JP3501182B2 - Magnetic resonance imaging device capable of calculating flow velocity images - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to nuclear magnetic resonance (hereinafter, referred to as
A magnetic resonance imaging apparatus for obtaining a tomographic image of a desired portion of a subject by utilizing a phenomenon (abbreviated as “NMR”),
In particular, it relates to a magnetic resonance imaging apparatus having a blood flow drawing means using a pulse sequence called a phase contrast method and a flow velocity image reconstruction means.

[0002]

2. Description of the Related Art A magnetic resonance imaging apparatus utilizes an NMR phenomenon to measure the density distribution, relaxation time distribution, etc. of nuclear spins (hereinafter simply referred to as "spins") at a desired inspection site in a subject. An image of an arbitrary cross section of the subject is displayed as an image from the measurement data. As shown in FIG. 3, a conventional magnetic resonance imaging apparatus includes a static magnetic field generating magnet 2 that applies a static magnetic field to a subject 1, a gradient magnetic field generating system 3 that applies a gradient magnetic field to the subject 1, and a living body of the subject 1. A sequencer 4 that repeatedly applies a high-frequency pulse that causes nuclear magnetic resonance to an atomic nucleus of an atom that forms a tissue in a predetermined pulse sequence, and an atomic nucleus of an atom that forms a biological tissue of the subject 1 by the high-frequency pulse from the sequencer 4. Using a transmission system 5 for irradiating a high-frequency magnetic field to cause nuclear magnetic resonance, a reception system 6 for detecting an echo signal emitted by the above-mentioned nuclear magnetic resonance, and an echo signal detected by this reception system 6. A signal processing system 7 for performing image reconstruction calculation is provided, and echo signals emitted by nuclear magnetic resonance are repeatedly measured to obtain a tomographic image.

In this device, as shown in FIG.
~ Static magnetic field generating magnet 2 for generating a static magnetic field of about 2 Tesla
The subject 1 is placed inside. At this time, the spins in the subject 1 perform precession around the direction of the static magnetic field at a frequency determined by the strength H ₀ of the static magnetic field. This frequency is called the Larmor frequency, and the Larmor frequency ν ₀ is

[Equation 1] , And has a unique value for each type of nucleus. If the angular velocity of the Larmor precession is ω ₀ , then there is a relationship of ω ₀ = 2πν ₀

[Equation 2] Given in.

The high frequency irradiation coil 14 in the transmission system 5
Larmor frequency ν of the nucleus to be measured by a
_When a high-frequency magnetic field (electromagnetic wave) having a frequency f ₀ equal to ₀ is applied, spins are excited and transit to a high energy state.
When this high-frequency magnetic field is cut off, the spins return to the original low energy state with a time constant corresponding to each state. The electromagnetic wave emitted at this time is received by the high-frequency receiving coil 14b in the receiving system 6, amplified by the amplifier 15, shaped in waveform, and then digitized by the A / D converter 17 to be stored in the central processing unit (hereinafter referred to as "CPU"). Send it to 8. In CPU8,
An image is reconstructed and calculated based on this data, and a tomographic image of the subject 1 is displayed on a display (hereinafter referred to as “CPT”) 20. The above-mentioned high-frequency magnetic field is obtained by sending to the high-frequency transmitting coil 14a a signal sent from the sequencer 4 controlled by the CPU 8 and amplified by a high-frequency transmitting coil power supply (not shown).

In addition to the above static magnetic field and high frequency magnetic field, the magnetic resonance imaging apparatus described above is provided with a gradient magnetic field coil 9 for producing a gradient magnetic field for obtaining positional information in space. These gradient coil 9
A gradient magnetic field is generated by being supplied with a current from a gradient magnetic field power supply 10 that operates by a signal from the sequencer 4.

