JPS60151548A - Method and apparatus for inspection by means of nuclear magnetic resonance - Google Patents

Abstract

PURPOSE:To perform the high speed collection of NMR data, by impressing the gradient magnetic field for satisfying a prescribed condition before and after 180 deg. pulse impression. CONSTITUTION:Current is passed through a coil 1 for static magnetic field from a control circuit 2 to give a static magnetic field H0. Current is passed through a coil 31 for Z gradient magnetic field from a controller 20 through a control circuit 4, 90 deg. X pulse is given while giving Z gradient magnetic field Gz<+> to select and excite a body to be inspected. A magnetic field Gz<-> is impressed successively to G<+> impressing. Next, magnetic fields Gx, Gy are impressed for a prescribed time. Next, the magnetization is turned over by 180 deg. pulse through a gate circuit 30. Above operations are repeated so that the fields Gx, Gy before and after 180 deg. pulse impressing satisfy next equation. gxpXtmp=g'xpXt'mp, gypX tmp=g'ypXt'mp, p=1-n, n is number of 180 deg. pulse, g, g' are intensity, x y are phase of RF pulse, tmp, t'mp are impressing time of gradient magnetic field before and after 180 deg. pulse impressing. Thereby, high speed collection of NMR data is made possible.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field to which the invention pertains] The present invention relates to nuclear magnetic resonance (nuclear magnetic resonance).
tic1゛QsOnanco) (hereinafter referred to as rNM
It is abbreviated as Rj.

) This relates to nuclear magnetic resonance testing methods and devices that utilize phenomena such as the distribution of specific atomic nuclei within a subject from outside the subject, and is particularly suitable for medical equipment. Concerning improvement of imaging equipment.

[Prior Art] First, an outline of the principle of NMR will be explained.

The atomic nucleus consists of protons and neutrons, which are considered to be rotating as a whole with a C1 nuclear spin angular motion of I.

Figure 1 shows the hydrogen nucleus ('l-1), which consists of one proton p as shown in (a) and has a spin of 1.
The rotation is expressed by the number of children 1/2.

Here, the proton p has a positive charge e+ as shown in (b), so as the nucleus rotates, the magnetic moment μ
occurs. In other words, each hydrogen nucleus can be thought of as a small magnet.

Figure 2 is a diagram schematically showing this point. In a ferromagnetic material such as iron, the directions of these micromagnets are aligned as shown in (a), and magnetization is observed as a whole. Ru. On the other hand, in the case of hydrogen, etc., the direction of the micromagnets (the direction of the magnetic moment) is random as shown in ([1)], and no magnetization is observed as a whole.

Here, when an eqA lift of 1-1° in the Z direction is applied to such a substance, each atomic nucleus is aligned in the Ho direction. That is,
The nuclear energy unit is quantized in the Z'IJ direction.

Figure 3 (a) shows this situation for a hydrogen nucleus. Since the spin hanger number of a hydrogen nucleus is 1/2, as shown in Figure 3 (b), -1/2 and + 1/2
It is divided into two energy units. The energy difference Δ' between two energy levels is expressed by equation (1).

ΔE-γ'hHo...(1) Where, γ is the gyromagnetic ratio 'hh-h/2π h is a blank constant.
H.

Since a force of <r is applied, the atomic nucleus precesses around the Z axis at an angular velocity ω, as shown in equation (2).

ω-γHa (Larmor angular velocity)...(2) When an electric wave (usually a radio wave) with a frequency corresponding to the angular velocity ω is applied to the system in this state, resonance occurs, and the atomic nucleus is expressed by equation (1). J channel-i! corresponding to the energy difference ΔF! −
It absorbs and transfers to a higher energy unit. Even if several types of atomic nuclei with nuclear spin angular motion are mixed, each nucleus has a different gyromagnetic ratio γ, so the resonant frequencies differ, and therefore only the resonance of a specific atomic nucleus can be extracted. Also, measure the strength of the resonance tt'l
For example, we can also know the existence of atomic nuclei. Furthermore, the atomic nucleus excited to a higher level returns to a lower level after a time determined by a time constant called relaxation time after resonance.

This relaxation time is the spin-lattice relaxation time (longitudinal relaxation time)
T+ and spin-spin relaxation time (transverse relaxation time > T2
By observing this relaxation time, data on material distribution can be obtained. Generally, in a solid, the spins are almost fixed at fixed positions on the crystal lattice, so it is inevitable that interactions between the spins will occur. Therefore,
Relaxation 11! T2 between i is short (, energy 1 obtained by nuclear magnetic resonance
! - first goes to the spin system and then to the lattice system. Therefore, the time T1 is significantly larger than T2. On the other hand, in a liquid, molecules move freely, so the likelihood of energy exchange occurring between spins and between spins and the molecular system (lattice) is about the same. Therefore, time T+
and T2 have approximately the same value. In particular, the time T1 is a time constant that depends on the way each compound binds, and the time T1 is a time constant that depends on the way each compound binds. It is known that the values differ greatly depending on the tumor.

Here, we have explained the ice-water nucleus (T1), but it is possible to perform similar measurements with other nuclei that have nuclear spin angular motion, and in addition to hydrogen nuclei, phosphorus nuclei (''P), It is applicable to carbon nuclei (IBC), sodium to lithium nuclei (Na), fluorine nuclei (19F), 11 atoms <17o), etc.

In this way, NMR allows the abundance of specific atomic nuclei and their relaxation times to be measured, so by obtaining various chemical information about specific atomic nuclei within a substance,
Various tests can be performed within the subject.

Conventionally, an inspection device using NMR has been used to excite the protons of a virtual cross section of the subject using the same principle as X-ray CT, and to detect the NMR corresponding to each projection.
There is a method (6) in which MR resonance signals are measured in many directions of the subject and the NMR resonance signal intensity at each position of the subject is determined by a reconstruction method.

FIG. 4 is an operational waveform diagram for explaining an example of an inspection method in this conventional apparatus.

First, a Z gradient magnetic field Gz+ as shown in FIG. 4 (ITJ) and an RF pulse (9o° pulse) with a narrow frequency spectrum (f) as shown in (A) are applied to the subject. In this case, Only the surface where the Larmor angular velocity ω = γ (Ho + △Gz) is excited, and the magnetization M rotates at the angular velocity ω as shown in Figure 5 (A). For example, the direction is changed by 90 degrees in the y' axis direction.Subsequently, as shown in Figure 4 (c) and (d),
Add a gradient magnetic field Gx and a y gradient lit & field G, thereby creating a two-dimensional gradient magnetic field, and perform NMR as shown in (e).
Detect resonance signals.

Here, as shown in Figure 5 (0), the magnetization N4 gradually disperses in the direction of the arrow in the x J y l plane due to the inhomogeneity of the magnetic field, so the N'MR resonance signal eventually decreases. , and disappears after Ta2 as shown in FIG. 4(e). The thus obtained NMI transforms resonance No. 18 into a Fourier transformation yiI! If υ, it becomes a projection in the direction perpendicular to the gradient magnetic field synthesized by the X gradient magnetic field Gx and the y gradient magnetic field Gy.

Thereafter, after waiting for a predetermined time Td, the next sequence is repeated in the same manner as described above. In each sequence, GxXGy is changed little by little.

By this, NMR corresponding to each projection
Resonance signals can be observed in numerous directions of the object.

By the way, in the conventional device A5, which performs this kind of operation,
As shown in FIG. 4, the time T5 until the NMR resonance signal disappears is 10 to 20 mS, but the predetermined time Td until moving on to the next sequence is 1 because of the relaxation time T+.
Sea Pi! degree is required. Therefore, if a cross section of a single object is reconstructed using, for example, 128 projections, the measurement would require a long time of at least 2 minutes, which is one of the major obstacles to achieving high speed. It is one. .

[Purpose of the Invention] In view of the above, the purpose of the present invention is to observe a large number of echo signals caused by RF pulses1, forcefully return the magnetization to a thermal equilibrium state, and shorten the retention time. Inspection method and inspection using NMR that can record data on M@t! - It's about providing.

[Summary of the Invention] In order to achieve such an object, the present invention includes the steps of: applying a first 90° pulse to cause nuclear magnetic resonance to the nuclei of atoms constituting the tissue of the subject; 9 (after application of 1' pulse, predetermined aI interplanar spacing 0
The process of repeatedly applying 180° pulses, #J 18
a step of applying a gradient magnetic field after reaching 0° pulse application; and li! of the 180° pulse. After the turning process, when n is an odd number, apply a 90° pulse followed by a 180° pulse, and when n is an even number, apply only a 90° pulse; a step of moving to the next step after a predetermined waiting time; a step of measuring a nuclear magnetic resonance signal generated under the applied magnetic field; 'I reconstructing an image related to the weave, and in the step of applying the gradient magnetic field, the following relationship is established before and after each application of the repeated 180° pulse. Features.

Qxp xtml)=ol Xpxi,' ml)Gy
p X1rnp =Q' yp X'j' mpHere, subscript p=1. 2. ,,,, n tmp is the gradient magnetic field application time during Tsns before applying the 180' pulse.

t'mp is (
gradient magnetic field application time.

Qxp is the magnitude of the X-axis gradient magnetic field at tmp.

(Jyp is the magnitude of the y-axis gradient ml at tmp.

Q' xp is the magnitude of the X-axis gradient magnetic field at Mmp.

Q'yp is the magnitude of the y-axis gradient m field that is equal to t'mp.

