JP2002052005A - Magnetic resonance imaging method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To dynamic-display a water-fat separation image in real time in a method of obtaining a fat suppression image by image calculations. SOLUTION: In a magnetic resonance imaging method, plural image data different in echo time are obtained, and a water-fat separation image is obtained by operation. Every time one image data is obtained, the operation is conducted between plural image data combined with the preceding obtained image data, and one water-fat separation image is displayed in the substantially same timing as the timing of obtaining one image data.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a magnetic resonance imaging method for obtaining a water / fat separation image using a magnetic resonance imaging apparatus (hereinafter, referred to as an MRI apparatus).
In particular, the present invention relates to a magnetic resonance imaging method for acquiring water / fat separated images continuously in time series.

[0002]

2. Description of the Related Art A magnetic resonance imaging method using an MRI apparatus targets a main constituent substance of a subject, proton, and images a spatial distribution of a proton density and a spatial distribution of a relaxation phenomenon of an excited state. Two-dimensional or three-dimensional imaging of the form or function of the human head, abdomen, limbs, etc.

In an MRI apparatus, images having various tissue contrasts can be obtained by utilizing the difference in the behavior of protons contained in each tissue. In the clinic, there are many cases where an image in which an MR signal due to fat is suppressed is often required, and various methods for obtaining an image in which fat is suppressed have been put to practical use. As an example, there is a method of acquiring a plurality of images having different echo times (TE) and obtaining a water / fat separation image by calculating between the images.

As a typical method, "" Simple Pro
ton Spectroscopic Imaging "; W. Thomas Dixon et al; RAD
IOLOGY, Vol. 153, 189-194 (1984) ". As a method other than the Dixon method for obtaining a water / fat separated image by calculation, there is a method described in the following literature: "" Water-Fat Imagin
g with Three-Point Direct Phase Encoding ”; Qing-
San Xiang and Li An; Proc. , SMR 3rd Meeting, 658
(1995) ",""Quadrature 2-Point Water-Fat Imagi"
ng ”; Li An and Qing-San Xiang; Proc., SMR 4thMee
ting, 1541 (1996) ",""Water-Fat Imaging with T
hree Orthogonal-Phase Acquisitions ”; Li An and Qi
ng-San Xiang; Proc. , ISMRM 6th Scientific Meetin
g, 1866 (1998). "

[0005] In these methods, since the echo signal is measured at the time when the phases of the water protons and the fat protons coincide with each other and at the time when the phases are shifted by a predetermined amount, the method is easily affected by the non-uniformity of the static magnetic field. For this reason, a method of calculating the rotation component due to the non-uniformity of the static magnetic field by calculating the signals having different echo times and correcting the influence of the non-uniform static magnetic field (two-point Dixon method with static magnetic field non-uniformity correction function, three-point Dixon method) Etc.) are known.

On the other hand, with the MRI apparatus, in order to examine the motor function of the limb, a method of repeating imaging while moving the joint, or using the MRI apparatus as monitoring when puncturing a tumor or the like, while performing puncturing, etc. A technique of continuously photographing and displaying a chronologically continuous image in near real time has been developed and put into practical use. Such a method of continuously acquiring images is known as dynamic imaging, and a method of realizing dynamic imaging in various imaging sequences has been proposed.

[0007] There is also a demand to separate water and fat in this dynamic imaging. For example, dynamic imaging of the coronary artery, which suppresses fat signals in the image to depict the tumor with high contrast during puncture,
There is a demand for rendering the coronary artery with high contrast with respect to the surrounding fat.

[0008]

However, when the above-mentioned method for obtaining a water / fat separation image is applied to dynamic imaging, image processing time becomes a problem. That is, in the water / fat separation image imaging method, a plurality of image data having different echo times must be acquired in order to obtain one water / fat separation image, and furthermore, it is necessary to perform an operation between the image data after acquiring the image data. Therefore, there is a problem that the interval at which one image is displayed becomes long. In particular, in dynamic imaging using an open type MRI apparatus having relatively large static magnetic field inhomogeneity, since the correction after the image data acquisition requires correction of the static magnetic field inhomogeneity, the calculation time is further increased, and the image processing time becomes longer. Even if data can be acquired with a relatively high time resolution, an image cannot be displayed with the time resolution of image data acquisition.

Accordingly, an object of the present invention is to realize near real-time processing when a water / fat separation image is captured by dynamic imaging. That is, an object of the present invention is to display an image of good image quality in a water / fat separated image imaging method in which the influence of a non-uniform static magnetic field is eliminated in near real time.

