JPH04208134A - Magnetic resonance image device - Google Patents

Abstract

PURPOSE:To reduce greatly the influence of noise upon an object to be inspected and the peripheries of the image device according to the invention by measuring the noise generated from gradient coils, presuming the lead pulse sequence which minimizes the power of the noise, and thereupon driving the gradient coils. CONSTITUTION:An image device comprises a measuring means 9 to measure the characteristic of noise generated by gradient coils 3 for lead, a presuming means 11 which presumes the optimum lead pulse sequence to minimize the power of the noise from the characteristic of the noise to be measured, and a means 4 to drive the gradient coils 3 in accordance with the optimum lead pulse sequence presumed. That is, the coils 3 are driven by a power supply 4 for coils, while this power supply 4, a signal transmission part 7, signal reception part 8, and data collection part 10 are controlled by a system controller 9. A pulse sequence as countermeasure to the noise is provided to suit the intended use. This greatly reduces the influence of the noise upon the object to be inspected 5 and the peripheries of the device concerned.

Description

Detailed Description of the Invention [Invention No. 1] (Industrial Application Field) The present invention relates to a magnetic resonance imaging apparatus, and in particular to a method for processing magnetic resonance signal data such as spin density image data within a subject at high speed. This article relates to ultra-high-speed magnetic resonance imaging equipment that collects images.

(Prior Art) As is well known, magnetic resonance imaging (MRI)
The chemical and This is a method of visualizing physical microscopic information. This magnetic resonance imaging method takes a very long time to collect data compared to other medical image diagnostic devices such as ultrasonic diagnostic devices and X-ray CT. Therefore, there are problems in that artifacts occur due to movements such as respiration of the subject, and it is difficult to image a moving heart or vascular system. Furthermore, since the imaging time becomes longer, the pain caused to the subject is also greater.

Therefore, as methods for collecting data at high speed in magnetic resonance imaging, the echo planar method by Mansfield and the ultrafast Fourier method by Hutchison et al. have been proposed. FIG. 12 shows a pulse sequence for collecting image data using the ultrafast Fourier method (also referred to as the multiple echo Fourier method). High frequency magnetic field R
As F, a 90° high-frequency pulse for selective excitation is applied, and at the same time, a gradient magnetic field Gs for slicing is applied to selectively excite magnetization in the slice plane, and then an 80° high-frequency pulse is applied, and then a slicing gradient is applied. A gradient magnetic field G for reading is applied in a direction parallel to the in-plane by switching between positive and negative at high speed, and at the same time a gradient magnetic field Ge for phase encoding is applied in a pulsed manner in a direction perpendicular to Gs and Gr every time Gr is switched. Here, the slicing gradient magnetic field Gs is a gradient magnetic field for slicing a desired cross section in the subject 5, and the phase encoding gradient magnetic field Ge is a gradient magnetic field for converting position information in the slice plane into phase information. The read gradient magnetic field G' is a gradient magnetic field for reading out magnetic resonance signals.

When this pulse sequence is applied, there is a time when the phase of magnetization in the slice plane is aligned every time the read gradient magnetic field is switched, so a large number of echo signal trains are observed. These signal trains can be collected as image data within the time for relaxation due to the relaxation phenomenon of magnetization in the slice plane or transverse magnetization, making ultra-high-speed imaging possible.

However, these methods have problems as described below. In order to realize ultra-high-speed imaging, it is necessary to supply a large current to the read gradient coil for generating the read gradient magnetic field Gr, and to switch this current at high speed. At this time, the current flowing through the gradient coil receives a force larger than the static magnetic field, and a very large noise is generated depending on the read pulse and the noise response function of the gradient coil. This phenomenon becomes stronger as the static magnetic field strength and gradient coil current increase, and since it is within the audible range in ultrahigh-speed imaging methods, it places a large acoustic burden on the subject. Conventional techniques for reducing this noise have been proposed, such as changing the weight, size, support method, structure, etc. of the gradient coil, and incorporating sound damping shrines, acoustic shielding shrines, etc. into the gradient coil. At present, neither method can be said to have a dramatic noise reduction effect compared to the ultra-high-speed imaging method.