[Example] A detailed explanation of the present invention will be given below with reference to the drawings. , FIG. 7 is a block diagram showing the configuration of an embodiment of an apparatus for implementing the method of the present invention. In the figure, 1 is a static magnetic field coil for generating a uniform static magnetic field H9 (the direction in this case is the Z direction), and 2 is a control circuit for this static magnetic field coil 1, such as a DC Contains a regulated power supply. The density Ho of the magnetic flux generated by the static magnetic field coil 1 is 0.
It is desirable that the uniformity is about 1T and the uniformity is 101 or more.

Reference numeral 3 generally indicates a gradient m-field coil, and 4 indicates a control circuit for this gradient magnetic field coil 3.

FIG. 6(a) is a configuration diagram showing an example of a cocoil for gradient ms, in which a coil 31 for Z gradient magnetic field, a coil 32 for y gradient magnetic field.
．． 33. Although not shown, a y-gradient magnetic field coil 32.
33 have the same shape and include an X-gradient magnetic field coil that is rotated by 90°.

This gradient magnetic field coil generates a magnetic field having linear gradients in the x-, y-, and z-axis directions in the same direction as the uniform static m-field Ho. The control circuit 4 is controlled by =1-ra 20 (details will be described later).

Reference numeral 5 denotes an excitation coil that provides an RF pulse with a narrow frequency spectrum f as an electromagnetic wave to the subject, and its configuration is shown in FIG. 8 (b).

6 is the frequency corresponding to the NMR resonance condition of the atomic nucleus to be measured (for example, for protons, 42.6 MHz, /T
), whose output is the control [1
-Goo 1~ whose opening and closing are controlled by the signal from La 20
A signal is applied to the excitation coil 5 via a circuit 30 (details will be described later) and a power amplifier 7. Reference numeral 8 denotes a detection coil for detecting an NMR resonance signal in the subject, and its configuration is the same as the excitation coil shown in FIG. Although it is desirable that this detection coil be installed as close to the subject as possible, it may also be used as an excitation coil if necessary.

Reference numeral 9 denotes an amplifier for controlling the NMR resonance signal width obtained from the detection coil 8, 10 a phase detection circuit, and 11 a signal from a wave memory circuit 11 that stores the phase-detected waveform signal from the amplifier 9, for example, through an optical fiber. A computer inputs the data through a transmission line 12 and performs predetermined signal processing to obtain a tomographic image, and 14 is a display of a television monitor that displays the obtained tomographic image.

The controller 1-○-ra 20 is a gradient magnetic field Gz, Gx.

It is configured to be able to output the signals (analog signals) necessary to control Gy and RF modulated signals, and the control signals (digital signals) necessary for transmitting RF pulses and receiving NMR multiplied signals. be.

FIG. 8 is a diagram showing an example of such a controller, which is particularly capable of high-speed control and allows easy changes in control sequences and analog waveforms. In the figure, 221 is a control circuit for writing data sent from the operator console 210 or the combination coater 13 via the operator console into each memory; 222 22! i, 228. 231 is a waveform storage memory in which waveform data of the x, y, z gradient signal and modulation signal given from this write control circuit are respectively occupied; 223; 226, 229, 232 are waveform storage memories 222. 224i, 228.
224. Latch circuit for temporarily holding the waveform data output from 231, respectively; 227. 230. 233 is this latch circuit 223. 226. 229. 232
1] and 8 converters convert the outputs from DA to DA, respectively. X2. V2. Z2. M2 is the DA! Grind circuit 224
, 227. 230. x, y, z gradient signal outputs and modulation signal ratio horns output from 233, respectively.

234.236. 238. 2401; t, transmitting/receiving circuit control signal, ie, AD conversion control signal, transmission gate control signal;

The data of the reception gate control signal 2-phase selection signal (a signal for selecting one type of pulse from among four types of RF pulses having mutually different phases) is written from the write control circuit 221, respectively. waveform storage memory, 2
3ri, 237. 239. 241 is this waveform storage memory 234. 23f'i, 238. Latch circuit that temporarily holds data output from 240, 2゜82, R
2, PS is this latch circuit 235. 237°239,
AD conversion control signal, transmission gate control signal output, and reception gate control signal output output from 241.

Phase selection (6 tan. 243 is each waveform storage memory mentioned above 222. 225. 228. 231. 234
．． 236. Read the contents of 238°240 to the latch circuit? The continuous output control circuit 242 sets the value of the continuous output/output start address during the operation or from the input/output (hereinafter simply referred to as the input/output unit), and also sets the value of the input/output start address to the value of the address. From the input/continuation control circuit.

Memory address register 2 that sequentially adds the given Hani+1 and outputs it as an input/output address.
44 indicates the number of output steps to be given from the computer, and 01 indicates the end of the output from the readout control circuit 243.
The output count register 245 is a one step length pulse generation circuit which generates a one step length pulse by counting the time length of one step (one step length) given from the combicoater.

The operation of such a controller is as follows.

(B) Loading and outputting operation In loading and unloading operation, the waveform data sent from the computer is loaded and outputted to the specified address of the waveform storage memory specified by the Combi coater. That is, first, memory address register 2
The write start address is set at 42. Is the data sent from the computer with the import/export command a write-in system? It is written to the address specified by the memory address register 242 in the waveform storage memory (eg, waveform storage memory 222) selected by the afl@path 221. After that, the write control circuit 221 automatically adds 1 to the memory address register 242 to make it the memory address for the next write. Sequentially write data to other waveform storage memories in the same manner as above.

(b) Read operation In the read operation, each memory is read in parallel. FIG. 9 shows an example of a time chart of read signal waveforms. The computer first sets the reading start address of the waveform storage memory in the memory address register 242. Next, the number of read steps is set in the output count register 244. In addition, the one step length (time per step during readout) is set in the pulse generation circuit 245 to be one step longer. Next, when the computer h and the other computers are instructed to start reading, the memory address register 242 indicates the waveform record 1d memory 222 at the address V. 22.5. 228. 23
1. 234. 236. The contents of 2311° 240 are read out to /all, and when all the data is available, the read control circuit 243 sends the contents to the latch circuits 223° 226, 22
9. 232. '235. 237. 239. 2
A latch pulse is output to 41 to latch the data. Next, the value of the memory address register 242 is incremented or decremented by 1.

If the output count register 244 indicates completion,
Read control circuit 243 hX et al. latch circuit 223. 2
26°229, 232. 235. 237. 23
9. A Clino pulse is output to 241 and the read operation is completed. When the output count register 244 has not finished, 1 is subtracted from the output count register 244, and the process moves to a tangent reading step in which the output from the 1 step length pulse generation circuit 245 waits for 1 step time length. Thereafter, the same process can be repeated to read out a waveform as shown in FIG. 9, for example. X, y, z gradient X2. V2
．． /'2 and modulation signal M2 further increase the latch circuit output.
) A converter 224. 227. 230. It is an analog signal obtained by DA conversion at 233, and the modulated signal M2+ is sent to the gate circuit 30, and x, y.

The Z gradient signals are each led to a control circuit 4 for the gradient magnetic field.

Since such an agent is equipped with dedicated hardware such as a waveform storage memory, it is possible to read and output a large amount of data at high speed. Furthermore, the contents of the waveform storage memory can be rewritten as needed, and any analog or digital signal waveform can be output.

Furthermore, it is easy to use part of the signal waveform (which is often actually used) by appropriately providing the read start address and the number of read steps.

Goo 1 to circuit 30 receive the RF multiplied signal from the oscillator 6,
In response to this, four types of signals having different phases by 90 degrees are generated, and one of the four types of signals is generated based on instructions from the controller 20.
1 is selected and further modulated with an RF modulation signal to obtain a drive signal for the excitation coil 5. The detailed configuration is shown in FIG. In the same figure, 311 is the input R'
A 90° phase shifter capable of obtaining two signals with a phase shift of 0° and 90° relative to the signal, 312. 313 is a 180° phase shifter that can simultaneously obtain two signals with a phase shift of 0° and 180° relative to the input signal. 90” as shown
By applying each output of the phase shifter 311 to each of the 180° phase shifters, O', 180°, 9
0”.

A signal with a phase difference of 270° is 1q. Each of these signals is connected to a high frequency switch (for example, a double balanced mixer (DBM) can be used) 314.
-317 to the combiner 321, and the four signals are added together. This base and high frequency switch are individually energized by the output of the decoder driver 320, and the decoder driver 320 is formed by decoding the phase selection signal ps given from the controller 20.
Any of the four output bays (<x, Y, -X, LY) is 1
/7 is active. In response to this, only the switch corresponding to J becomes conductive (the other three switches are non-conductive). Therefore, only one signal is input to the combiner 321.

The output of the coupler 321 is inputted to the modulator 323 after passing through the amplifier 322, where the RF modulation signal (pulse signal) received from the controller 20 is used to control the rotation of the magnetization M by the pulse width and peak value of the signal. ) Accumulate.

), and for example, the signal is modulated into a Gaussian waveform i as shown in (a) of FIG. 4 and output.

According to the gate circuit having such a configuration, based on one RF multiplied signal, an RF signal having a phase difference of 4i'!, 0'', 90', 1806. There is an advantage in that it is extremely easy to selectively select a desired signal and modulate the signal with a desired waveform at an appropriate timing.

The operation of the apparatus of the present invention configured in this way can be explained using the two-dimensional PR (Projection Recons) shown in FIG.
The steps will be explained step by step with reference to a time chart for the case of using the tru-Cth on ) method.