Here, the imaging principle of the magnetic resonance imaging apparatus will be described with reference to FIG. First, FIG.
As shown in (a), an atomic nucleus written in a static magnetic field H ₀ in the Z direction behaves like a single bar magnet in classical physics, and has the Z-axis at the Larmor frequency ν ₀ described above. Are doing precession around. This frequency is given by the above equation (1) and is proportional to the strength H _{0 of the} static magnetic field. Γ in the formulas (1) and (2) is called a gyromagnetic ratio and has a value unique to the nucleus. Generally, the number of nuclei to be measured is enormous, and each rotates in an arbitrary phase, so when viewed as a whole, the components in the XY plane cancel each other out, and the Z direction component is a macroscopic The magnetization remains. In this state, as shown in FIG. 4B, when a high frequency magnetic field H ₁ having a frequency equal to the Larmor frequency ν ₀ is applied in the X direction, the macroscopic magnetization begins to fall in the Y direction. This tilting angle is proportional to the product of the amplitude of the high frequency magnetic field H ₁ and the application time, and the high frequency magnetic field H ₁ when tilting 90 ° with respect to the pulse application time is 90 ° pulse, and the high frequency magnetic field H ₁ when tilting 180 ° is 180 °. It is called a pulse. The X, Y, and Z axes in FIGS. 4A and 4B are Cartesian coordinate axes that are orthogonal to each other.

A two-dimensional Fourier imaging method is a method generally used for imaging in such a magnetic resonance imaging apparatus. FIG. 5 is a timing diagram schematically showing a pulse sequencer of a typical spin echo method among the above two-dimensional Fourier imaging methods. Figure 5
2A, the irradiation timing of the high-frequency magnetic field signal and the envelope for selectively exciting the slice position of the subject are shown in FIG. (B) shows the application timing of the gradient magnetic field Gz in the slice direction, and (c) shows the application timing of the phase encoding direction gradient magnetic field Gy and the measurement by changing the amplitude, and (d). The figure shows the application timing of the gradient magnetic field Gx in the frequency encoding direction. Further, FIG. 7E shows the echo signal (NMR signal) to be measured.

In the pulse sequence shown in FIG. 5, first,
After applying the 90 ° pulse, a 180 ° pulse is applied at Te / 2 when the echo time is Te. the above
After the 90 ° pulse is applied, each spin starts rotating in the XY plane at its own velocity, so that a phase difference occurs between the spins over time. Here, when a 180 ° pulse is applied, each spin reverses symmetrically with respect to the X axis, and since it continues to rotate at the same speed thereafter, the spin converges again at time Te shown in FIG. 5, as shown in FIG. To form an echo signal. The signal is measured as described above, but the spatial distribution of the signal must be obtained in order to form a tomographic image. For this purpose, a linear gradient magnetic field is used. A spatial magnetic field gradient can be created by superimposing a gradient magnetic field on a uniform static magnetic field. As described above, since the spin rotation frequency is proportional to the magnetic field strength, the spin rotation frequency is spatially different in the state in which the gradient magnetic field is applied. Therefore, the position of each spin can be known by examining this frequency. For this purpose, the phase encode direction gradient magnetic field Gy and the frequency encode direction gradient magnetic field Gx shown in FIG. 5 are used.

Using the pulse sequence described above as a basic unit, the intensity of the phase-encoding direction gradient magnetic field Gy is changed every time, and is repeated a predetermined number of times, for example, 256 times, at a constant repetition time Tr. The spatial distribution of macroscopic magnetization shown in FIG. 4A is obtained by subjecting the measurement signal thus obtained to two-dimensional inverse Fourier transform. In the above explanation,
If the three types of gradient magnetic fields do not overlap with each other, X, Y, Z
Or any combination thereof. The basic principle of the above magnetic resonance imaging is described in detail in "NMR Medicine (Basic and Clinical)" (edited by the Nuclear Magnetic Resonance Medical Research Society, Maruzen Co., Ltd., issued January 20, 1984). .

Next, the principle of blood flow drawing in the nuclear magnetic resonance imaging apparatus will be described. As a method of drawing blood flow in a magnetic resonance imaging apparatus, a time-of-flight (Time-of-f
light: hereinafter abbreviated as TOF) method, phase sensitive (Ph
ase-sensitive (hereinafter abbreviated as PS) method, and phase-contrast (hereinafter abbreviated as PC) method that inverts the polarity of phase diffusion due to blood flow and makes a difference. It is used. Among these, regarding the increase in blood flow signal due to the inflow effect of the TOF method, "MagneticRe
sonance Imaging. Stark DD et al. edited, The C. V.
Mosby Company. pp108-137, 1988 ”, and for the PS method, the difference method is“ Cerebra ”MR.
Research on Angioimaging (Cerebrovascular Magnetic Resonance Imaging) -First Report- "(Keiji Fukui et al., CT Study 10 (2), 1988)
Pp. 133-142.