On the other hand, a technique has been proposed in which a signal with the same amplitude and opposite phase as the noise perceived by the subject's hearing is generated to cancel the noise. This technology is an effective method in terms of reducing the acoustic burden on the subject itself, but it also has the disadvantage that the noise generated from the lead gradient coil, such as in ultra-high-speed imaging, has a significant impact on the surroundings of the device. Well,
It's not a solution at all.

In addition, in conventional magnetic resonance imaging, a technique has been proposed to reduce noise by changing the spectral properties of gradient pulses by controlling the pulse's sooting characteristics. At present, very little consideration has been made.

(Problems to be Solved by the Invention) As mentioned above, in order to realize ultra-high-speed imaging, when a large current is supplied to the read gradient coil and this is switched at high speed, the current flowing through the gradient coil is lower than the static magnetic field. A large force is experienced and a very loud noise is generated depending on the read pulse and the noise response function of the lead gradient coil. Conventional techniques for reducing this noise include changing the gradient coil itself and its surrounding structure, but none of them can be expected to have a dramatic noise reduction effect on ultra-high-speed imaging methods, and they are similar to the noise that the subject perceives with his/her hearing. The method of canceling noise by artificially generating signals with opposite phases in amplitude does not reduce the effect of noise on the surroundings of the device, and it is also necessary to control the spectral properties of the gradient pulse to control the switching characteristics of the pulse. The method of reducing noise by changing the noise cannot be said to be an effective method even for ultra-high-speed imaging methods.

SUMMARY OF THE INVENTION An object of the present invention is to provide a magnetic resonance imaging apparatus having noise reduction means that is effective even for the noise generated during ultra-high-speed imaging as described above.

[Structure of the Invention] (Means for Solving the Problems) The present invention applies a high-frequency magnetic field and gradient magnetic fields for slicing, phase encoding, and reading to a subject placed in a uniform static magnetic field. In a magnetic resonance imaging apparatus that detects and visualizes the magnetic field signal from the subject by applying magnetic field according to the pulse sequence of : an old control stage to be measured; an estimation means for estimating an optimal lead pulse sequence that minimizes the power of the noise from the characteristics of the noise controlled by this means; By including a means for driving a lead gradient coil, the acoustic influence of noise on the subject and the surroundings of the apparatus is reduced.

The means for estimating the optimum pulse sequence includes a means for variably setting the read pulse sequence within a predetermined range while keeping the imaging specifications (spatial resolution, etc.) constant, and a means for estimating the spectrum of the read pulse sequence set by this means. means for estimating the power of the noise generated from the lead gradient coil from the spectrum estimated by this means and the characteristics of the noise measured by the measuring means; and a minimum power of the noise estimated by this means. and means for estimating a read pulse sequence.

In the case of a read pulse waveform or a rectangular wave, the means for variably setting the read pulse sequence sets the gradient strength of the read gradient magnetic field and the sampling pitch in the read direction to be constant,
Means for changing the read pulse switching interval and data collection time within a predetermined range, or changing the gradient strength of the read gradient magnetic field, the sampling pitch in the read direction, the read pulse switching interval and the data collection time within a predetermined range This is achieved by means of

In addition, when the read pulse waveform is a trapezoidal wave, the means for variably setting the read pulse sequence is such that the gradient strength of the read gradient magnetic field, the sampling pitch in the read direction, and the switching time of the read pulse are kept constant, and the switching interval and read pulse sequence are set constant. Means for changing the data collection time within a predetermined range, or by keeping the gradient strength of the read gradient magnetic field and the sampling pitch in the read direction constant, the switching time, switching interval, and data collection time of the Coulomb pulse within a predetermined range. means to change it,
Alternatively, this can be achieved by changing the gradient strength of the read gradient magnetic field, the sampling pitch in the read direction, the read pulse switching time, the switching interval, and the data collection time within a predetermined range.

Furthermore, in the case of a read pulse waveform or a sine wave, the means for variably setting the read pulse sequence includes, while keeping the gradient strength of the read gradient magnetic field constant, the sampling pitch at unequal intervals in the read direction, the switching interval of the read pulse, and the data acquisition time. within a predetermined range, or the gradient strength of the read gradient magnetic field, unequal sampling pitch in the read direction, read pulse switching interval, and data collection time within a predetermined range. achieved by means.