1) Time t.

A current is passed from the control circuit 2 to the static m-field coil 1, and the static magnetic field 1 is applied to the test object (the test object is installed inside the cylinder of each coil). is applied, the controller 20 controls the control circuit 4
A current is passed through the Z gradient magnetic field coil J31 through the first
When a 7-gradient magnetic field Gz+ is applied as shown in Figure 1 (b), a first 90''X pulse is applied to b as shown in Figure 11 (a) to selectively excite the subject.

2) Following the above-described application of Gz-, Gz- is applied.

This is to match the phase of the NMR resonance signals of different parts of the object, and this technique (this is a known technique). Cover with 12.

3) After that, the time Gx and Gy of ti+ are respectively Gx+
, (apply with the magnitude of JX2.

4) Phase difference 1 selected and output in circuits 1 to 30 after rs+ time from the application of the first 90°X pulse.
The subject is excited with a 12F signal modulated into an 80° rectangle.

Before and after this 180°y pulse, (m
1) Add homogeneity spoil pulses at Gx, Gy, and Gz as in (d). Due to this homogeneity spoil pulse, 180° pulse irregularity Me
It is possible to suppress the noise caused by the noise.

Here, the subscripts x and y of 90° and 180° indicate the phase of the RF pulse, and X and y differ by 90° (Q phase).

4) Next, when Gx and Gy are set to Ω'x++Q'y+ as shown in (c) and (d) of the same figure, a spin echo signal is generated as shown in (b) of the same figure. Ox+ xtm,=Q' x+ x-i' m+Qy+
The echo signal reaches its maximum when Xl:ml=Q'y+Xj'ml.

5) Next, change the magnitudes of Gx and Gy to gx2+Qy2 and repeat the same operations as 2) to 4) above.

At this time, it is necessary to satisfy the following formula.

GIxpX'l: mp=Q'xp'Xi' mpQy
pXjmp=Q'ypXi' mp here, subscript 1)
−1, 2, , , n is the number of 180° pulses between the first 906 pulse and the second 906 pulse (details will be described later).

6) For a predetermined number of times n (after giving a 1180° pulse, at the timing when the echo signal is at its maximum (at the end of t' mn), the magnetization vector is selectively set by a 90" pulse and Gz, when n is an odd number. is directed in the two-axis negative direction (downward), and 14, where n is an even number, is directed in the Z-axis direction (upward).

7) Only when n is an odd number, then turn the entire magnetization vector upward with a pulse 180'.

8) After these, Bomogenity in Gx, Gy and Gz
Gives spoil pulse. By applying this pulse, correlation with the next sequence can be eliminated.

9) Wait for Td time and repeat the same diatance.

In such a sequence, each time parameter Ts
p, T'sp, 'Td, and n are appropriately selected depending on the respective usage conditions.

L: The movement of the magnetization M in the sequence described above is measured at the center of the slice plane (by applying a 90° pulse, the magnetization f
The part where vl correctly rotates 90 degrees), the slice and boundary (
When a 90° pulse is applied, the magnetization M rotates θ゛, and when a 180° pulse is applied, Gz=O, so the part rotates 180°), outside the slice plane (not affected by a 90° pulse application, and the 180° pulse 12 and 13 for each part of the part where the direction of magnetization M is reversed by
Accept the illustration.

Figure 12 shows when n is an odd number, and the final 90°X pulse turns the total magnetization downwards (after that, it turns upwards at 180°X).The slice boundary ((c) in the same figure) is 90°. θ with pulse
It rotates only 90° (0<θ), but the second 90°
Just before the pulse, θ from the negative direction of the two axes.

, so it is oriented upward at 180°-X.

No. 131-A is when n is an even number, and just before the second 90° pulse, [since θ from the Z-axis stop direction, it is sufficient to point upward at 90°-X.

NMR signal obtained in such a sequence (first
(E) in FIG. 1 is Jζ resampled in the wave memory, and the data is subjected to two-dimensional image reconstruction processing in the combicoater 13, and a cross-sectional image of the object is displayed on the display 14.

It should be noted that the present invention is not limited to the above-mentioned embodiments, and may be applied to various methods or systems such as those described below.

1) As shown in FIG. 14, n=2. gx, >>G'x
++(Jy+>CJ'y++GX2(Q'X2゜Gy2
(Set as Q'y2. In this case, the signals of periods Ts+ and T'52 are affected by RF pulses and noise from Gx, Gy, G2, etc., so they are not used, and the signals of periods T'S+ and Ta2 are used. ηru.

According to this method, I' m + <
: t'm+ and tm2>t'm2, so s'ml
It has the advantage that the Xtm2 time is long and signals can be recorded with good S, -'N.

2) When applied to the spin warp method, as shown in FIG. 15, I:mp (p = 1 to n) is kept constant and Ox+ , , , +CJxr+ are successively changed, Measure the NMR signal.

3) A case where the Fourier method is applied, in which Qxp (p-1 to 1)) is kept constant in FIG. 15, and tmp is successively changed.

7I) is applied to the 1-planar method, and the method is as shown in FIG. 16 (the diagram when n is an odd number). The conventional echo planar method uses magnetic field reversal (e.g. Gy
However, in the present invention, it is performed by applying a 180° pulse. Finally, the magnetization M is forcibly turned upward (when n is an odd number, use a 90° pulse and a 180° pulse; If the number is even, use a 90″ pulse).

5) When applied to the selective excitation line method, Figure 17 (diagram when n is an odd number)
The method shown in .

6) Inversion recovery method. In other words, it is a method in which a 180° pulse is added before a series of pulse sequences, and a homogeneity spoil pulse is added here to prevent adverse lateral effects caused by the inaccuracy of the 180° pulse. Gx.

It is commonly added to Gy and Gz. However, the application of the body spoil pulse is not necessarily necessary and can be omitted if the adverse effect in the lateral direction does not occur or can be ignored.

This inversion recovery method can be applied to the above-mentioned total method, and both cases with emphasis on T+ (longitudinal relaxation time) can be obtained.

7) For each method listed above, use J5 to generate a plurality of non-selective 180° pulses. For example, 18
90°, 180°y and 90″X instead of 0°9
pulse train is used. This makes it possible to cancel out inaccuracies in pulse intensity.

8) A method in which a T+ image, a T2 image, a spin-density image, and a combination thereof are obtained through inter-image calculation.

For example, when n=1, the signal strength V in the sequence shown in Figure 11 is r:r:'', and MU uses the spin@return relationship to change Ds+, T's+, and Td. Calculate from multiple images.

9) Apply multi-slice method. That is, using the waiting time of Td, other planes are excited and information about them is obtained.

10) The phase relationship of the pulses in each of the above methods is C+
O"z-180'4-180'l ---'(O'--
c (Constant Eye) QO''-1180'1 180%-'I
O'z180'-E0+O"-x-(+8oNie18
0 degree engineering) A-----qo 212 (r+: 4? Interpretation) qOF: C (180'-　180z.) 4--
---qo2z180-c [Effects of the Invention] As explained above, according to the present invention, after observing a large number of spin echoes, the magnetization is forcibly returned to a thermal equilibrium state,
Since the entire magnetization M is all directed upward (Z-axis stopping direction), the next operation can be started with a short waiting time Td, and the overall scan time can be shortened.

In addition, in order to improve image quality, it takes much less time to collect a large number of similar images and average them as time-series data, or to convert them into images and then average them, compared to conventional methods. The result is that it can be done in time.

Furthermore, the following effects are also exhibited.

G) Before and after the non-selective 180° pulse, the magnetization "above the For example, in Fig. 13, at t and -t2, the magnetization vector is pulled upward by T+, and the circle button in Fig. 13 (c) becomes smaller, and from t3 to
At t4, the magnetization of T1 is pulled upward, and the same figure (c)
The cones are in the expanding direction and cancel each other out.

■The homogeneity spoil pulse causes the components of magnetization to become opaque and there is no correlation between scans, so the magnetization moves correctly and there is less noise.

■When n=2, both TS + and T' 52 (T' s
+ 10Tsp) by making it sufficiently smaller than
It is possible to eliminate the influence of noise etc. on the applied signal, and at the same time, it is possible to eliminate the influence of noise etc. on the applied signal. By shortening T'S2, a high level NMR signal can be obtained.

(By using multiple pulses for the 180° pulse,
Since errors such as strong tW are canceled, the magnetization rotates correctly.

■You can easily obtain an image that suits your purpose by using inter-image calculations.

■) With multi-slice, the speed can be further increased by 11 times.

