JPH0889493A - MRI imaging method - Google Patents

Abstract

(57) [Summary] [Object] To provide an MRI imaging method capable of acquiring images with different relaxation states of magnetization in a short time with high resolution. [Constitution] Regarding n vertical relaxation states, a part obtained by dividing each image in a measurement space is acquired during one relaxation process. By repeating this, all signals necessary to create n images are acquired. According to this method, it is possible to use many echo signals to create one image and to select and use a portion of a series of echo signals having high intensity. In addition, deterioration of time resolution can be prevented.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention measures magnetic resonance (MR) signals from atomic nuclei of hydrogen, phosphorus, carbon and the like contained in a subject, and visualizes information on tissues in a living body. The present invention relates to an MRI imaging method that enables a (MRI) apparatus to acquire a plurality of images having different longitudinal relaxation states of magnetization in a short time with high spatial resolution.

[0002]

2. Description of the Related Art An MRI apparatus acquires magnetic resonance signals from atomic nuclei such as hydrogen, phosphorus, and carbon contained in a region of interest of a subject, and noninvasively images the inside of the subject. A lot of physicochemical information is mixed in this magnetic resonance signal, and each has a clinically important meaning. One of the information is spin-lattice relaxation (hereinafter, longitudinal relaxation). The spatial distribution of the time constant T ₁ of vertical relaxation can be obtained by acquiring a plurality of images having different states of vertical relaxation. However, in the method currently used in clinical practice, it takes a lot of time to obtain the spatial distribution of T ₁ , which is a problem.

In order to solve this, the following method has been reported as a method for acquiring a plurality of images having different longitudinal relaxation states in a short time.

With respect to the magnetization of atomic nuclei in a subject placed in a static magnetic field and aligned in one direction (hereinafter referred to as the longitudinal direction),
A 180-degree inversion RF pulse is applied to invert the direction of magnetization. After the lapse of time Ta, a first θ degree (θ <90) RF pulse is applied to invert the magnetization by θ degrees with respect to the longitudinal direction. The collapsed magnetization is decomposed into a component in the vertical direction and a component in the direction orthogonal to it (hereinafter referred to as the horizontal direction), and from the component in the horizontal direction, an image in the first longitudinal relaxation state is created. Acquire all necessary echo signals by the echo planar method. Following the first θ degree (θ <90) RF pulse, θ at time Tb intervals
The application of the RF pulse is repeated n times. Every echo signal necessary for creating each image is acquired for each application of each θ-degree RF pulse. As a result, n images in different longitudinal relaxation states can be acquired during one relaxation process.

From the acquired images of different relaxation states, a distribution image of T ₁ can be obtained with the time constant T _{1 of} longitudinal relaxation and the nuclear density M ₀ as unknowns.

The detailed principle of this method is described in Magnetic Resonance in Medicine.
nance in Medicine) Volume 30, Pages 351 to 354 (19
1993).

In the echo planar method used for image acquisition in the above method, a large number of echo signals are generated by rapidly reversing the polarity of a readout gradient magnetic field applied for signal readout after applying an RF pulse. This is an imaging method.

The magnetization excited by the RF pulse decays exponentially with the passage of time. Therefore, when a plurality of echo signals are generated by applying an RF pulse once as in the echo planar method, the signal intensity becomes weaker in the later generation order. The spatial resolution of an image depends on the number of echo signals used for image creation, and in order to obtain an image with high spatial resolution, it is necessary to generate a large number of echo signals having an intensity that can be used for image creation. However, since the number of echo signals that can be generated by applying a single RF pulse is limited due to the limit of the driving capability of the gradient magnetic field, the spatial resolution of the obtained image is also limited accordingly.

As a method for solving this problem and obtaining an image with high spatial resolution, the following method has been reported.