On the other hand, estimating and implementing the optimal lead pulse sequence can be achieved, for example, by estimating the sequence that minimizes noise for each cross section when imaging an arbitrary cross section, and then individually implementing those sequences. You can
Alternatively, when imaging an arbitrary cross section, estimate a sequence that minimizes noise for each cross section, and then construct an optimal sequence group that can be applied to all cross sections.1. (may also be carried out).

The minimum value of the noise power is estimated, for example, by estimating such that the square integral value of the product of the response function spectrum of the noise, which is a noise characteristic, and the spectrum of the read pulse becomes the minimum value.

(Function) In the present invention, by estimating and implementing an optimal pulse sequence that minimizes the power of noise generated when applying a gradient magnetic field for reading, significant changes can be made to the gradient coil itself and its surrounding structure. The influence of noise on the subject and the surroundings of the device is significantly reduced by = 11-.

(Embodiment) FIG. 1 is a block diagram showing the configuration of a magnetic resonance diagnostic apparatus according to an embodiment of the present invention. In the figure, a static magnetic field magnet 1 is driven by an excitation power source 2, and a gradient coil group 3 is driven by a gradient coil power source 4. As a result, a uniform static magnetic field and a gradient magnetic field having a linear gradient magnetic field distribution in three directions perpendicular to each other in the same direction as the static magnetic field, namely, the slice direction, the phase encode direction, and the read direction are applied to the subject 5. Ru.

The gradient coil group 3 includes gradient coils for slicing, phase encoding, and reading. The probe 6 is supplied with a high frequency signal from the transmitter 7, applies a high frequency magnetic field to the subject 5, and receives a magnetic resonance signal from within the subject 5. The probe 6 may be provided for both transmission and reception, or for both transmission and reception. The magnetic resonance signal received by the probe 6 is subjected to, for example, phase detection in the receiving section 8 , and then transferred to the data collecting section 10 , where it is A/D converted and then sent to the data processing section 11 .

Excitation power source 2, gradient coil power source 4, and transmitting section 7
, the receiving section 8 and the data collecting section 10 are all controlled by a system controller 9. In the system controller 9, noise countermeasure pulse sequences are prepared according to various uses. The system controller 9 and data processing section 10 are controlled by a console 12. The data processing unit] 1 performs image reconstruction processing such as Fourier transform of the magnetic resonance signal input from the data acquisition unit 10, thereby calculating the density distribution of desired atomic nuclei within the subject 5, and converting it into image data. Desired. This image data is input to the image display 13, and images such as density distribution are displayed.

FIG. 2 shows a flowchart for determining a read pulse sequence that minimizes the acoustic noise caused by the read gradient magnetic field of ultrahigh-speed MRI in this embodiment. However, this flowchart only gives the noise response function spectrum, and is a determination procedure when optimizing the read pulse sequence by changing it. Each step will be explained in detail below.

In FIG. 2, first in step 21, the characteristics of the noise generated from the lead gradient coil, such as the noise response function spectrum Δr(r), are estimated. However, this spectrum Ar(1') is determined by A-characteristic measurement, which is a characteristic of human hearing. To estimate the noise response function spectrum, an acoustic sensor such as a microphone is installed at a predetermined position in the inner row of the lead gradient coil, and the frequency of the input signal of the gradient coil is swept to calculate the noise amplitude for each frequency by 4.
Possible methods include measuring the input signal of the gradient coil and using an impulse, and Fourier transforming the response (noise).

Figure 3 shows a typical noise response function spectrum Ar(r)
It shows. Normally, the fundamental frequency of the read pulse in ultra-high-speed MRI is approximately 1. OOHz to several kiloliters (z), and the wavelength of the sound wave in such a frequency band is on the order of the same as the size of the gradient coil, so the noise response function spectrum Ar(r) has multiple peaks ( Therefore, by appropriately shifting the fundamental frequency of the read pulse sequence from the above-mentioned peak, the noise level can be significantly reduced.