[Brief explanation of drawings]

Figure 1 C is a diagram explaining the spin of a hydrogen atom, Figure 2 is a diagram schematically showing the magnetic moment of a hydrogen atom, and Figure 3 is a diagram explaining a state in which 10 k atoms of a hydrogen atom are aligned in the direction of the magnetic field. ,
Fig. 4 shows an example of the test pulse waveform by conventional NMR, Fig. 5 shows the magnetization M on rotational coordinates, and Fig. 6 shows an example of the test pulse waveform by conventional NMR.
7 is a structural diagram showing an example of the LA field coil, and FIG.
, FIG. 9 is a diagram for explaining the operation of the controller 20, FIG. 10 is a diagram of the gate circuit 3o,
FIG. 11 is an operation waveform diagram for explaining the sequence according to the present invention, FIGS. 12 and 13 are schematic diagrams showing the direction of magnetization in the operation shown in FIG. 11, and FIGS. FIG. 7 is a diagram for explaining other embodiments or adaptations of the invention. 1, , , l11111JU 2l -i' Jl
/, 2.4. , , il field i1J control circuit, 301.
Gradient magnetic field coil, 519. Excitation coil, 6111 oscillator, 849. Detection coil, 10, phase detection circuit,
11. , , wave memory circuit, 13. ,, Combi coater, 20. ,. controller, 30. ,, Goo 1~Circuit. Showa date i Case description Patent application 1987-7707 July 2 Title of the invention Inspection method and apparatus by nuclear magnetic resonance 3 Relationship with the person making the amendment A'1 Patent applicant 4, attorney A1 Address: 172-9-32 Yokogawa Hokushin Electric Co., Ltd., Musashino-shi, Tokyo (0422) (54) 11116, “Title of the invention”, “Scope of claims” of the specification to be amended; The columns of "Detailed Description of the Invention" and "Brief Description of the Drawings" as well as Drawing 7 and Contents of Amendment 1) The full text of the specification m will be corrected as shown in the attached sheet. 2) Correct Figure 13 of the drawings as shown in attached Figure 13. 3) Correct Figure 18 of the drawings as shown in attached Figure 18. 4) Figures 19 to 28 of the attached sheet after Figure 18 of the drawings
Add diagrams. Specification n# 1, Name of the invention, Nuclear magnetic resonance testing method and apparatus 2, Claims: Original U! u! The nuclear magnetic resonance signal g generated under I is expressed as (2
) In one culm to which the gradient magnetic field is applied, a part of the gradient magnetic field applied before and after the repeated 180° pulse is strengthened so that the application time is shorter than the application time of other gradient magnetic fields, and together with this, Measure the four nuclear magnetic resonance curves under an applied magnetic field, and reconstruct both edges related to the tissue of the subject based on the obtained signals.
J. An examination method using nuclear magnetic resonance according to claim 1, characterized in that only echo signals are observed during another gradient magnetic field application period that is longer than the short gradient field application period. Scolding IH'& LL, ``Pestle'' 1 person shi-sai - T applying walking distance
Then, each of the 180° pulses of ty k'-2 h repetitions is 4fi! IIc of the slice when applying the first 906 pulses
N>'s slope! 14G+ is generated, and the gradient magnetic field of the two-part, the second, and the second and third and third-party encodes the gradient magnetic field 9x. Inspection method based on ψ゛゛p of Claim 1, Clause 1, 8゛. A uniform image is obtained from multiple images obtained by
In the spin J image and the combined image of these, the man J:ru 4 Tadasu Moriue''L engineering■U;. The wave pulse ζ is the first 906 pulses. 3. Detailed description of the invention (industrial application field) This invention GJ, nuclear magnetic resonance (formerly + clear magnet
etic resonance) (hereinafter referred to as rNM)
NMR that utilizes the phenomenon (abbreviated as RJ) to determine the distribution of specific atomic nuclei within the specimen from outside the specimen.
The present invention relates to an inspection method and an apparatus therefor, and particularly relates to an improvement of an NMR imaging apparatus suitable for medical intoxication. BACKGROUND OF THE INVENTION An NMR imager places a living body, usually a patient, in a magnetic field. Then, a predetermined pulsed W6 magnetic wave is applied to the living body to excite only a specific atomic nucleus of interest among the various atoms that make up the living body. Once excited, the atomic nucleus returns to its original energy state, but at this time, it emits the absorbed energy to the outside as electromagnetic waves. In an NMR imager, this emitted magnetic field is detected by a coil. This detection signal is a nuclear magnetic resonance signal (NMR I=... echo signal and F
ID (free 1induction decay)
＞) and contains various information about the target atomic nucleus. The NMR image @ device is a device that analyzes this and visualizes a part of the living body as a tomographic image, which is useful for diagnosis, treatment, etc. of the living body. First, an outline of the principle of NMR will be explained. The atomic nucleus consists of protons and neutrons, and the nucleus as a whole is considered to be rotating (rotating) with a nuclear spin angular momentum r. Figure 1 shows the hydrogen nucleus ('l-1), which consists of one proton P, as shown in (a), and rotates as expressed by the spin quantum number 1/2. . Since the proton P has a positive charge e+ as shown in (b), magnetic moments 1 to μ are generated as the atomic nucleus rotates. In other words, each hydrogen nucleus can be thought of as a small magnet. Figure 2 is an explanatory diagram that schematically shows this point. In a sealed body, the directions of these micromagnets are aligned as shown in (a), and magnetization can be observed as a whole. On the other hand, in the case of hydrogen, etc., the magnetization is random as shown in the direction of the micromagnets (the direction of the II magnetic moment), and no magnetization is observed as a whole. Here, when a static magnetic field 1-10 in seven directions is applied to such a substance, each atomic nucleus is aligned in the H6 direction. Figure 3(a) shows this situation for a hydrogen nucleus. Since the spin quantum number of a hydrogen nucleus is 1/2, J shown in Figure 3 (b): -1/2 and +17
It is divided into 72 two energy units. The energy difference ΔF between two energy units is expressed by equation (1). /\E-γ is also Llo' (1) However, γ is also the rotational ratio (a constant specific to each atomic species) -h
/2y? h is a blank constant, where each nucleus has a static value of 11'A HO,
μX H. As a result of this force, the atomic nucleus precesses around the Z-axis at an angular velocity a) as shown in equation (2). m=71''o (Larmore own speed r!IL>(2>i.e.
Each type of atomic nucleus precesses at a different Larmor angular velocity ωm. In this way, the living body placed in static Ilml-10,
For example, the frequency corresponding to Larmor angular velocity ω1 (ft=ω
When an electromagnetic wave (usually a radio wave) of 1/2π) is applied, resonance occurs in the precessing atomic nucleus corresponding to this frequency f1, and the atomic nucleus has an energy outside Δ shown by equation (1).
It absorbs the energy equivalent to F and transitions to a higher energy unit. Here, one body is usually composed of multiple types of nuclei, but under the environment of static fa field I'o, the applied frequency f
Only one type of atomic nucleus resonates with one electromagnetic wave. Therefore, by selecting the strength of the static magnetic field 1'0 applied to the living body and the applied frequency f, it is possible to extract only the resonance of a specific type of atomic nucleus. By measuring the strength of the resonance, we can determine the amount of nuclei present. In addition, the atomic nucleus excited to a higher level is
After resonance, it returns to a lower unit after a time determined by a poetic constant called relaxation time. At this time, the absorbed energy is released to the outside, so by measuring the temporal change in the resonance intensity, the time described below can be determined. The relaxation time is spin-lattice relaxation time (longitudinal relaxation time) T+
and spin-spin relaxation time (transverse relaxation time>T2). Data on material distribution can be obtained by observing this relaxation time. In general, in solids, the transverse relaxation time T2 is short, and nuclear magnetism The energy obtained by resonance first spreads to the spin system and then to the lattice system.Therefore, the longitudinal relaxation time T1 is significantly larger than T2.On the other hand, in a liquid, molecules move freely. Therefore, the likelihood of j-energy exchange occurring between spins and between spins and the molecular system (lattice) is about the same.Therefore, the times T1 and T
2 becomes almost equal value. Here, we have explained the hydrogen nucleus ('I'),
Similar measurements can be performed on other atomic nuclei with nuclear spin angular motion, such as the phosphorus nucleus ("P").
, carbon nucleus (C), sodium nucleus (5Na), etc. In this way, by using NMR, it is possible to measure the abundance of specific atomic nuclei and their relaxation times, so by obtaining all the chemical information about specific atomic nuclei within a substance,
Various tests can be performed within the subject. Conventionally, the PR method (PRO method) has been used as a method of obtaining images regarding the tissue of a subject using such NMR phenomenon.
injection reCOIIStrUClioll
metllOd (also called projection restoration method). The principle of image reconstruction using this PR method is
The principle is almost the same as that of the device. First, the body axis direction of the subject (
By applying a gradient magnetic field (in the Z-axis direction), the virtual sliced portion (z
Excite the plane perpendicular to the axis.Although we will explain the case where the tomographic plane is perpendicular to the body axis of the subject, any plane can be selected by changing the gradient magnetic field. Next, the gradient Ii field is applied in the X and y directions, respectively. In this state, the NMR beam is detected, and the projection in the direction perpendicular to the x and y composite gradient!i field is waited. And x, y
By repeating the operation of changing the value of the composite gradient magnetic field, obtaining NMR signals corresponding to the NMR signals, and performing Fourier transform on each, projections are determined in many directions of the subject. The PR method is a method of constructing 100 images of a subject using the 01 method using this projection. FIG. 4 is an operational waveform diagram for explaining an example of an inspection method for a conventional device using this PR method. First, a Z gradient magnetic field Gz+ as shown in FIG. An RF pulse (90'' pulse) is applied. A gradient magnetic field G2 is applied in the 7-axis direction (body axis direction) of the living body, and the protons precess at a period proportional to the strength of the magnetic field. Here, the position of the Z axis (1-1
0+ΔGz).() is the applied RF
Prone is precessing at the same Larmor angular velocity ω, -γ(1'o+ΔGz) as the frequency of the pulse (ωJ=2πfJ). Therefore, protons (), which are precessing at an angular velocity near this frequency as the center frequency, are excited. That is, the gradient magnetic field Gz in the Z-axis direction acts to determine the position of the slice plane of the living body. If the magnetization M of the excited proton is shown on a rotating coordinate system rotating at an angular velocity ωJ as shown in FIG. Subsequently, as shown in FIGS. 4(c) and 4(d), an X gradient magnetic field OX and a y gradient magnetic field Gy are applied simultaneously. A two-dimensional gradient I1m is created by combining these two gradient magnetic fields, and under this environment, an NMR signal shown in (e) is detected. Here, as shown in FIG. 5 (1), the magnetization M changes due to the non-uniformity of the la field and the T2 relaxation.