In the application of the RF pulse once, a part of all the echo signals necessary for image formation is divided into a plurality of pieces and is acquired. The application of RF pulse is repeated by the same number as the division number,
Acquire all echo signals required for image creation. As a result, many echoes can be used to create one image, and an image with high resolution can be obtained. In this specification, this method will be referred to as measurement space division. Here, the measurement space is the application amount of the readout gradient magnetic field applied for signal readout on the horizontal axis, and the application amount of the encoding gradient magnetic field applied orthogonally thereto for adding position information on the vertical axis. The sampling points of the echo signal are arranged according to the applied amount of the gradient magnetic field. The report of the measurement space division imaging using the echo planar method is described in, for example, Japan Society for Magnetic Resonance Medicine, 11S,
175 pages (1993).

[0011]

The above-mentioned conventional techniques have the following problems, respectively.

In the first prior art, the θ degree RF pulse is used as the RF pulse applied for image acquisition. All echo signals required to create one image are obtained by the echo planar method from the magnetization excited by applying a single θ-degree RF pulse. The lateral component of the magnetization used for image creation is sin θ times the magnitude of the longitudinal magnetization (θ <9
Since 0, sin θ <1), the signal strength becomes weaker,
Fewer echo signals are available. Therefore, the signal-to-noise ratio of the image obtained by this method is low, and the spatial resolution is limited due to the limitation of the driving ability of the gradient magnetic field.

Further, when a distribution image of T ₁ is obtained from a plurality of images having different longitudinal relaxation states obtained by using the above conventional technique, if the spatial resolution of each image is low, T ₁ within one pixel Since a plurality of atomic nuclei different from each other are often included, there arises a problem that the accuracy of T ₁ obtained deteriorates. Improving the temporal resolution in obtaining the distribution of T ₁ has been discussed so far. However, at the same time, it has not been discussed to improve the spatial resolution and prevent the accuracy deterioration of T ₁ .

In the second conventional technique, all echo signals necessary for producing one image are divided and acquired, and the spatial resolution of the image is improved. This imaging method applies the same number of RF pulses as the number of divisions, but the time between RF pulses is several hundred ms.
To s for several seconds, the time resolution of image creation is reduced. When capturing an image of one longitudinal relaxation state,
This decrease in time resolution is hardly a problem, but when multiple images with different longitudinal relaxation states are acquired by this imaging method, it is necessary to change the conditions and repeat this, so the deterioration of time resolution is large. It becomes a problem.

An object of the present invention is to provide an MRI imaging method which can reduce the requirement for driving a gradient magnetic field and can acquire a plurality of images having different relaxation states in a short time with high resolution. It is to solve the two problems at the same time.

[0016]

[Means for Solving the Problems] A means for achieving the above object is to divide n images of different relaxation states in a measurement space, and to make a part of each of the n images during one relaxation process. A method of getting it is possible. Specifically, it is performed as follows.

First, the magnetization of atomic nuclei contained in the subject is reversed by a 180 degree RF pulse. After the time Ta has elapsed, the first θ degree (θ <90) RF pulse is applied. At this time, all the echo signals necessary for creating the image in the first longitudinal relaxation state are divided into a plurality in the measurement space, and a part of them is acquired. The second to nth θ-degree RF pulses are repeatedly applied at time Tb intervals from the first θ-degree RF pulse. For each application of each θ-degree RF pulse, all echo signals necessary for creating the second to nth longitudinal relaxation state images are divided in the measurement space, and a part of each is acquired. This makes it possible to acquire some of the signals necessary to create n images with different longitudinal relaxation states during one relaxation process.

After the application of the nth θ-degree RF pulse, the magnetization is waited for recovery, the magnetization is inverted again by 180 degrees, and the θ-degree RF is repeated n times.
The process of pulse application and signal acquisition is repeated multiple times. This makes it possible to acquire all signals necessary to create n images with different relaxation states.

[0019]

The means for solving the above problems have the following actions.

(1) Since the echo signals required for image formation are divided and acquired, the number of signals generated by one RF pulse can be reduced and the demand for driving the gradient magnetic field can be reduced.

(2) Since many echo signals can be used to create one image, an image with high spatial resolution can be obtained.

(3) It is possible to select and use a portion having a high intensity of a series of echo signals.

(4) With respect to a plurality of longitudinal relaxation states, a part of each image divided in the measurement space is acquired during one relaxation process, so that deterioration of time resolution can be prevented.