The PC method will be described below. In the magnetic resonance imaging apparatus, several kinds of gradient magnetic fields are applied as described above in measuring an echo signal (NMR signal), but the spins excited by the application of these gradient magnetic fields depend on the position and the moving speed. It undergoes phase rotation. That is, as shown in FIG. 6A, for example, at time t ₁ , two spins S ₁ and S ₂ exist at the position X ₀ , one spin S ₁ is stationary, and the other spin S ₁ is stationary. If S ₂ is moving in the X direction at speed v,
As shown in FIG. 6B, the gradient magnetic field −Gx in the frequency encoding direction from time t ₁ to t ₂ and time t ₂ to t.
By applying the gradient magnetic field Gx whose polarities are up to ₃ , the phase changes Φs ₊ and Φf ₊ are respectively received.

[Equation 3]

[Equation 4] Here, if Δt = t ₂ −t ₁ = t ₃ −t ₂ , the formula (4)
Can be replaced by

[Equation 5] Similarly, from time t ₁ to t ₂ , + Gx, and from time t ₂ to t _3.
Up to -Gx and the positive and negative gradient magnetic fields are applied in the reverse order, the phase changes Φs shown in equations (6) and (7), respectively.
_{-, Φf -} the subject.

[Equation 6]

[Equation 7]

From the above equations (4) to (7), the phase change of the stationary spin (S ₁ ) is canceled and the phase change of the moving spin (S ₂ ) is canceled by the application of the pair of positive and negative gradient magnetic fields. It can be seen that it is proportional to v. Such a pair of positive and negative gradient magnetic field pulses is called a flow encode pulse. If the polarity of the flow encode pulse is reversed (that is, the order of +-is reversed), the polarity of the phase rotation is also reversed as described above, and therefore the complex difference (vector A ₊ - vector a _- Taking a), stationary parts of the signal not subjected to phase rotation as shown in FIG. 7 (a vector S)
Is removed and only the blood flow signal (vector F ₊ −vector F ₋ ) is detected. The obtained signal intensity changes depending on the flow velocity, and the signal intensity becomes maximum when the flow velocity is such that the phase difference between Φf ₋ and Φf ₊ is π. Therefore, a sequence having a positive flow encode pulse and a sequence having a negative flow encode pulse (shown in FIG. 11, FIG. 12, and FIG. 13) are operated and acquired on any one axis of the Cartesian coordinate system. By performing the complex difference calculation shown in FIG. 8 on the data, the blood flow flowing along the uniaxial direction can be visualized.

By operating this on all three axes, that is, by operating six sequences, blood flow in all directions can be visualized. FIG. 14 shows an example of a sequence for drawing blood flow in the three-axis directions with six sequences. Furthermore, in order to shorten the total measurement time, the blood flow in three directions can be visualized by a sequence of operating four sequences as shown in FIG. In this case, S-axis, P-axis,
Although the flow encode pulse is applied to the F axis at the same time, the reference sequence in which the positive flow encode pulse is applied in the three directions is applied to any one of the S, P, and F axes in the other three sequences. Negative flow encode pulse is applied to, and by taking the difference between each data and reference data, it is possible to remove the influence of the flow in the other two axis directions and extract the blood flow signal of only one axis. it can. For this PC method, refer to "Magnetic resonan
ce angiography. Dumoulin CL, et al. Radiology 16
1: 717-720, 1986 ”.

Next, the blood vessel image projection method will be described.
It is possible to acquire three-dimensional blood vessel data using any of the above blood flow imaging techniques. Of course, it is possible to acquire two-dimensional data, but by acquiring three-dimensional data,
Wide range and high resolution information can be obtained. The obtained three-dimensional data set is a set of a large number of two-dimensional images each partially containing a blood vessel, and each two-dimensional image is thin slice data. Therefore, one blood vessel is divided into several slices and visualized, and it is difficult to grasp the running and shape of the blood vessel as it is.
Therefore, from these three-dimensional data, an X-ray angiographic image or a DSA (Digital Subt
raction Angiograpyh) is used to create a projected blood vessel image. In angiography in a magnetic resonance imaging apparatus, a projection image can be created by changing the viewing direction by post-processing in order to obtain a three-dimensional data set, but X-ray angiography can obtain only one two-dimensional projection image. Is. FIG. 9 shows a method of obtaining a projection image from a large number of continuous two-dimensional images each partially including a part of a blood vessel. Since the obtained data is three-dimensional data, the projection direction may be any direction. In general, projection is performed in the direction of coronal section, sagittal section, or transverse axis, but in order to obtain depth perception after the anterior-posterior relationship of blood vessels, rotation about a certain axis, for example, an angle of about ± 45 ° is used. It is easy to recognize the structure of the blood vessel by creating projected images every 5 ° to 10 ° from the projected images and displaying them as moving images.