In step 22, various parameters (read pulse switching interval Tr,
Set the reading gradient magnetic field strength (Gr) to an appropriate value. Step 26) sets the values of these parameters under predetermined constraint conditions, and searches for an optimal dopa sequence. That is, in step 23, the human particle vector (spectrum of the set pulse sequence) Φr(f) is calculated. At this time, the spectrum of the noise output with respect to the input of the lead gradient coil is Ar(
f') ・Φr(r). Next step 24
Then, as the noise power Pr, the noise output power spectrum l Ar (f) shown in the following formula % formula %
- Calculate the integral value of Φr(f') l 2 by it.

Pr-f l Ar(f) ・Φr(f) l df
-(1) Step 25 is a step of determining whether the noise power Pr is the minimum value or not, and searching for the minimum value Pr, min. Performed iteratively on parameter values. Finally, in step 27, Pr is set to the minimum (P
An optimal read pulse sequence is determined to be (r, m1n). The above optimal sequence can be determined using software using a computer or the like, but it may also be determined while actually dynamically changing the read pulse sequence so that the noise power is minimized.

Next, we will discuss the imaging specifications of ultra-high-speed MRI (spatial resolution,
A detailed description will be given of a method for variable setting parameters of a curved pulse sequence that minimizes the noise power Pr while keeping the imaging area (imaging area, etc.) constant. Three types of read pulse waveforms will be described: rectangular, trapezoidal, and sine wave.

In addition, since the imaging specifications are kept constant, the following parameters are constant (unchanged) in all examples.
shall be.

Xr/read direction imaging range Xe: phase encode direction imaging range Δxr, read direction spatial resolution Δxe: phase encode direction spatial resolution Number of rings) Δte: Phase encode pulse width (ultrafast Fourier method) First, read pulse waveform or rectangular, variable setting method of sequence parameters is read gradient magnetic field (Gr) and read → sampling time (Δtr). - A case will be described in which a method is used in which the read pulse switching interval (Tr) is varied while keeping the read pulse switching interval (Tr) constant (see FIG. 4). As mentioned above, since Ne (the number of pixels in the encoding direction) is constant, changing Tr corresponds to changing it in proportion to the data acquisition time (DAT) or T'.

In other words, data collection time 1) There is always the following relationship between AT and Tr.

DAT=Ne-Tr-(2
) FIGS. 4(A) and 4(B) respectively show the read pulse sequence (3 types of Tr) and the corresponding spectrum (input spectrum I/L) Φr(f).

Here, it is assumed that the conditions to be satisfied by the data collection time DAT can be written as follows.

Nc (NrΔtr+Δte) ≦DAT ≦DAT
max...(3) Here, DATmaX on the right side is the longest allowable data collection time with the data collection time of ultra-high-speed MRI and 1-, and this has a different value for each part of the human body. Conceivable. Specifically, in the human chest, where blood flow and movement of the heart muscle are unavoidable, DATmax = 30 to 40 m5. In addition, the movement of internal organs can be significantly suppressed by intentionally stopping breathing.l) ATmax + 50-
80m5. Furthermore, for a head that does not move and has relatively good static magnetic field uniformity and properties, it is considered that DATmax may be set as 70 to 80 m5. However, ideally, the shorter the data collection time, the better.1) If the noise level can be reduced at will without the image quality of ultra-high-speed MRI being degraded, then the significance of this is Very large. As a specific example, the above-mentioned long head 1) A T m
If there is a read pulse sequence in which the noise is significantly reduced with respect to ax, it will be possible to carry out an ultra-high-speed MRI examination of the head without generating excessive noise.

On the other hand, the left side of equation (3) corresponds to the shortest value allowed for data collection time 1) AT. Formulas (3) and (2)
Therefore, the constraint condition regarding the switching interval Tr can be written as shown above.

NrΔ Lr+ Δte≦Tr= DATmax/
Nc - (4) On the other hand, when the human waste vector Φr(f) in FIG. 4 is written as a function of the switching interval T', the following equation is obtained.

Regarding the Tr dependence of Φr(f, Tr), as the switching interval Tr becomes longer, the fundamental frequency interval becomes narrower, and furthermore, as the data acquisition time DAT becomes longer, the spectral width near each frequency becomes narrower.

Moreover, at this time, the peak value of each spectrum increases in proportion to Tr. When the switching interval Tr becomes short, Φr(r, Tr) exhibits a behavior completely opposite to that described above (see FIG. 4(B)). In this way, the optimum read pulse sequence is determined by varying the switching interval Tr within the range of equation (4) and finding the value that minimizes the noise power Pr determined by equation (1).