[0010]

In order to achieve the above object, a magnetic resonance imaging method of the present invention acquires a plurality of image data for each echo time from a plurality of nuclear magnetic resonance signals having different echo times, In a magnetic resonance imaging method for obtaining a water / fat separation image by an operation between these images, a series of measurements for acquiring the plurality of image data is repeatedly performed, and after the first measurement, every time one image data is acquired. In addition, an operation is performed between a plurality of image data combined with the image data acquired earlier,
It is characterized in that a set of water / fat separated images is displayed at substantially the same timing as when acquiring one image data.

According to this method, the acquisition of image data and the calculation between images are performed in parallel, so that each time one image data is acquired, one set of water / fat separated images is obtained.
Therefore, the water / fat separated image can be displayed at substantially the same timing as the acquisition of one image data.

In the magnetic resonance imaging method of the present invention, a two-point Dixon method for obtaining a water / fat separation image by calculating two image data having different echo times,
Obtain water / fat separation image by calculating image data 3
It can be applied to any of the point Dixon methods.

According to the magnetic resonance imaging method of the present invention, when a water / fat separation image is obtained by calculating three image data having different echo times, a series of measurement is performed for image data used only for static magnetic field non-uniformity correction. Instead of acquiring every time, it is possible to acquire intermittently.

That is, in the measurement for acquiring a plurality of image data, at least two image data of image data having different echo times are repeatedly acquired, and image data other than the two image data is intermittently acquired. get.

When the three-point Dixon method is applied, two-point Dixon
Although the measurement time is longer than that of the method, the measurement time, that is, the image update interval can be shortened by thinning out image data that is not directly required for the calculation between images.

[0016]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 3 is a diagram showing the configuration of a typical MRI apparatus for performing the magnetic resonance imaging method of the present invention. The MRI apparatus includes a magnet 302 that generates a static magnetic field in a space where a subject 301 is placed, a gradient coil 303 that generates a gradient magnetic field in this space, and an RF coil 304 that generates a high-frequency magnetic field in an imaging region of the subject. And an RF probe 305 for detecting a nuclear magnetic resonance signal generated by the subject 301. Further, a bed 312 for transporting the subject 301 to the static magnetic field space is provided.

The gradient magnetic field coil 303 is constituted by gradient magnetic field coils in three directions of X, Y and Z, and generates a gradient magnetic field in accordance with a signal from the gradient magnetic field power supply 309. RF coil 304
Generates a high-frequency magnetic field according to the signal of the RF transmission unit 310.

The signal from the RF probe 305 is sent to the signal detector 3
The signal is detected at 06, processed by the signal processing unit 307, and converted into an image signal by calculation. The image is displayed on the display unit 308.

The control unit 311 controls the gradient magnetic field power supply 309, the RF transmission unit 310, and the signal detection unit 306. The control unit 311 controls the application of the gradient magnetic field and the high-frequency magnetic field and the timing of signal detection according to a predetermined pulse sequence determined by an imaging method. . This pulse sequence is set in the memory of the control unit 311 in advance. In the present invention, a sequence for obtaining a water / fat separation image is stored as a pulse sequence.

Next, a first embodiment of the magnetic resonance imaging method of the present invention will be described. In this embodiment,
A pulse sequence based on the two-point Dixon method is executed as a pulse sequence for obtaining a water / fat separation image. FIG. 1 shows a sequence of the two-point Dixon method based on the gradient echo method.

In this sequence, first, a high-frequency pulse 101 for selectively exciting a specific section of the subject is applied, and then a gradient magnetic field pulse 102 is applied so as to obtain a gradient echo signal after a predetermined echo time TE1. Measure the echo signal. Next, the same echo signal measurement is repeated by changing the echo time. Although the gradient magnetic field for selecting a section and the gradient magnetic field for phase-encoding these echo signals are omitted in the figure, the measurement of the echo signals includes the application of these gradient magnetic fields.

Here, the echo time TE1 of the echo signal acquired for the first time is given by: TE1 = 2nτ (n, where Δf is the difference between the resonance frequencies of water protons and fat protons, and 2τ = 1 / Δf.
Is a positive number, the same applies hereinafter), and the second echo time
TE2 is set to TE2 = TE1 + τ.

Assuming that the difference between the resonance frequencies is Δf (=） τ), the water protons and the fat protons which are oriented in the same direction at the time of excitation will be oriented in the same direction after 2τ, and
In each case, they turn in the opposite direction (180 °), the same direction (360 °), ... Therefore, by setting the echo time as described above, the first echo signal has the same phase of the water proton and the fat proton, and the second echo signal has the phase difference of 180 ° between the water proton and the fat proton. Become. In the figure, a water signal is indicated by a black arrow, and a fat signal is indicated by a white arrow. At the time of the first imaging, the water signal 104 and the fat signal 105 have the same phase. On the other hand, during the second imaging, the water signal 106 and the fat signal 107 have opposite phases.