Now, when a projection image is created, a ray trajectory method is used to project three-dimensional data from a certain viewpoint.
g) is used. When one optical axis is provided from the viewpoint to the projection plane, the blood vessel candidate on the optical axis can be regarded as having a larger signal value than the background noise. Therefore, the highest signal value on a given optical axis is very likely to be a blood vessel. Therefore, it can be easily inferred that a blood vessel image can be obtained by creating one projection image with only this maximum value. This method, maximum intensity projection (M aximum I ntensity P rojectio
n), which is the most frequently used method. The maximum intensity projection method is characterized in that it is less susceptible to noise than the method of adding pixel values on a projection line. If images of a plurality of projection angles are created using these projection methods, the human vascular system can be observed three-dimensionally.

By the way, in the PC method, a phase change proportional to the flow velocity can be extracted by making a phase difference using the same measurement data in addition to creating a blood vessel image. As shown in FIG. 8, when the data set obtained by two sequences of the positive polarity flow encode pulse sequence and the negative polarity flow encode pulse sequence is subjected to the phase difference instead of the complex difference, only the phase difference between the two sequences is obtained. Can be extracted. This phase difference data corresponds to (Φ ₊ −Φ ₋ ) in FIG. 7, but if a pixel including only blood vessels is selected, (Φ ₊ −Φ ₋ ) ≈ (Φ f ₊
−Φf ₋ ), which indicates a value proportional to the blood flow velocity.

[0017]

In the above-mentioned prior art, when calculating and displaying the flow velocity on the entire screen, as shown in FIG. Since random phase differences (random values of −π to π) are displayed in the corresponding area), the pixel values in the same area change significantly, making diagnosis difficult. In particular, when cine display is performed, the phase difference of the noise region changes for each image, which causes the screen to flicker. The present invention removes unnecessary noise in such a region that does not hit human tissue,
The purpose is to provide a good flow velocity image.

[0018]

In order to achieve the above object, in the present invention, a static magnetic field generating means for applying a static magnetic field to an object, and a gradient magnetic field generating means for applying a gradient magnetic field to the object,
A sequencer that repeatedly applies a high-frequency pulse that causes nuclear magnetic resonance to the atomic nuclei of the biological tissue of the subject in a predetermined pulse sequence, and a nuclear magnetic resonance to the atomic nucleus of the biological tissue of the subject by the high-frequency pulse from the sequencer. A transmission system that irradiates a high-frequency magnetic field to cause
A reception system for detecting an echo signal emitted by the nuclear magnetic resonance, a signal processing system for performing an image reconstruction operation using the echo signal detected by the reception system, and a display unit for displaying the reconstructed image. In a magnetic resonance imaging apparatus for obtaining a tomographic image by repeatedly measuring an echo signal emitted by nuclear magnetic resonance, the signal processing system includes a pair of complex image data measured by a phase contrast method.
The absolute value image is acquired from either one of the
Proportional to blood flow velocity from phase difference of paired complex image data
A phase image having a phase value is acquired as a velocity image, and the image data value of the phase image corresponding to an area in which the image data value is smaller than a predetermined threshold level value is fixed based on the absolute value image. By substituting the value (-
Random value of π to + π) Noise removal level
A processing means for acquiring a phase image to obtain the velocity image is provided. (Claim 1)

Further, in the present invention, the threshold level value is set by multiplying the maximum signal intensity in the image data of the absolute value image by a ratio smaller than 0.5. (Claim 2)

Further, in the present invention, the threshold level value is set by multiplying the maximum signal intensity in the image data of the absolute value image by a ratio depending on the attenuation amount upon signal reception and the bit shift amount upon Fourier transform. It was made to be done. (Claim 3)

[0021]