[0024]

1 is a block diagram showing an example of the overall configuration of a magnetic resonance apparatus according to the present invention. This magnetic resonance device is
A tomographic image of the subject 101 is obtained by utilizing the R phenomenon, and the static magnetic field generating magnet 102, the signal processing unit 103, the RF transmitting unit 104, the RF receiving unit 105, the gradient magnetic field generating unit 106, the display unit 107, and these are provided. It comprises a control unit 108 for controlling.

The static magnetic field generating magnet 102 is for generating a strong and uniform static magnetic field in the space where the subject 101 is arranged.
It is arranged in the space around the subject 101. The output of the RF transmitter 104 is sent to the transmitter coil 109 to generate an RF magnetic field. The output of the gradient magnetic field generator 106 is sent to the gradient magnetic field coil 110 and generates gradient magnetic fields Gx, Gy, Gz in three directions of X, Y, Z. A tomographic plane with respect to the subject 101 can be set depending on how to apply the RF magnetic field and the gradient magnetic field. The RF receiving unit 105 includes the receiving coil 1
11 signal is received. The output of the RF receiving unit 105 is processed by the signal processing unit 103 such as image reconstruction, and then displayed on the display unit 107. Here, the image reconstruction means a process of creating one image by performing a two-dimensional complex Fourier transform in the measurement space. Note that, in FIG. 1, the transmission coil 109, the reception coil 111, and the gradient magnetic field coil 110 are arranged in a space around the subject 101.

FIG. 2 shows a sequence (RF
Pulse and gradient magnetic field application procedure) will be described. According to this sequence, in order to obtain n 32 × 32 matrix images with different relaxation states, measurement is performed in the following procedure.

(1) A 180-degree RF pulse 201 is applied to reverse the magnetization of atomic nuclei in the subject.

(2) A dephasing gradient magnetic field 203 is applied to cancel the remaining lateral magnetization.

(3) The θ-degree RF pulse 202 and the slice selection gradient magnetic field 204 are applied simultaneously to determine the target slice of the subject, and the magnetization of the nuclei in the slice is inverted by θ degrees with respect to the vertical direction.

(4) Rephase gradient magnetic field 205 for the slice selection gradient magnetic field 204, and dephase gradient magnetic field 20 for the encode gradient magnetic field for adding position information
6, and a dephasing gradient magnetic field 208 for the readout gradient magnetic field for signal reading is applied.

(5) The encode gradient magnetic field 207 and the readout gradient magnetic field 209 are applied as shown in the figure. At this time, echo signals for 8 encodings are acquired from the echo signals necessary for image creation.

(6) By a procedure 210 in which steps (2) to (5) are sequentially repeated n times, eight encoded echo signals are obtained for each of n images having different relaxation states.

(7) The magnitude of the dephasing gradient magnetic field 206 is changed, and the procedure 210 is repeated four times. This allows
It is possible to acquire all the echo signals necessary to create n images.

In the sequence of FIG. 2, the time between the 180 degree RF pulse 201 and the θ degree RF pulse 202 is Ta, θ.
The time between the RF pulses 202 is Tb.

FIG. 3 shows the behavior of longitudinal magnetization in the procedure 210 when the sequence of FIG. 2 is executed in accordance with the RF pulse application timing. Apply 180 degree RF pulse 305 to the initial magnetization 301 to set the magnetization to 1
Invert 80 degrees. The reversed magnetization is longitudinally relaxed during the time Ta to become the magnetization 302. The first θ-degree RF pulse 306 is applied to the magnetization 302 to reverse the magnetization by θ degrees. θ
The lateral component of the magnetized magnetization is detected by the echo planar method as a signal that contributes to image formation in the first relaxation state. The remaining vertical component is further longitudinally relaxed during the time Tb to become the magnetization 303. In the same manner, the acquisition of a signal for the nth magnetization 304 is repeated, and a part of the echo signals necessary for creating n images with different relaxation states can be acquired.