Such a sequence is performed plural times, for example, 128 times, 256 times while changing the magnitude of the phase encoding gradient magnetic field.
By repeating, for example, times, image data consisting of the signal of the echo time TE1 and image data consisting of the signal of the echo time TE2, two image data in total are obtained.

Although a gradient echo type pulse sequence is shown in FIG. 1, two image data can be similarly obtained by a spin echo type pulse sequence as shown in FIG. In the case of the spin-echo type pulse sequence, a high-frequency pulse 202 for inverting the spin excited by the high-frequency pulse 201 is used.
Assuming that after / 2, the phases of water protons and fat protons are aligned after TE from 201 after application, so the signal is measured at this point when the phase rotation difference based on the resonance frequency difference becomes 360 ° and 180 °. Should be set to the echo time. Ie
TE can be set arbitrarily, and the difference between the first signal measurement time and the second signal measurement time becomes τ.

Next, image processing for obtaining a water / fat separation image from image data obtained by executing the pulse sequence shown in FIG. 1 or 2 will be described.

The image (x,
Assuming that the magnitude of the signal due to water and the magnitude of the signal due to fat at the position y) are W (x, y) and F (x, y), respectively, the first and second signals Sl (x, y), S2 (X, y) are respectively expressed as follows. S1 (x, y) = W (x, y) + F (x, y) (1) S2 (x, y) = W (x, y) −F (x, y) (2) Then, S1 (x , y) + S2 (x, y) = 2W (x, y) (3) S1 (x, y) −S2 (x, y) = 2F (x, y) (4) (x, y) gives a fat image F (x, y) as a subtraction image.

However, since the static magnetic field generated by the static magnetic field magnet is actually non-uniform, the above equations (1) and (2) are as follows. Sl (x, y) = (W (x, y) + F (x, y)) exp (iα (x, y)) (5) S2 (x, y) = (W (x, y) -F ( x, y)) exp (i (α (x, y) + α ′ (x, y))) (6) In equation (5), α (x, y) and α ′ (x, y) are both static. Phase rotation component due to non-uniform magnetic field, depending on position.

Therefore, when calculating image data obtained by the above-mentioned two-point Dixon method, a water image and a fat image cannot be obtained only by simply adding or subtracting according to the above equations (3) and (4). . Therefore, prior to the calculation between the images, the phase rotation amount due to the non-uniform static magnetic field is obtained from the phase difference between the signal S1 and the signal S2, and the phase rotation amount due to the non-uniform static magnetic field is corrected for the signal S2 (x, y). I do.

This process is described in the following document: "" Two-Point Dixon Technique for Water-Fa
t Signal Decomposition with Bo Inhomogeneity Corre
ction “; Bernard D. Cooms et al .; Magnetic Resonance in
Medicine, Vol. 38. 884-889 (1997) ". When the phase rotation amount due to the non-uniformity of the static magnetic field is obtained, a discontinuous jump (around the main value of the phase) may occur in the phase value, and an unwrap process is performed to remove the jump. Further, since the unwrapping process is easily affected by noise, a process of masking in advance an easily affected region to remove the effect of noise is performed. FIG.
Shown in

In the present invention, such water / fat separation photographing is performed as dynamic imaging, and a method thereof will be described below.

FIG. 5 is a diagram conceptually showing dynamic imaging according to the present invention. In the pulse sequence shown in FIG. 1 or FIG. 2, measurement for acquiring image data of the first echo time TE1 (measurement of TE1) is a white rectangle. The measurement for acquiring the image data of the second echo time TE2 (the measurement of TE2) is indicated by a hatched rectangle. The horizontal axis is the time axis. That is, in the measurement of TE1 indicated by the white rectangle, the pulse sequence for measuring the echo signal at the echo time TE1 is repeated for a required number of phase encodes, for example, 128 times, and the image data for the echo time TE1 is obtained. In the measurement of TE2 indicated by the hatched rectangle, the pulse sequence for measuring the echo signal at the echo time TE2 is a required phase encode number, for example, 128.
The image data for the echo time TE2 is obtained repeatedly and repeatedly. To obtain water / fat separation images continuously, TE
Image data is acquired in the order of l, TE2, TEl, TE2,.