FIG. 1 is a diagram showing the outline of noise removal of a velocity image in the magnetic resonance imaging apparatus of the present invention. Further, FIG. 2 shows a profile for one line in each image of FIG. In the present invention, the absolute value image and the velocity image are obtained by performing the image reconstruction calculation using the data measured by the phase contrast method. Of these images, FIGS. 1 (a) and 2 (a) are absolute value images before difference, FIGS. 1 (b) and 2 (b) are velocity images before noise removal, FIG. 1 (c), and FIG. 2 (c) is a velocity image after noise removal. FIG. 2A shows a profile of one line of the absolute value image. In this figure, the central portion is a region with human tissue, and both ends are regions without human tissue (regions where an NMR signal is not originally generated). . In addition, there is a blood vessel in the center of a certain region of human tissue. Since there is a large difference in the signal intensity of a portion of human body tissue as compared with the signal intensity of a region without human body tissue, it is possible to distinguish between the two by setting an appropriate threshold level value for the signal intensity. (Besides, the bone portion is a region where an NMR signal is hard to be generated, so it is treated in the same manner as a region without human body tissue.)

FIG. 2B shows a velocity image 1 before noise removal.
It is a line profile. In a region with human body tissue, the variation in velocity data is small, but in a region without human body tissue, the variation in velocity data is large because the phase data is indefinite, which causes noise on the image. In order to remove this noise, the difference in signal intensity is used to identify the presence or absence of human tissue in FIG. 2A, and the threshold level value of signal intensity is set to an appropriate value. A value obtained by multiplying the maximum signal strength value in the absolute value image before difference by the ratio K is used as the threshold level value of the signal strength. The value of ratio K is 0 to 0.5
The range is selected. When the threshold level value is set too high, the necessary velocity information of the human body tissue will be deleted, and when it is too low, background noise cannot be completely removed. Further, the threshold level value has an optimum value that varies depending on the S / N of the target image, so it is preferable to set it depending on the image. At this time, the absolute signal strength obtained by considering the attenuation amount at the time of receiving the NMR signal and the bit shift amount at the time of Fourier transform may be used as a criterion for determining the S / N of the image.

FIG. 2C is a profile for one line of the velocity image after noise removal. In this figure,
In (a), the velocity data value at the position where the signal intensity value of each pixel of the absolute value image falls below the threshold level value is replaced with a constant value, preferably 0, and (-π
(Random value of ++ π) is removed . By performing this replacement, in the reconstructed velocity image, the NM of the air portion other than the human body tissue, bone, etc.
The indefinite phase information generated by the portion where the R signal is not generated can be removed, and a noise-free velocity image as shown in FIG. 1C can be obtained. This effect can be clearly confirmed by comparison between FIG. 1 (b) and 1 (c), and comparison between FIG. 10 (a) (with photograph) and FIG. 10 (b) (with photograph). it can.

[0024]