FIG. 4 is an explanatory diagram in which signals acquired when eight images having different relaxation states are obtained by executing the sequence of FIG. The images up to the first image 403 are shown. Here, the measurement space is the application amount of the readout gradient magnetic field on the horizontal axis kx,
The vertical axis ky represents the application amount of the encoding gradient magnetic field, and the sampling points 401 of the echo signal are arranged according to the application amount of the gradient magnetic field. As shown in the figure, the sampling point 401 of the echo signal is the lines 404, 405,
They are arranged on 406 and 407, respectively, and the measurement is sequentially performed along this line. This line is referred to herein as the signal acquisition sequence. Each image is measured from the first sheet to the eighth sheet in sequence for one signal acquisition sequence 404, and then from the first sheet to the eighth sheet for the next signal acquisition sequence 405.
By repeating this for 406 and 407, all signals of n images can be acquired.

FIG. 5 shows another example of the sequence for realizing the present invention. In order to obtain n 32 × 32 matrix images with different relaxation states by this sequence,
The measurement is performed according to the following procedure.

(1) A 180 degree RF pulse 501 is applied to reverse the magnetization of atomic nuclei in the subject.

(2) A dephasing gradient magnetic field 503 is applied to cancel the residual transverse magnetization.

(3) The θ-degree RF pulse 502 and the slice selection gradient magnetic field 504 are applied simultaneously to determine the target slice of the subject, and the magnetization of the nuclei in the slice is inverted by θ degrees.

(4) Rephase gradient magnetic field 505 for slice selection gradient magnetic field 504, Dephase gradient magnetic field 50 for encode gradient magnetic field for adding position information
6 is applied.

(5) An encode gradient magnetic field 507 and a read-out gradient magnetic field 508 are applied as shown in FIG. At this time, echo signals for 8 encodings are acquired from the echo signals necessary for image creation.

(6) By a procedure 509 in which steps (2) to (5) are sequentially repeated n times, echo signals of 8 encodings are obtained for each of n images having different relaxation states.

(7) The magnitude of the dephasing gradient magnetic field 506 is changed, and the procedure 509 is repeated four times. This allows
It is possible to acquire all the echo signals necessary to create n images.

In the sequence of FIG. 5, the time between the 180 degree RF pulse 501 and the θ degree RF pulse 502 is Ta, θ.
The time between the RF pulses 502 is Tb.

The sequence shown in FIG. 5 is different from the sequence shown in FIG. 2 in that the encoding gradient magnetic field and the readout gradient magnetic field are applied, and the arrangement of sampling points of measured echo signals in the measurement space is different. . This is shown in FIG.

FIG. 6 shows the arrangement of signals in the measurement space when the sequence of FIG. 5 is executed and eight images of different relaxation states are obtained. The measurement space 603 of the eye image is shown. A / D
Sampling point 601 is the signal acquisition sequence 604, 6
Measured in sequence along 05, 606, 607. Each image is measured from the first sheet to the eighth sheet in sequence for one signal acquisition sequence 604, and then from the first sheet to the eighth sheet in the next signal acquisition sequence 605. This is 606
By repeating 607, all signals of eight images can be acquired.

The time constant T ₁ of the longitudinal relaxation time can be obtained by using a plurality of images having different relaxation states obtained by the above method. The method is shown below. Equation 1 is an expression expressing the behavior of the longitudinal magnetization shown in FIG. 3, where Mn is the magnitude of magnetization before the nth θ-degree RF pulse is applied, M ₀ is the magnitude of the first magnetization, and n is the number of images to be taken. Is represented. In equation 1, α is cos
θ 1 and β 2 are equations 2, γ is equation 3 and Ta and Tb are relaxation waiting times. In addition, ψ is the angle at which the nuclear magnetization falls first
It is represented by. In the _equation 4, θ ₀ represents an angle set by the device for inclining the nuclear magnetization, and θ represents an angle at which the nuclear magnetization actually falls. In Equation 1, attention is paid to each pixel of n (in this case, n ≧ 3) images, and from the signal strength change, T ₁ , M
Each value can be determined by performing non-linear least-squares approximation with ₀ and θ as unknowns.

[0049]

[Equation 1]

[0050]

[Equation 2]

[0051]

(Equation 3)

[0052]

[Equation 4]

Angle θ at which the magnetization falls by the RF pulse
Is proportional to the intensity of the applied RF pulse and varies depending on the location in the subject. Moreover, it is generally difficult to determine the absolute value of θ. The situation is shown below.