'As it gradually disperses in the direction of the arrow within the plane, N
MR signal decreases 1. As shown in FIG. 4 (e), it disappears after a period of time Ts@. If NMR (:8) obtained in this way is subjected to Fourier transformation, it becomes a projection in the direction perpendicular to the gradient magnetic field synthesized by the X gradient magnetic field G X X'/gradient magnetic field Gy. After a time Td, repeat the following sequence with the same operation as above. In each sequence, the values of Gx and Gy are changed little by little, and the direction of the composite gradient magnetic field is varied. This makes it possible to obtain NMR signals corresponding to each projection in many directions of the object. (Problems to be solved by the invention) In the conventional apparatus that operates in this manner, as shown in FIG. As shown in , the time Ts until the N IVI P +g disappears is 10 to 20 ms, but the predetermined time Td until moving to the next sequence is the longitudinal relaxation time T. Therefore, about 1 SeC is required. Therefore, if a cross-section of a single object is to be reconstructed using, for example, 12,810 movements, the measurement will require at least two minutes, which is a major obstacle in achieving high speed. In addition, the application of the gradient-like i-field is sequentially switched, and the application time or amplitude of the gradient magnetic field is changed to
In the conventional NMRj!ii imager using the two-dimensional Fourier method, which uses a two-dimensional tool to obtain an image of the slice plane, Td is similarly 1 sec. The shortcoming is that the total measurement time is long.In order to solve this problem, the DFFT method (drivan equili
l+rium fourier transform)
When applied to an NMRiffii imager, the following drawbacks arise. In conclusion, N
It is inappropriate to use the DEFT method in the tvlRiiIiii image device, and there is no known technical example in which the DEFT method is used in the NMRiilftJA device. The DEFT method proposed for use in NMR analysis is described in ([Pulse and Fourier Transformation NMRJ Farrer, Beraker Name: Yoshioka Mtenj]. This DEFT method is a pulse sequence for speeding up. , (90°X...τ...iao'y...τ・
...90o-8...Td)". When performing two-dimensional imaging using this DEFT method, a 9° pulse (U, iπ selective excitation method (gradient magnetic field is applied at the same time) ) to excite only a specific slice plane, and there is no problem with IJ in this respect. However, for the 180° pulse, both selective and non-selective excitation are possible. , the distribution of the magnetization M2 on the Z-axis in the dynamic equilibrium state in the thickness direction of the slice immediately before the first 90° pulse is calculated using the equation of 31och, t, 44 The horizontal axis in Fig. 27 is the slice thickness direction in two directions, and the slice thickness is normalized to be in the range of ±1. In Fig. 27, DEFT method (pull 180°
In the case of selective and non-selective excitation of pulses and 'of the present invention'
Three similion results for the J combination are shown. here,
The 900 pulses were subjected to Thousandsian modulation for selective excitation. This (well, the average T1, T2 and TT of a living body =
It Eli was performed using 100 mS (repetition time). M2 is M before executing the pulse sequence
2 is set to 1, and the sensitivity corresponds to the NMR signal strength. (a) In the case of a non-selective excitation pulse of 18° in the DEFT method, as shown by the chain line A in Fig. 27, the MZ
becomes very small. In general, during the waiting time d of the pulse sequence, the same pulse sequence is sequentially applied to multiple slice planes in this way, and the original pulse sequence has a sufficiently long Td so that Mz is After T1 longitudinal relaxation and enlargement, the next pico (via
A multi-slice method is being used to perform NMR imaging for (w). This eliminates the decrease in the NMR signal (magnitude of MZ) and simultaneously reduces the data of the minority plane by 1q, so it is effective as a pseudo high-speed method. However, in the multi-slice method, the condition f1 is that Mz outside the slice plane is large without being influenced by excitation of other slice planes. Under these conditions, the DEFT method using a non-selective excitation 180° pulse has the drawback that the multi-slice method cannot be used in combination because the Mz outside the slice plane becomes small. The actual slice shape is obtained by multiplying Mz in FIG. 27 by a function of the slice shape (in this case, Gaussian shape), which is shown in FIG. 28. The horizontal axis in Figure 28 is the second
Same as Figure 7. (I)) Next, when considering the case of 180° pulse of selective excitation in the DEFT method, it can be seen as indicated by the dashed line B in FIGS. 27 and 28. In this case, Mz is not a big problem outside the slice plane as shown in Figure 27, but the slice shape is not shown in Figure 28. Yes, that's a problem. The reason why the three peaks are formed is that the magnetization at the slice boundary M IJ < When the 180° pulse of selective excitation is performed, a complicated yaw motion occurs, and the solid direction of each Mz becomes different, resulting in a decrease in the signal. be. As described above, the DE FT method, which is a well-known technique, can be used as the NMRii! It is unsuitable for use in II image devices. In view of these points, what is the purpose of the present invention? An object of the present invention is to provide an inspection method A3 using nuclear magnetic resonance that shortens the scan time without degrading the quality of images obtained, and an apparatus therefor. (Means for Solving the Problem) In order to achieve such an object, the present invention forcibly returns the magnetization to a thermal equilibrium state after observing a large number of NMR signals due to the R pulse. The present invention is characterized by being able to shift to the next pulse sequence with a short waiting time and making the overall scanning time faster. (Example) The present invention will be explained below with reference to the drawings. Fig. 6 shows , is a block diagram showing the configuration of an embodiment of the device according to the present invention. Shizuka I!l
Field coil 2 is a control circuit for this static magnetic field coil 1,
For example, it includes a DC stabilized power supply. The density Ho of the magnetic flux generated by the static 11 field coil 1 is 0. It is desirable that the ITN degree and the uniformity be 10゛4 or more. Reference numeral 3 generally indicates a gradient magnetic field coil, and 4 indicates a control circuit for this gradient magnetic field coil 3. In the apparatus of the present invention, first. A second gradient magnetic field is generated, but only the first. If it is explained simply as a second gradient magnetic field, it is abstract and the invention is difficult to understand. Therefore, in this specification, the first gradient 11 field will be described as a two-gradient field, and the second gradient magnetic field will be described as a composite magnetic field of an X gradient magnetic field and a y gradient magnetic field. However, this combination may be of any type, and the first and second
It suffices if the gradient magnetic fields of are in different directions. 3
1. The gradient magnetic fields in other directions other than the x, y, and z gradient magnetic fields may be combined (I may also be used. Also, at Mt. 111 during the Ice Age, as a means to generate the 1st and 2nd gradient ?ti field) , each with dedicated coil means (2
Coil for gradient magnetic field, coil for X gradient Il field. Although the explanation will be given using an example in which a y gradient magnetic field coil is provided, the present invention is not limited to this. That is, first, second
For example, one means can be used to generate a gradient magnetic field of the first . Both of the second gradient magnetic fields may be generated. FIG. 7(A) is a configuration diagram showing an example of the gradient magnetic field coil 3. FIG. The coil shown in FIG. 3A includes a Z gradient [a field coil 31 and a y gradient magnetic field coil 32 and 33. Furthermore, although not shown, a y gradient magnetic field coil 32
．． X, which has the same shape as 33 but is rotated 90 degrees and installed
It also includes gradient field coils. This gradient magnetic field coil 3 is in the same direction as the uniform static magnetic field 1-1°, x, y, z.
Generates a magnetic field with a linear gradient in each axial direction. The control circuit 4 is controlled by a controller 20. Reference numeral 5 denotes an excitation coil for applying a high frequency pulse with a narrow frequency spectrum f to the subject as an electric li wave, the configuration of which is not shown in FIG. 7(b). 6 is the frequency corresponding to the NMR conditions of the atomic nucleus to be measured (for example, for protons, 42.6 tv11-1 z
/ T > is an oscillator that generates a signal, and its output is
A gate circuit 30 whose opening/closing is controlled by a signal from the controller 20 is applied to the excitation coil 5 via the power amplifier 7 . 8 is a detection coil for detecting the NMR signal applied to the subject; its configuration is the same as the excitation coil shown in FIG. It is installed. The detection coil 8 is preferably installed as close as possible to the subject, but it may also be used as the excitation coil 5 if necessary. 9 is an amplifier that amplifies the NMR signal (FID signal or echo signal) obtained from the detection 21 signal 8; 10 is an I☆
The phase detection circuit 11 is a wave memory circuit that stores the phase-detected waveform signal from the amplifier 9, and includes a /\7'F) converter. 13 is a transmission line 1 for transmitting the signal from the wave memory circuit 11, such as an optical fiber.
There is a combicoater that inputs the tomographic image through 2 and performs the predetermined 18 processing to produce a 1q tomographic image. Necessary information is transmitted from the l-roller 20 to the convenience store 1-ta 13 by a signal line 21. The controller 20 transmits first and second gradient magnetic fields (gradient magnets $1Gz, Gx, Gy), signals required to control the amplitude of RF pulses (analog signals), and control signals (digital signals) required for transmitting R pulses and receiving NMR signals can be output. This controller 1 connector 520 has a sequence function that is a feature of the device according to the present invention, that is, a function that controls the operation timing of the RF pulse and the operation timing of the first and second gradient magnetic fields. However, the element that performs this sequence function is not limited to the controller 20, and may include other elements such as a combination, a
The present invention can also be achieved by providing this function. Figure 8 shows the child's (work = 1 nto 〇 - especially high 3!l1l)
ll! FIG. 2 is a configuration diagram showing an example of a two-in-one roller that can be controlled and easily change control sequences and analog waveforms. In the figure, 221 is a self-programming control circuit that writes data sent from the combi coater 13 to each memory via the operation console 210 or during operation; 222. 225°22
8, 231 is x given by this write control circuit,
a waveform storage memory in which waveform data of the y, z gradient signal and the modulation signal are stored, respectively; 223. 226, 22
9. 232 is the waveform memory memo',) 222. 2
25. A latch circuit for temporarily holding the waveform data output from 228 and 231, respectively, 224. 227. 23
-0. 233 is this latch@path 223. 226.