Embodiments of the present invention will now be described in detail with reference to the accompanying drawings. FIG. 3 is a block diagram showing the overall configuration of the magnetic resonance imaging apparatus according to the present invention. This magnetic resonance imaging apparatus obtains a tomographic image of a subject by utilizing a nuclear magnetic resonance (NMR) phenomenon, and as shown in FIG. 3, a static magnetic field generating magnet 2, a gradient magnetic field generating system 3, and a transmission magnetic field generating system 3. System 5, receiving system 6, signal processing system 7, sequencer 4,
And a central processing unit (CPU) 8. The static magnetic field generating magnet 2 is for generating a uniform static magnetic field around the subject 1 in the body axis direction or in the direction orthogonal to the body axis.
A magnetic field generating means of a permanent magnet type, a normal conducting type or a superconducting type is arranged in a space having a certain space around the subject 1. The gradient magnetic field generation system 3 includes X, Y,
It is composed of a gradient magnetic field coil 9 wound in three Z-axis directions and a gradient magnetic field power supply 10 for driving each gradient magnetic field coil, and drives the gradient magnetic field power supply 10 of each coil according to an instruction from a sequencer 4 described later. Therefore, X,
The gradient magnetic fields Gx, Gy, and Gz in the Y-axis and Z-axis directions are applied to the subject 1. The slice plane for the subject 1 can be set by the method of applying this gradient magnetic field. The sequencer 4 repeatedly applies a high-frequency magnetic field pulse that causes nuclear magnetic resonance to the atomic nuclei of the atoms constituting the biological tissue of the subject 1 in a predetermined pulse sequence, operates under the control of the CPU 8, and operates the subject 1 Various commands necessary to collect the tomographic image data are sent to the transmission system 5, the gradient magnetic field generation system 3 and the reception system 6.
The transmission system 5 irradiates a high frequency magnetic field in order to cause nuclear magnetic resonance in the atomic nuclei of the atoms constituting the biological tissue of the subject 1 by the high frequency pulse sent from the sequencer 4, and the high frequency oscillator 11 and the modulator 12 are provided. And a high-frequency amplifier 13 and a high-frequency coil 14a on the transmission side. The high-frequency pulse output from the high-frequency oscillator 11 is amplitude-modulated by a modulator 12 according to a command from the sequencer 4, and the high-frequency pulse thus amplitude-modulated is high-frequency amplifier. After being amplified in 13, the electromagnetic waves are applied to the subject 1 by being supplied to the high-frequency coil 14a arranged close to the subject 1. The receiving system 6 detects an echo signal (NMR signal) emitted by nuclear magnetic resonance of atomic nuclei of the biological tissue of the subject 1, and includes a high-frequency coil 14b on the receiving side, an amplifier 15, a quadrature detector 16, A
An electromagnetic wave (NMR signal) of a response of the subject 1 due to the electromagnetic wave emitted from the high-frequency coil 14a on the transmission side is detected by the high-frequency coil 14b arranged close to the subject 1. , Amplification 15 and quadrature detector 16 are input to A / D converter 17 to be converted into a digital quantity, and two sequences of data sampled by quadrature detector 16 at the timing instructed by sequencer 4 are collected. It is set as data, and its signal is sent to the signal processing system 7. This signal processing system 7
Is the CPU 8, the magnetic disk 18, and the magnetic tape 19.
And the like, and a display 20 such as a CRT. The CPU 8 performs Fourier transform, correction coefficient calculation, image reconstruction, and other processing, and performs an appropriate calculation on a signal intensity distribution of an arbitrary cross section or a plurality of signals. The obtained distribution is imaged and displayed on the display 20 as a tomographic image. In FIG. 3, the high-frequency coils 14 a and 14 b on the transmitting side and the receiving side and the gradient magnetic field coil 9 are installed in the magnetic field space of the static magnetic field generating magnet 2 arranged in the space around the subject 1.

Here, in the present invention, the sequencer 4 starts the phase contrast blood vessel depiction sequence. The measured data is subjected to appropriate processing, and depending on the number of axes of the flow encode for each slice, for example, 1 axis, 2 axes, 3 axes, 2, 4, 6 or 2, 3, 4, Are stored in the memory as a set. Usually, a complex difference process is subsequently performed to create an absolute value image in which only blood flow signals are extracted, and after the difference of a plurality of slices, the projection process shown in FIG. 9 is performed using the set of absolute value images (three-dimensional data). Is converted into an arbitrary two-dimensional projection image.

Further, in the present invention, a velocity image is created from the stored measurement data by the PC method described above, if necessary, by the phase difference processing shown in FIG. At this time, one absolute value image of the positive / negative flow encode data is created once in the process of processing, and the maximum signal intensity value on the image is detected. A threshold level value for removing noise in the velocity image is set based on the maximum signal strength value. The threshold level value is obtained by multiplying the maximum signal strength value by a constant ratio K. The value of the ratio K is selected as an appropriate value between 0 and 0.5. Next, each pixel value of the absolute value image is compared with the above-mentioned threshold level value, and the pixel position showing a value less than or equal to the threshold level value is stored in the memory. Subsequently, a velocity image obtained by performing the phase difference is created, and the velocity data of the stored pixel positions are all replaced with a constant value, for example, "0", and (-
Remove the random value) and a noise of π~ + π. After the replacement is completed for all pixels, the velocity image is displayed on the display (see FIGS. 1 and 2). By performing the above, a noise-free velocity image as shown in FIG. 1 (c) or FIG. 10 (b) (a photograph is attached) can be obtained.

If the threshold level value set above is too high, the velocity information of the human body tissue having a low level is deleted, and if it is too low, the background noise is not completely removed. Becomes Also, the threshold level value has an optimum value that differs depending on the S / N of the target image. For example, when the S / N is large, it is appropriate to set it to a small value, so it is set to change depending on the image. Is desirable. At this time, S /
As the criterion of N, it is preferable to use the absolute signal strength obtained by considering the attenuation amount at the time of receiving the NMR signal and the bit shift amount at the time of Fourier transform.

