIG. 8 is a diagram showing a longitudinal relaxation time distribution to be simulated in T ₁ image imaging according to the first embodiment.

FIG. 9 shows 1 in T ₁ image imaging of the first embodiment.
The figure showing the simulation result of the 1st proton image

FIG. 10 is a diagram illustrating a simulation result of a T ₁ image according to the _first embodiment.

FIG. 11 is a pulse diagram of T ₂ image imaging according to the first embodiment.

FIG. 12 is a diagram showing a proton density distribution to be simulated in T ₂ image imaging according to the first embodiment.

FIG. 13 is a diagram showing a distribution of transverse relaxation times to be simulated in T ₂ image imaging according to the first embodiment.

FIG. 14 is a diagram showing a simulation result of a first proton image in T ₂ image imaging of the first embodiment.

FIG. 15 is a diagram illustrating a simulation result of a T ₂ image according to the first embodiment.

FIG. 16 is a flowchart of proton image imaging according to the second embodiment.

FIG. 17 is a pulse diagram of proton image imaging according to the second embodiment.

FIG. 18 is a pulse diagram of T ₁ image imaging according to the second embodiment.

FIG. 19 is a pulse diagram of T ₂ image imaging according to the _second embodiment.

FIG. 20 is an explanatory diagram showing that a gradient magnetic field can be selected such that resonance lines corresponding to respective lattice points do not overlap with a discrete lattice-like proton density distribution.

FIG. 21 is an explanatory view showing that it is not possible to select a gradient magnetic field such that resonance lines corresponding to lattice points do not overlap with a spatially continuous proton density distribution.

FIG. 22 is an explanatory diagram of a plurality of gradient magnetic fields capable of observing a discrete lattice-like proton density distribution from one projection spectrum.

[Explanation of symbols]

RF radio frequency pulses G _x x-direction gradient field G _y y-direction gradient field G _z z-direction gradient field FID NMR signal

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Akihiko Uchiyama Kanagawa Prefecture, Yokohama City, Kanazawa Ward, Tomioka Higashi 3-2-17 704

Claims (5)

(57) [Claims] 1. A nuclear magnetic resonance imaging apparatus for obtaining an image from the nuclear magnetic resonance signals in a three-dimensional image space obtained by superimposing a temporally constant gradient magnetic field in the static magnetic field, its (a) the three-dimensional image space Divide into multiple in each direction
Divided into a plurality of unit video bodies Te, (b) in accordance with the way divided in the unit video material, each unit video bodies
Mutually different strength of the magnetic field determines a gradient magnetic field to be applied to the three-dimensional image space each, by applying a radio wave pulses (c) the three-dimensional image space, (d) O to the falling edge of the radio wave pulse At the same time, the gradient magnetic field determined in (b) is applied to the three-dimensional image space. A nuclear magnetic resonance imaging apparatus that collects image data assuming the generated signal, and (f) reconstructs an image using the image data collected in (e). 2. A nuclear magnetic resonance imaging apparatus for obtaining an image from the nuclear magnetic resonance signals in a three-dimensional image space obtained by superimposing a temporally constant gradient magnetic field in the static magnetic field, its (a) the three-dimensional image space Divide into multiple in each direction
Divided into a plurality of unit video bodies Te, (b) in accordance with the way divided in the unit video material, each unit body
Mutually different strength of the magnetic field determines a gradient magnetic field to be applied to the three-dimensional image space, (c) gives the gradient magnetic field in the three-dimensional image space, (d) the 3-dimensional this gradient magnetic field is applied (E) Nuclear magnetic resonance signals generated from each unit image body by applying the radio wave pulses to a signal generated from a lattice-like point having a uniform density of nuclear magnetic resonance nuclei. (F) A nuclear magnetic resonance imaging apparatus that performs image reconstruction using the image data collected in (e) above. 3. The gradient magnetic field is determined so that the frequency spectrum of each lattice point does not overlap when the nuclear magnetic resonance signal generated from each unit image is regarded as a signal generated from the lattice point. The nuclear magnetic resonance imaging apparatus according to claim 1, wherein 4. An applied radio wave pulse having an excitation angle of 9
The nuclear magnetic resonance imaging apparatus according to any one of claims 1 to 3, wherein the angle is 0 °. 5. The nuclear magnetic resonance imaging apparatus according to claim 1, wherein the method of dividing the unit image bodies and the determination of the gradient magnetic field are set in advance.