22J DΔ which changes the output from 232 to each DA
It is a conversion circuit. ×2. V2+Z2+M2 are the x, y, and z gradient position encoding outputs and modulation signal outputs, respectively, output from the DΔ transformation circuit. 234. 236. 238. Reference numeral 240 denotes a waveform storage memory, which is supplied via the write control circuit 221 to transmit/receive circuit control signals, △D conversion control signal, transmission gate control No. 13, reception signal 1 to control signal 1 phase selection signal (the phases are different from each other). This is a signal for selecting one type of pulse from among four different types of RF pulses. ) data is written. 235, 237. 23'l, 241 are latch circuits that temporarily hold data output from this waveform storage memory;
T2, S2. R2 and PS are latch circuits 235. 237
AI) Conversion control signal output from 239 and 241, transmitter 1 to control signal output, receiver 3 gate control signal output. This is a phase selection signal. 2431J @ readout control circuit that reads out the contents of each waveform storage memory to the latch circuit; 242 sets the value of the write 7' readout start address from the operation console or =1 computer (hereinafter referred to as computer); At the same time, a memory address register 244 sequentially adds 1ifJ −t−1 given from the write/read control circuit to the value of the address and outputs this as a write/read address. An output count register 245 indicates the number of output steps to be applied and notifies the readout control circuit 243 of the end of the output, and 245 is an output count register 245 which indicates the end of the output when the number of output steps (1 step) is hit. This is a 1-step-length pulse generation circuit that generates a 1-step-length pulse. The operation of the controller having such a configuration is as follows. (-a) Write operation Writes the waveform data sent from the combicoater to the designated address of the waveform storage memory designated by the combicoater. First, a store start address is stored in the memory address register 242. The data sent from the combi coater to the old writing command is sent to the write control circuit 2.
The data is written to the address specified by the memory address register in the waveform storage memory selected by 21. After that, the write control circuit 221 automatically adds 1 to the memory address register 242 to make it the memory address for the next iB entry. Data is sequentially loaded into and output from other waveform storage memories by the same operation as described above. (b) Read operation The contents of each memory are read in parallel. FIG. 9 shows an example of a time chart of read signal waveforms. The computer first sets the reading start address of the waveform storage memory into the memory address register 242. Next, output the number of read steps Callan 1-Register 244
Set to . In addition, one step length pulse generation circuit 245
Sera 1~. Next, in response to a reading start command from the computer, the waveform storage memory 222, 225, 228. 23
1. 234. 236. 238. 240 are simultaneously read out, and when all the data is available, the read control circuit 243 sends the latch circuits 223 . , 226°229
, 232. 235. 237. 239. 241
Outputs a latch pulse to latch the data. Next, if the memory address register 244 indicates completion, the read control circuit 243 transfers the data to the circuit 223. 226.
229°232, 235. 237. 239.
A clear pulse is issued at 2111 and then at 8 to complete the operation. When the output registers 1 to 1 registers 244 are not completed, 1 is subtracted from the output registers 1 to 1 registers 244, and the time length of one step is determined by the output from the pulse generator circuit 245. The process moves on to the successive step I. By repeating the same procedure, it is possible to read out a waveform such as that shown in FIG. , 227.
230. It is an analog signal obtained by D△ conversion at 233, and the modulated signal @M2 is sent to the gate circuit 30 and x
, y, and z gradient signals are respectively led to a control circuit 4 for controlling the gradient magnetic field. Since such a controller is equipped with dedicated hardware such as a waveform recorder 10 memory, it is possible to read and output a large amount of data at high speed. Furthermore, since the contents of the waveform storage memory can be rewritten as necessary, eleven unique analog/digital signal B waveforms can be output. Furthermore, by appropriately changing the bladder output start address and the number of reading blocks, it is easy to use a portion of the signal waveform (which is often actually used). The gate circuit 30 receives the RF multiplied signal from the oscillator 6;
For this, four types of signals with different phases by 90 degrees are created, one of the four types of signals is selected based on the instructions from the controller 1 and roller 20, and this is further modulated with an RF modulation signal to excite it. This is used to obtain a drive signal for the coil 5, and the detailed configuration of A is shown in FIG. In the same figure, 311 is the input R
A 90° phase shifter capable of simultaneously obtaining two signals with phase shifts of 0° and 90° with respect to the F-fold signal, 312. Reference numeral 313 is a 180° phase shifter that simultaneously outputs two signals having a phase shift of Oo and 180° with respect to the input signal. As shown, each output of 90° phase @ 311 is 1806
By adding to each of the phase shifters, the 0
Signals having phase differences of @, 1806.90', and 270" are outputted to 1q. These (, M high) high frequency switches (for example, double balanced mixer: DBM can be used) 314 to 317 In this case, the high frequency switches are individually activated by the output of the decoder driver 320, and decode the phase selection signal PS given from the controller 20. The four outputs (X,
'l', -X, -'l'> Any one of IJ becomes active. As a result, only the corresponding switch becomes conductive (the switches of the other 31 prisoners are non-conductive)
. Therefore, only one signal is input to the coupler 321. The output of the coupler 321 passes through the amplifier 322 and then is input to the modulator 323, where the R "modulation signal @ (pulse signal, which is a pulse signal, and the child pulse width and peak (The rotation of the magnetization M is determined by the value. Originally 0°, 90°, 180°, 270° ((I
The RF multiplied signal 1q1 exhibiting K can be selectively selected from among these signals, and furthermore, the child signal can be modulated with a desired waveform at an appropriate timing (the advantage is that it can be easily achieved by ~). The operation of this Shomei device configured in this way will be explained next with reference to time charts 1 to 1 for the two-dimensional PR method in Fig. 11. 1) Control circuit 2 for static magnetic field A current is applied to coil 1, and a static magnetic field H9 is applied to the test object (the test object is placed inside the cylinder of each coil).
In the state in which Z slope nj! is applied from the controller 20 through the control circuit 4, A current is applied to the magnetic coil 31, and the first gradient uk field (Z
As shown in Fig. 11 (a), apply the first 90°
Excite n selectively. Let this point be to. 2> Apply Gz- following the above-mentioned Gz+ application. This is for matching the phases of NMR resonance signals from different parts of the object, and this J:Una technique is a known technique. The end point of this Gz-application is 12
shall be. 3) After that, the times Gx and Gy of t+r++ are respectively tI
It is applied with the magnitude of Qx+ and Qy+. 4) After 51 hours of applying the first 90°X pulse, a 180° RF multiplied signal modulated into a rectangular wave with a phase difference of 90' is selected and output by the circuit 30 to the gate 1.
The pulse causes l1iI+ in the subject and reverses the magnetization. This 18f')' is the 11th pulse after t.
Homogenity spoil pulses are applied at Gx, Gy, and Gz as shown in (b) to (d) in the figure. This homogeneity spoil pulse can suppress the noise generated due to the inaccuracy of the 180° pulse. Here, the subscripts x and y of 90° and 180° indicate the phase of the RF pulse, and X and y have phases different by 90°. 5) Next, when Gx and Gy are added to Q'x+-+Q'y+ as shown in (C) and (D) of the same figure, J as shown in (Bo) of the same figure is obtained.
, a spin echo signal is generated. Then, gx+ Xjm+ =Q' x+ Xi' yn+Qy
+ X[m+ =Q' y+ 6) Next, change the magnitudes of Gx and Gy to QX2 + Gy2 and repeat the same operations as 2) to 5) above. At this time, it is necessary to satisfy the following formula. gxp Xjynp -C2' xp xt:' ml
l+C1yp Xtmp =Q' yp xt' mP
Here, the subscript p is 1, 2, . , , , rl, where n is the number of 180° pulses between the first 90° pulse and the second 90° pulse (details described below). 7〉Predetermined number of times n (, after passing the J180° pulse,
At the timing when the Jco signal is at its maximum (at the end of t' mn), the magnetic 1hi is selectively applied with a 90° pulse and Gz.
When tl is an odd number, it is pulled in the Z-axis angular direction (downward). When 11 is an even number, it is pulled in the Z-axis direction (upward).
(shown in Figure 19 for the case where is an even number). 8) Only when n is an odd number, follow with a 180°-X pulse to turn the total magnetization upwards. 9) These are followed by homogeneity spoil pulses with Gx, Gy, and G2. By applying this pulse, the signal can be set to 1 without correlation with the next sequence. 10) It is Td time (the same sequence is repeated by the master). With J5 in such a ceiling, each time parameter Tsp
, T' SP+ T, i, and n are appropriately selected depending on the respective usage conditions. In the sequence described above, the movement of the magnetic f and
Pulse application time limit M rotates the θ° plane, and when applying a 180° pulse, it becomes G2-0, so 180° rotation J
12 and 13, respectively, for each part outside the slice plane (the part that is not affected by the application of a 90° pulse and the direction of magnetization M is reversed by the 180° pulse). In Figure 12, when n is an odd number, the total magnetization is turned downward with the last 90° tilt pulse, and then turned to the soil with 180°8. The slice boundary ((c) in the same figure) is θ° with a 90° pulse.
(0°<θ<906) L does not rotate, but the second 90°
Immediately before the pulse, it is angled θ° from the Z-axis angular direction, so it is directed upward at 180'-X. In FIG. 13, when n is an even number, there is C, and in the second 90° pulse direct flight, it is 0 toward the square of the two axes, so it is only necessary to push 1 toward G at 90°-X. At such a theetance, tJ7 N M Rf
The ffi month ((e) in FIG. 11) is limped by the wave memory, and the data is reconstructed into a two-dimensional image by the combicoater 13, and a cross-sectional image of the subject is displayed on the display 14. . Note that the present invention is not limited to the embodiments described above, and may be implemented in various ways or methods as described below. 1) As shown in FIG. 14, n = 2. gx, ＞
>g'XI・Qy + )C1' y + , QX2
(Q'
The s2 signal is not used because it is affected by RF pulses and noise from Gx, Gy, G2, etc., and is not used during the period T'.
51 and TSi2 signals are used. According to this method, the above article ('l, Jζsu, tmI
(t'ml and tm2 >>j'm2, so tm
l and tm2 have the advantage that the signal can be collected with a good s7・'N. Furthermore, in Fig. 14, Ts + = T' S I , TS2
= T's2 How 1m, I + 1°m l
, '−□2゜t'IT+2 and NO' x+ =
When the gradient magnetic field is gX2 I Q' y+ =gy2, the Echo-4i receives RF pulses, Gx, Gy,
Since c7 is not affected by noise such as Gz, S,
/N good signal can be obtained. 2) When applied to the echo planar method, the first
The system is as shown in Figure 6 (the diagram when nh<odd number). A typical 1-cope loop is a reversal of the magnetic field (e.g. G