In FIG. 5, the read pulse waveform is rectangular and Gr is used as a variable setting method for sequence parameters.

The case where both Δtr and Tr(DAT) are changed is shown. At this time, it is assumed that the gradient magnetic field strength Gr for Li-1 is determined as a function of the switching interval Tr as follows.

Here, γ is the gyromagnetic ratio, and the other parameters are as described above. By determining the reading gradient magnetic field strength Gr as shown in equation (6), the data collection time DAT can be made as short as possible.

On the other hand, there is an upper limit to the gradient magnetic field strength for reading depending on the performance of the gradient power supply, etc., and if this is Gr and max, then Gr≦Gr, max - (7)
The lower limit of the switching interval T' is determined from equations ([i) and (7), and together with the constraints from the allowable data collection time DATmax, Tr must satisfy the following constraint conditions.

...(8) = 21- Furthermore, when the input waste vector Φr(f) in FIG. 5 is written as a function of the switching interval Tr, the following equation is obtained.

Φr (r, Tr) Regarding the Tr dependence of Φr (f, Tr), when the switching interval Tr becomes longer, the behavior is almost the same as in the case of Fig. 4, but at each fundamental frequency (rm = (2m-1) /2Tr
) (;The spectral peak value near I is almost constant.

Furthermore, when the switching interval Tr becomes short, the behavior is completely opposite to that described above. The method for determining the minimum noise power Pr is as described in section 1 above.

The advantages of such a variable setting method of sequence parameters are that the data collection time is relatively short compared to the case shown in FIG. If it exists, the reading gradient magnetic field strength GrO value itself will be smaller than Equation (6), so a synergistic effect of noise reduction can be expected.

Figure 6 shows that the read pulse waveform is trapezoidal and the sequence parameter is variable setting method.
Tr(D
This shows the case where the AT) is changed. At this time, the switching interval Tr must satisfy the following constraint conditions.

NrΔtr+Ts≦1゛r≦I) ATmax/ Ne−
(10) However, the switching time Ts is assumed to be longer than Δte (encode pulse width). When the input waste vector Φr(f) in Fig. 6 is written as a function of the switching interval Tr and the switching time Ts, we get (1) r(f, Tr, Ts) (11) Equations (II), (12) If the switching time Ts is set to zero, it can be seen that the read pulse matches the rectangular Φr(r, Tr) (Equation (5)).

Regarding the Tr dependence of Φr (r, Tr, Ts) (in the method shown in FIG. 6, the switching time Ts is constant), as the switching interval Tr becomes longer, the fundamental frequency interval becomes narrower, as in the case where the read pulse is rectangular. The spectral width near each frequency also becomes narrower. Furthermore, the peak value of each spectrum approaches that of the rectangular read pulse as the switching interval Tr increases. Furthermore, when the switching interval Tr becomes shorter, the behavior is completely opposite to that described above, and the peak value of each spectrum decreases relative to that of the rectangular read pulse. The method for determining the minimum noise power Pr is as described above.

In FIG. 7, the read pulse waveform is trapezoidal, and the read gradient magnetic field strength Gr,
This shows a case where the switching time Ts and Tr(1)AT) are varied while keeping the sampling time Δ[r constant. At this time, the following relationship exists between the switching time Ts and the switching interval Tr.

Ts=Tr−NrΔtr−(+3
) If the shortest switching time that can realize a predetermined reading gradient magnetic field strength Gr is Ts, min, then the switching interval Tr must satisfy the following constraint conditions.

Input waste vector Φr (f, Tr, Ts) in Fig. 7
is obtained by replacing NrΔ t"1's/Tr= 1----^=--(15)Tr in FIG. 6(a)(+)). Φr(".

As for the Ts dependence of Tr, Ts), when the switching time Ts becomes longer (when Tr becomes longer), the fundamental frequency interval and each spectrum width change as shown in FIG.
The peak values of each spectrum hardly change. When the switching time Ts becomes shorter (Tr becomes shorter), the behavior is opposite to that of the double series.