[0028]

Since the present invention is constructed as described above,
As shown in FIG. 10 (b) (attached with a photograph), a good blood flow velocity image without noise is created by the indefinite phase information of the tissue that does not generate the NMR signal such as the region corresponding to the background or the bone. You can

[Brief description of drawings]

FIG. 1 is a diagram showing an outline of noise removal of a flow velocity image in a magnetic resonance imaging apparatus of the present invention.

FIG. 2 is a diagram showing a profile for one line in each image of FIG.

FIG. 3 is a block diagram showing an overall configuration of the present invention and a conventional magnetic resonance imaging apparatus.

FIG. 4 is an explanatory diagram showing the behavior of nuclear spins for explaining the imaging principle of the magnetic resonance imaging apparatus.

FIG. 5 is a timing diagram schematically showing a pulse sequence of a typical spin echo method among two-dimensional Fourier imaging methods in a general magnetic resonance imaging apparatus.

FIG. 6 is an explanatory diagram showing a state of phase rotation of a moving spin and a stationary spin by a flow encode gradient magnetic field.

FIG. 7 is an explanatory diagram showing a state in which only a blood flow signal is extracted from data measured by the PC method using a phase difference.

FIG. 8 is an explanatory diagram showing two types of difference images obtained by using PC measurement data.

FIG. 9 is an explanatory diagram showing a projection process for creating a two-dimensional blood vessel image with an arbitrary projection angle from three-dimensional blood vessel data.

FIG. 10 is a diagram showing an effect of noise removal processing of the present invention with an actual speed image (with a photograph attached).

FIG. 11 is a diagram showing a sequence example in which a flow encode pulse is applied in a slice selection gradient magnetic field application direction.

FIG. 12 is a diagram showing a sequence example in which a flow encode pulse is applied in a phase encode gradient magnetic field application direction.

FIG. 13 is a diagram showing a sequence example of applying a flow encode pulse in a frequency encode gradient magnetic field application direction.

FIG. 14 is a diagram showing an example of a sequence for drawing blood flow in three axial directions with six sequences.

FIG. 15 is a diagram showing an example of a sequence for drawing blood flow in three axial directions with four sequences.

[Explanation of symbols]

1 subject 2 Magnetic field generator 3 Gradient magnetic field generation system 4 Sequencer 5 Transmission system 6 Reception system 7 Signal processing system 8 CPU 9 gradient coil 10 gradient magnetic field power supply 14a High frequency coil on transmission side 14b RF coil on receiving side

─────────────────────────────────────────────────── ─── Continuation of front page (58) Fields surveyed (Int.Cl. ⁷ , DB name) A61B 5/055

Claims (3)

(57) [Claims] 1. A static magnetic field generating means for applying a static magnetic field to a subject, a gradient magnetic field generating means for applying a gradient magnetic field to the subject, and a high frequency for causing nuclear magnetic resonance in atomic nuclei of atoms constituting the biological tissue of the subject. A sequencer that repeatedly applies a pulse in a predetermined pulse sequence, a transmission system that irradiates a high-frequency magnetic field to cause nuclear magnetic resonance in the nuclei of the biological tissue of the subject by the high-frequency pulse from the sequencer, and the nuclear magnetic resonance A nuclear magnetic system comprising a receiving system for detecting an echo signal emitted, a signal processing system for performing an image reconstruction operation using the echo signal detected by the receiving system, and a display unit for displaying the reconstructed image. In a magnetic resonance imaging apparatus for obtaining a tomographic image by repeatedly measuring an echo signal emitted by resonance, the signal processing system uses a phase contrast method. A pair of measured complex image data
The absolute value image is acquired from either one of
Phase proportional to blood flow velocity from phase difference of complex image data
By obtaining a phase image having a value as a velocity image, and replacing the image data value of the phase image corresponding to an area whose image data value is smaller than a predetermined threshold level value based on the absolute value image with a constant value − π to +
A phase image from which noise that is a random value of π has been removed
Magnetic resonance imaging apparatus characterized by acquiring and comprising processing means to the velocity image. 2. The magnetic resonance imaging apparatus according to claim 1, wherein the threshold level value is set by multiplying the maximum signal intensity in the image data of the absolute value image by a ratio smaller than 0.5. Magnetic resonance imaging apparatus. 3. The magnetic resonance imaging apparatus according to claim 1, wherein the threshold level value is the maximum signal intensity in the image data of the absolute value image, the attenuation amount at the time of signal reception and the bit shift amount at the time of Fourier transform. A magnetic resonance imaging apparatus characterized by being set by multiplying by a dependent ratio.