However, in the present invention, it is performed by applying a 180° pulse. Finally, the magnetization M is forcibly turned upward (when n is an odd number, 9o° pulse and 180° pulse, 1
H′ where 1 is an even number? i is performed with a 90' pulse). 3) When applied to the cereutive excitation line method, Figure 17 (diagram when n is an odd number)
The method shown in . 4) When the inversicon recovery method is applied, that is, as shown in Figure 18, a 180° pulse is added before a series of pulse sequences (the part surrounded by a chain line), and Here, we introduce small modality spoil pulses Gx , Gy , G
It is commonly added to z. However, this homogeneity
The spoil pulse mark 110 is not necessarily required C.
If there is no adverse effect in the lateral direction or can be ignored, it can be omitted. This inversion recovery method can be applied to all of the above-mentioned methods, and is suitable for iQing an image in which 1-3 (longitudinal relaxation time) is emphasized. 5) In each of the methods listed above, the non-selective 180° pulse is a plurality of pulses. For example, 180
Instead of 90°X, 180°y, and 90°X pulse trains are used. This makes it possible to cancel out inaccuracies in pulse intensity and the like. 6) In the inter-image calculation, it is also possible to obtain an image using T, an image, a T2f image, a spin density image, or any combination thereof. For example, when n-1, the signal strength V jumps at sequence 17 in Fig. 11, and M uses the following relationship to obtain T S + + T ' S
Calculations are made from multiple images obtained by changing I + T and d. 7) Use the multi-slice method). However, the waiting time of Td is utilized to excite the surface of the state and obtain its NIVIR signal. 8) The phase relationship of the pulses in each of the above methods is qo''
”x-+8O'y-IF3cr'y--Cto'-x(
n:'191. QO"x -1&)y -1&)'y
-・-'(C;"x I&)'-X (n: odd number), not limited to QO'x (1&)lx 180"x') --this 1
0o-x (n・2k i1id)C10还-(1≧Goto)2x 18O,')"--/to2×-qd5-×1
It is also possible to apply it to the two-dimensional Fourier method.
The child's behavior is as follows. In addition, Fig. 20 is a diagram showing the waveforms of each part that is applied to the ceiling, and Fig. 12 is a diagram showing the movement of magnetization ivl during the sequence. (1) Shown in (a) and (b) of the same figure. selectively excite with 90'8 pulse and 111J G z+. (2) Apply magnetic field G2- as shown in the figure to align the phase of the excited spins in 7 coordinate directions. At the same time, 9x and changes the phase in the X direction. In this case, the output of gX determines the change in the phase in the X direction (this is phase encoded: pbase enc
O (referred to as 1e). (3) After the same figure (2, as shown in λ, apply the gradient magnetic field Qy for protraction during the Ts+ period and observe the FID signal @ (Figure (E)). (4) ) Apply a 180°y pulse. In this case, before and after each pulse, f, G x as shown in (b) to (d)
. Add homogeneity spoil pulse with Gy and Gz. (5) Observe Subineni 1-Shinyumi while applying Oy again. (6) Repeat (4) and (5) above. (7) After applying the 180° pulse for a predetermined number of times n (odd number of times), apply the same g and Gz- as applied to the TSI section, and selectively spin at the 90° pulse Gz+. is directed in the Z-axis angular direction (downward), and then a 180° pulse is applied to direct the total magnetization to the Z-axis stopping direction. (8) Apply spoil pulses at tllGx, Gy, and G2, and (9) T Wait for 6 hours and repeat the same sequence. However, the intensity of Gx or application time 1 is changed sequentially. After completing the entire 00 sequence, J, performs reconstruction calculations while performing the sequence. For example, collect the first echo signal, Haruno, and
A two-dimensional 71-trix whose vertical dimension is Gx is assumed, and the matrix is subjected to two-dimensional Fourier transformation. To perform calculations on a computer, it is preferable to use Fast Fourier Transform +@ (FFT). The result is a two-dimensional tomographic image of the subject, which is read out as appropriate and displayed on a display. Note that when the number of repeated surfaces n is an even number, the operation follows the sequence shown in No. 21, and the movement of the magnetization M is as shown in FIG. 13. The operation order is the same as the operation when the number of repetitions n is an odd number. However, only item (7) above is as follows. (7) After applying 8 180° pulses a predetermined number of times (even times), Gx with the opposite polarity to that applied in the T s + section
and Gz are applied, and the 90° pulse and Gz direct all spins toward the Z-axis stopping direction. In this way, images can be obtained at high speed. Note that images can be similarly obtained for each echo signal such as the second echo signal and the third echo signal. These images can be averaged to 4-Ge, and can also be used for image calculations to be described later. Note that the averaging may be performed on time-series data before it is converted into an image. Note that the two-dimensional Fourier method is not limited to the first embodiment (various modifications are possible). For example, as shown in Fig. 26, " = 2 + Q x, +
)Q'x + , QX2 <Q'X2. In this case, the signals in the periods Ts+ and T'S2 are not used because they are affected by RF pulses and noise from Gx, Gy+Gz, etc., and the signals in the periods T's+ and TS2 are used. According to this method, the article f′4Jζs,j−m,<
: t'ml and tIT12) 1-'m2, the time between t'm+ and tm2 is sufficiently short (, good S/
It has the advantage of being able to collect signals from anywhere. Furthermore, in FIG. 26, Ts' + = T' s + , T
tm I + 17m so that s 2-T-'52
+ + tm2 +t'm2, Q' XI =
When we set the gradient magnetic field to QX21 Q' y+ -Gx2, the echo signal is completely unaffected by RF pulses, Gx, Gy, and Gz forces, so it is possible to obtain a signal with a good S/N ratio. can. ■Phase - [N code ■ is not constant, and as shown in Figure 22 (number of repetitions n/]'ia number) and Figure 23 (number of repetitions is an even number), each time a 180° pulse is applied. They may be changed sequentially. In this case, F
The phase of the ID is T and the code m is A1, the phase of the first echo is B (Q'x away from △), the phase of the second echo is As shown in the following, the phase T chord is monotonically decreasing or increasing. At this time, Gx changes polarity between odd and even numbers. Then, the GX gradient magnetic field immediately before the second 90° pulse (Jx is applied to prevent the last phase Tn code from returning. ■ Using the method described in F, Q' x + Q″x + - Knee may be applied at the same time as the spoil pulse immediately after each 180° pulse, that is, by adding both pulses. ■The above 9'x9g'' ■As shown in Figure 24 (when the number of repetitions is an odd number) and Figure 25 (when the number of repetitions is an even number), by inverting the projection gradient position 1 'li3 G y, the echo signal can be obtained. In this case, the FID signal immediately after the 90° pulse can be used. In the figure, the phase T code suspension is gx Xjm' (k=1
〜１１〉 is decided as J:, and the first and second -L Coco-8
The echo signals have the same magnitude, the third and fourth echo signals have the same magnitude, and so on. (2) It is also possible to prefix each sequence with 1806 pulses for inversion recovery as shown in FIG. 18 to obtain a T1-weighted image. ■Comboji 1 to 18 of each of n 180'' pulses
06 pulse may be used. ■ Between images (Ih calculation: T1 edge IT-2 image, spin! 7 degrees doctor, or a combination of these iM eyes). In this case, the signal strength V is (10 anti-1g0L) in FID. 30)--'jo'l) go
z (breasts, visit'lD'z-(+8o:, -1go',
-)'-------n(Ll-qO varnish JgO1(vlk
+i; It can also be spelled as sail etc. (Effects of the Invention) According to the present invention, after observing a large number of spin echoes, the magnetization is forcibly returned to a thermal equilibrium state, and the total magnetization is direction), it is possible to move on to the next operation G in a short time Ta, and the overall scan time can be shortened. In addition, in order to improve image quality, it is possible to collect a large number of similar data and average it as time-series data, or convert it into images and then average them, in a much shorter time than with conventional methods. play. Furthermore, the slice shape is better than the DEFT methods A and B, as shown by the solid line C in FIGS. 27 and 28. However, Fig. 27 shows that the magnetization outside the slice plane is larger than A, and Fig. 28 shows that the obtained signal (1
It can be seen that i is close to another J:i rectangle. Furthermore, the next J: Una effect is also exhibited. ■Before and after the 180° pulse of non-selective excitation, the magnetization "above the xy plane" and "below the xy plane" alternate, so the influence of T+ relaxation is small. For example, in FIG. 13, 11-
At t2, the magnetization vector is pulled upward by 1', causing the cone in the same figure (c) to become narrower, and at t3 to t4, the magnetization is pulled to '2' by T +, resulting in the direction shown in the same figure (c). The cones cancel each other out in the direction of expansion. (g) Magnetization by small t-genity spoil pulse
The μ component disappears, and the correlation between the squeegees disappears.
The magnet moves correctly and there is little noise. ■ TSL and T'S2 together at n -2 (T's+ 10T
s2), it is possible to eliminate the influence of noise on the applied signal, and at the same time, by shortening TSL and T'S2, the N
M R4:A ”?+ is calculated by ft1J. ■ By using multiple pulses for the 180° pulse, 31 differences in intensity etc. are canceled, so the magnetization shows the correct rotation. ■ The image-to-image calculation meets the purpose. Images can be easily scaled up by 1q. ■Multi-slices can further increase the speed of viewing by 111 x 1. 4. Brief explanation of the drawings Figure 1 1; L 71 (explains the spin of elementary atoms) Figure 2 is a schematic diagram of the magnetic moment of a hydrogen atom, Figure 3 is a schematic diagram of the magnetic moment of a hydrogen atom.
The figure explains the state in which the nuclei of hydrogen atoms are aligned in the direction of the magnetic field.J
FIG. 4 is a diagram showing an example of the test pulse waveform by N M tt, FIG. 5 is a diagram showing magnetization M in a rotating coordinate system, and FIG. Fig. 7 is a structural diagram showing an example of a magnetic field coil, Fig. 8 is a detailed configuration diagram of the controller 1 to roller 20, Fig. 9 is a diagram for explaining the operation of the controller 20, and Fig. 10 is a diagram showing the configuration of the controller 20. 1 to 30, FIG. 11 is an operation waveform diagram for explaining the sealing according to the present invention, and FIG. 12 415 and FIG.
14-26 are operational waveform diagrams for explaining the sequence of the operation shown in FIG. The figure is the first
Figure 28 shows the binding obtained by computer simulation of the distribution of magnetization Mz in a state where the dynamic equilibrium state is reached by carrying out the sealing shown in Figure 1 in a similar manner.
The first 90° pulse and 7 gradient magnets to the state of 2'! J.G.
FIG. 3 is a diagram showing the NMR signal intensity after selective excitation by applying () z. DESCRIPTION OF SYMBOLS 1... Static magnetic field coil, 2... Control circuit for static magnetic field coil, 3... Gradient magnetic field coil, 4... Control circuit for gradient magnetic field coil, 5... Excitation coil, 6 ...R
F oscillator, 7... power amplifier, 8... detection coil, 9... amplifier, 10... phase detection circuit, 11...
・Wave memory circuit, 13... Combi coater, 14
. . . Display unit, 20 . . .] nt O-ra, 30 .
...Y gradient magnetic field coil. Figure 26 Figure 27 Figure 28 Mark Δ