In Figure 8, the read pulse waveform is trapezoidal, and the sequence parameter variable setting method is +7Gr, Δtr.

This shows the case where both Ts and lr(DAT) are changed. At this time, the switching time Ts is the shortest switching time for realizing the gradient magnetic field strength Gr for Li-1, and is proportional to Gr.

Ts=Ks-Gr-(1G
) That is, the switching interval Tr is a function of the read gradient magnetic field strength Gr and 1-, and can be written as the following equation.

Tr=NrΔ tr-1-Ts The conditions that the read gradient magnetic field strength Gr must satisfy are Or≦Gr, max (Equation (7)), as in the case of FIG.
In history, there is something from the right side of equation (10) or (14) and equation (17). To express Φr(r, Tr, Ts) as a function Φr(r, Gr) of read gradient magnetic field strength Gr, the formula (
11) or by substituting Ts(Gr) of equation (I6) and Tr(Gr) of equation (17) into equation (12). Formula (7
) and (18), the read gradient magnetic field strength Gr is changed and the read pulse sequence is determined by determining Gr at which the noise power P' is minimum.

FIG. 9 shows a case where the read pulse waveform is a sine wave and the sequence parameter variable setting method is to change Tr (DAT) while keeping the amplitude Gr of the sine wave constant. Sampling in the read direction is performed at non-uniform intervals so that sampling is performed at equal intervals in Fourier space, and the sambling timing changes as the switching interval Tr changes. The sine wave read pulse waveform is Gr(
t), then Gr(t) = Gr sin (mar 1)
-(19) Tr Also, if the nonuniform zumbling time that gives the read direction coordinate kr in Fourier space is tr(kr), then the switching interval Tr becomes the minimum when satisfies the following formula, and min satisfies the following formula: must be fulfilled.

Therefore, as a constraint on the switching interval Tr, on the other hand, if the input waste vector Φr(f) in the case of FIG. 9 is expressed as a function of the switching interval Tr, it can be written as follows.

As for the dependence of Φr(f, Tr) on T, as the switching interval Tr becomes longer, the fundamental frequency interval and each spectrum width become narrower, almost the same as in FIG. 4 where the read pulse waveform is rectangular. The peak value of the spectrum is T
It increases in proportion to r. When the switching interval Tr becomes shorter, the behavior is completely opposite to that described above. The method for finding the optimal sequence that minimizes Pr is as described above.

In FIG. 10, the read pulse waveform is a sine wave, and the read gradient magnetic field strength Grs, uneven interval sampling timing, and T
The case where both r(DAT) are changed is shown. The switching interval Tr and read gradient magnetic field strength Gr are always expressed by the formula (
21), the reading gradient magnetic field strength Gr can be written as a function of the switching interval Tr as shown in the following equation.

On the other hand, the read gradient magnetic field strength Gr has an upper limit, and is expressed by the formula (7
), Gr≦Gr, max is not satisfied, so the minimum value of the switching interval Tr is finally determined from equations (19) and (7) by Tr
, min... (2G) Therefore, as in the case of Fig. 9, the constraint condition for the switching interval Tr is as follows: It can be written as follows.

2 yr (f-1/2Tr)...(27) Regarding the Tr dependence of Φr(f, Tr), as the switching interval Tr becomes longer, the fundamental frequency interval and each spectrum width become narrower, and the peak value of the spectrum becomes approximately constant.

When the switching interval Tr becomes short, the behavior is completely opposite to that described above. Note that the advantages of FIGS. 8 and 10 are that the switching interval Tr can be adjusted in the same way as the method of FIG.
If there is a minimum noise power Pr in the direction of increasing the length, the read gradient magnetic field strength Gr or the value of Gr itself can be small, so a synergistic effect of noise reduction can be expected.

Next, arbitrary slice imaging (double oblique)
When performing ultra-high-speed MRI, the noise power Pr
This section describes the procedure for determining the optimal read pulse sequence that minimizes the read pulse sequence.

Figure 11 shows the unit vector e s perpendicular to the arbitrary slice plane.
(= (sinθ cosφ, sinθ co
sθ.

COSθ)) is shown. At this time, two gradient coils, x and y, are sufficient for generating the read gradient magnetic field, and the read gradient magnetic field generated from each coil is
Letting x, r, Gy, r, Gz, r, it can be written as follows.