FIG. 7 shows an example of the RF magnetic field strength distribution. This is the RF intensity distribution created by the circular surface coil 701 having a diameter of 12.5 cm and one turn. According to Equation ₁ , T ₁ can be determined even if the RF intensity is nonuniform. Further, the angle θ at which the magnetization actually falls, whose absolute value is difficult to determine, can be calculated according to Equation 1.

FIG. 8 is a diagram showing changes in longitudinal magnetization when two atomic nuclei having different T ₁ are mixed in the same pixel. A state 801 of change in longitudinal magnetization when a single atomic nucleus such that T ₁ = T _1A exists in one pixel, and T ₁
The state 803 of change in longitudinal magnetization when a single atomic nucleus such as = T _1B exists in one pixel, whereas when these two atomic nuclei exist in the same pixel, the signal intensity change of that pixel is changed. _{Shows the} same change 803 as in the nucleus where apparently T ₁ = T _1k . At this time, T _1k is a value expressed by Equation 5 using T _1A and T _1B .

[0056]

(Equation 5)

When the obtained plural images having different longitudinal relaxation states have low spatial resolutions, and thus a plurality of atomic nuclei having different T ₁ are mixed in the same pixel, the accuracy of the obtained T ₁ distribution deteriorates. . This method is also effective for this problem.

[0058]

According to the present invention, it is possible to provide an MRI imaging method capable of acquiring a plurality of images having different vertical relaxation states in a short time with high resolution.

[Brief description of drawings]

FIG. 1 is a block diagram showing a device configuration for implementing the present invention.

FIG. 2 is a diagram showing a sequence of the first embodiment of the present invention.

FIG. 3 is a timing chart showing longitudinal magnetization in an example of the present invention.

FIG. 4 is an explanatory diagram showing an array of acquisition signals in the embodiment of the present invention.

FIG. 5 is a diagram showing a sequence of a second embodiment of the present invention.

FIG. 6 is an explanatory diagram showing an array of acquisition signals in the second embodiment of the present invention.

FIG. 7 is a diagram showing an example of magnetic field strength of RF.

FIG. 8 is a graph showing changes in longitudinal magnetization when two atomic nuclei having different Ti are mixed in the same pixel.

[Explanation of symbols]

201, 202 ... RF pulse, 203 ... Spoiler gradient magnetic field, 204 ... Slice selection gradient magnetic field, 205 ... Rephase gradient magnetic field, 206, 208 ... Dephase gradient magnetic field, 207 ... Encoding gradient magnetic field, 209 ... Readout gradient magnetic field, 210 ... One pattern of signal acquisition.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Etsuji Yamamoto 1-280, Higashi Koigokubo, Kokubunji, Tokyo Inside the Central Research Laboratory, Hitachi, Ltd.

Claims (3)

[Claims] 1. An MRI imaging method for acquiring an image in which longitudinal relaxation states of magnetization of atomic nuclei are different from each other, 180 degree RF
Step 1 of applying a pulse to reverse the magnetization of atomic nuclei in the subject, and after the time Ta of Step 1 has elapsed,
a step 2 of applying a θ degree (θ <90) RF pulse and acquiring a part of all echo signals necessary for creating one image by an echo planar method at a time Tb interval n times; An MRI imaging method characterized in that n steps of images having different longitudinal relaxation states are acquired by repeating the step 1 and the step 2 with a magnetization recovery period interposed therebetween. 2. A with different n images of relaxed state acquired by the imaging method according to claim 1, determine the constant T ₁ time of spin density M _0, longitudinal relaxation for each pixel, M _0, T _An MRI imaging method for obtaining a distribution image of ₁ . 3. A spin density M ₀ , using n images of different relaxation states acquired by the imaging method according to claim 1.
An MRI imaging method for obtaining a distribution image by obtaining M ₀ , T ₁ , and θ for each pixel by a non-linear least-squares approximation method using three parameters of a longitudinal relaxation time constant T ₁ and a magnetization inclination angle θ.