Claims (1)

[Scope of Claims] (1) Nuclear magnetism in the nuclei of atoms constituting the tissue of the subject;
a step of applying a first 90° pulse that causes a gas resonance;
A step of repeatedly applying an 80° pulse, a step of applying a gradient magnetic field in connection with the application of the 180° pulse, and a step of repeating the 180° pulse, and then 90° when n is an odd number. A process of applying a pulse and a subsequent pulse (180° pulse, and only a 90' pulse when n is an even number), and a process of moving to the next process after a predetermined waiting time after the above process. , measuring a nuclear magnetic resonance signal generated under the applied magnetic field, and reconstructing an image related to the tissue of the subject based on the obtained nuclear magnetic resonance signal, and applying the gradient magnetic field. In the step of applying the repeated 180° pulse, the nuclear magnetic resonance inspection method is configured such that the following relationship holds true before and after each application of the repeated 180° pulses: Qxp X F, mp = G' xp X j' rn
pQyp X jmp =L; J' yp X t'
mp Here, the subscripts p −1,,2,,,,ntmp are the gradient magnetic field application time that is 1 times the Tsn time before applying the 180° pulse. t'mp is the gradient magnetic field application time at T'sn time after applying the 180° pulse. (jxp is the magnitude of the X-axis gradient magnetic field at tmp. Qyp is the magnitude of the y-axis gradient magnetic field applied to tmp. Q' xp is the magnitude of the X-axis gradient magnetic field at t' mp. CJ'yp is t (2) In the step of applying the gradient magnetic field, a part of the gradient magnetic field applied before and after the repeated 180° pulse is strengthened, and the application time is shorter than the magnetic field application time (
With this, apply the above! In the step of measuring the nuclear magnetic resonance signal generated under the i field and reconstructing an image related to the tissue of the subject based on the iq field, the gradient magnetic field application period is longer than the short gradient magnetic field application period. Nuclear magnetic resonance according to claim 1, characterized in that only the echo signals are observed during the application of the gradient ll field if1. The step of measuring the signal and reconstructing an image related to the tissue of the subject based on the obtained signal is one of the PR method, spin warp method, echo planar method, or selective one-dimensional line method. A nuclear resonance inspection method according to claim 11n, characterized in that the step of applying the first 90° pulse comprises applying the 90° pulse. The nuclear magnetic resonance testing method according to claim 1, characterized in that the method includes applying an 1806 pulse for inversion recovery before the 1806 pulse. Claim 1, characterized in that three pulses at 4' are used by administering two 90° pulses whose phases are 90° different from each other but in the same phase.
d1° (6) In the step of applying the gradient magnetic field, apply a homogeneity spoil pulse with a gradient f41m immediately before and immediately after applying the repeated 180° pulse. Inspection by nuclear magnetic resonance according to claim 1, characterized in that the test 7! li method. (7) In the step of applying a 90° pulse or a 180° pulse after repeatedly applying the 180° pulse,
2. The nuclear magnetic resonance examination method according to claim 1, wherein a homogeneity spoil pulse is applied using a gradient magnetic field after the application of the 90' pulse or the 180° pulse. (8) In the step of applying the gradient magnetic field, the application time is changed every repetition, and in the step d5 of reconstructing the image, an inter-image calculation is performed from a plurality of images obtained by changing the time parameter to T1. image. At least one image of T21'lJ, a spin density image, and a combination image thereof is made to look like (Claim 11)] The nuclear magnetic resonance inspection method according to claim 1. (9) After the step of repeating the 180° pulse, when n is an odd number, add 906 pulses and the following 180° pulse, and when n is (!Jt pl!), add 906 pulses and the following 180° pulse
In the process of applying only a 0° pulse, multi-slicing is performed by selectively exciting another slice plane during the waiting time before moving on to the next sequence. The testing method using nuclear magnetic resonance according to scope 1. (10) means for applying a high-frequency pulse for causing nuclear magnetic resonance in the nuclei of atoms constituting the tissue of the subject; means for applying a gradient magnetic field for applying a gradient I & field to the nuclei; In a nuclear magnetic resonance inspection apparatus equipped with means for measuring a generated nuclear magnetic resonance signal, the means for applying a high-frequency pulse applies a 90° pulse and, after application of the 90° pulse, applies a 180° pulse n times. The application is repeated, and then 906 pulses are applied when n is an even number, and 906 pulses are applied once when n is an odd number.
80° pulses are applied, and the gradient magnetic field applying means is configured so that the following relationship is established before and after each application of the 18° pulses. Inspection device using resonance. Qxp Xjmp =Q' xp X't' mpGy
'pXjmp=G'ypXj' ml) Here, subscript p=1. 2,80. , ntmp is J5(
Apply a gradient magnetic field 141tfl. t'mp is the gradient magnetic field application time at the r'sn time after applying the 180° pulse. gxp is the magnitude of the X-axis gradient magnetic field at the tip. (Jyp is the magnitude of the y-axis gradient magnetic field at tip. 0"xp is the size of the x-axis gradient magnetic field at t'mp. 'l'yp is the magnitude of the y-axis gradient ll field at l:'mp.