(Gx, r, Gy, r, Gz, r) = (Gr
sinφ, −Grcosφ, O) −
(28) On the other hand, the noise response function spectrum in the case of an arbitrary slice is considered to depend not only on f but also on φ, so if Ar(f, φ), the noise power Pr is φ
As a function of , it can be written as the following equation.

From equation (29), for each φ (for each slice plane), Pr(
The read pulse sequence that minimizes φ) may be different. In such a case, the method for setting the pulse sequence may be to use a fixed downward optimal sequence group within a predetermined range of φ, or to individually set the optimal pulse sequence for each slice plane. May be set.

In the second example, we will explain in detail how to find the optimal sequence that minimizes the noise power by changing the read pulse sequence, assuming that the response function spectrum of all noises is fixed (constant). stated. Naturally, the lead pulse sequence should be kept constant and the acoustic properties of the gradient coils (weight, size,
It is also possible to consider a method of optimizing the noise response function spectrum Ar(1') by changing the structure, etc.). In addition, as a combined method of both methods, Rietparsunkens Φr (
It is also possible to consider a method of determining a combination that minimizes the noise power by changing both f) and the noise response function spectrum Ar(r). Furthermore, it goes without saying that the method of the present invention is applicable not only to ultrahigh-speed MHI but also to general MRI, high-speed MR1, and the like.

[Effects of the Invention] According to the present invention, a total of 71111 noises generated from the readout gradient coils are estimated, a read pulse sequence that minimizes the power of the noise is estimated, and the readout gradient coils are operated according to this read pulse sequence. By driving the gradient coil, the influence of noise on the subject and the surroundings of the device can be significantly reduced without making any major changes to the gradient coil itself or the surrounding structure.

[Brief Description of the Drawings] Fig. 1 is a block diagram showing the configuration of a magnetic resonance imaging apparatus according to an embodiment of the present invention, and Fig. 2 is a block diagram showing the configuration of a magnetic resonance imaging apparatus according to an embodiment of the present invention. FIG. 3 is a diagram showing the response function spectrum of noise, and FIGS. 4 to 10 are flowcharts showing the procedure for determining the optimal lead pulse sequence for the ultrafast Fourier method in the same example. FIG. 11 is a diagram for explaining a specific example of the setting method, FIG. 11 is a diagram showing a unit vector perpendicular to the slice plane in an arbitrary direction, and FIG. 12 is a diagram showing a pulse sequence of the ultrafast Fourier method. ] ... Static magnetic field magnet 2 ... Excitation power supply 3 ... Gradient coil group 4 ... Gradient coil power supply 5 ... Subject 6 ... Probe 7 ... Transmission section 8 ... Receiving section 9 ... System controller] 0... Data acquisition section 1 Data processing section Fig. 1 Fig. 2 Fig. 3 (A) Read pulse sequence (B) Input 2S "Pr (f) (A) Read pulse sequence 1 O diagram

Claims (2)

[Claims] (1) A high-frequency magnetic field and gradient magnetic fields for slicing, phase encoding, and reading are applied to a subject placed in a uniform static magnetic field according to predetermined pulse sequences, and magnetic resonance signals from the subject are detected. In a magnetic resonance imaging apparatus that detects and images images, there is provided a measuring means for measuring the characteristics of noise generated from a lead gradient coil for generating a lead gradient magnetic field, and a measuring means for measuring characteristics of noise generated from a lead gradient coil for generating a lead gradient magnetic field; A magnetic resonance imaging system comprising: estimating means for estimating an optimal read pulse sequence that minimizes noise power; and means for driving the read gradient coil in accordance with the optimal read pulse sequence estimated by the means. Device. (2) The estimating means for estimating the optimum read pulse sequence includes means for variably setting the read pulse sequence within a predetermined range, means for estimating the spectrum of the read pulse sequence set by this means, and estimating by this means. means for estimating the power of the noise generated from the lead gradient coil from the spectrum and the characteristics of the noise measured by the measuring means, and a read pulse sequence that minimizes the power of the noise estimated by this means. Claim 1 characterized in that it has means for estimating.
The magnetic resonance imaging device described.