When the first image data for TE1 and the first image data for TE2 are obtained, the image calculation processing shown in FIG. 4 is started. In parallel with this image calculation processing, TEl, TE2
Is continued. As described above, the time efficiency can be improved by performing the shooting and the image calculation processing in parallel. In FIG. 5, the numbers 610, 611, 612,... In Kakko below the measurements 601, 602, 603,... Are sequentially using two pieces of image data obtained by the two measurements made by Kakko. This means performing image calculation processing. That is, using the image data obtained in the measurement 601 and the image data obtained in the measurement 602, the phase rotation amount of the signal due to the non-uniformity of the static magnetic field is obtained, and the measurement is performed.
A phase correction process is performed on the image data obtained in 602, and a water / fat separation image is obtained by an inter-image operation between the corrected image data and the image data obtained in the measurement 601 (process 610), and this is an MRI apparatus Display on the monitor.

As described above, the next measurements 603 and 604 are in progress during such calculation, and when the measurement 603 is completed, the image data obtained by the previous measurement 602 and the measurement 6 are completed.
Image processing is performed using the image data obtained in step 03 (step 611) to obtain a second set of water / fat separated images. This image replaces the image previously displayed on the monitor and updates the image.

Thereafter, every time measurement is performed for one echo time, the image data obtained by the measurement and the image data obtained by the immediately preceding measurement are subjected to image calculation processing to obtain a new image. Obtain a proper water / fat separation image and update the display image. In this way, a new image can be updated at a time interval substantially equal to the time interval of measurement for one echo time. Such a measurement is repeated a desired number of times for the time to be monitored.

As described above, the case where two image data with different echo times are obtained, that is, the embodiment applied to the two-point Dixon method has been described. However, as shown in FIG. The same can be applied to a case where three image data are repeatedly obtained (three-point Dixon method).

Hereinafter, application to the three-point Dixon method will be described. In the three-point Dixon method, the measurement is performed three times while changing the TE. 1
The second and the second measurement are the same as the two-point Dixon method,
For the high-frequency excitation pulse 401, TE is 2τ in the first measurement.
And a read gradient magnetic field pulse 402 is applied. In the second measurement, the readout gradient magnetic field pulse 403 is applied so that TE becomes longer by τ than in the first measurement. In the third imaging, TE is further extended by τ from the time of the second imaging (2τ from the time of the first imaging), and a readout gradient magnetic field pulse 404 is applied.

At the time of the first photographing, the water signal 405 and the fat signal
406 has the same phase and phase 407. This value is α. Similarly, at the time of the second imaging, the water signal 408 and the fat signal 409 have opposite phases. The phase of the water signal is 410 and its value is α +
α '. At the time of the third imaging, the water signal 411 and the fat signal
412 has the same phase again, and the value of phase 413 is α + 2α ′. The signal S3 (x, y) at the time of the third imaging is expressed as follows. S3 (x, y) = (W (x, y) + F (x, y)) exp (iα (x, y) + 2α ′ (x, y)) (7)

Since the water signal and the fat signal have the same phase at the time of the first and third photographing, the phase of S3 (x, y) / Sl (x, y) is obtained to determine the phase of the static magnetic field. The amount of phase rotation can be determined. arg (S3 (x, y) / Sl (x, y)) = 2α '(x, y) (8) arg () means to obtain the phase.

After obtaining the value of equation (8) for all (x, y), and performing an unwrapping process for removing around the main value, 2
To obtain the amount of phase rotation α ′ (x, y) due to the inhomogeneity of the static magnetic field. Using the obtained α ′ (x, y), the signal S2 obtained in the second measurement is corrected, and S2 ′ (x, y) = S2 (x, y) exp (−iα ′ (x, y)) = (W (x, y) −F (x, y)) exp (iα (x, y)) (9) By calculating (9), a water image is obtained as an addition image and a fat image is obtained as a subtraction image. Can be FIG. 7 shows a flow of image processing by the three-point Dixon method.

It is to be noted that even in the three-point Dixon method, the G shown in FIG.
In addition to E-sequence, high-frequency pulse 401 and gradient magnetic field 40
It is possible to use an SE sequence in which an inverted high-frequency pulse is added between the two.

FIG. 8 shows T using the three-point Dixon method.
FIG. 7 is a diagram conceptually illustrating dynamic imaging for continuously obtaining water / fat separated images from three pieces of image data having different Es. Again, white rectangles 701, 704, and 707 indicate measurement for acquiring image data for the first echo time TE1, and hatched rectangles 702 and 705 indicate measurement for acquiring image data for the second echo time TE2. Shown, hatched rectangle 70
Reference numerals 3 and 706 denote measurements for acquiring image data for the third echo time TE3. The horizontal axis is the time axis.

In this embodiment, TE1, TE2, TE3, TE1,
Repeat the measurement for image data with different echo times, such as TE2, TE3,.
When steps 1, 702 and 703 are completed, the image processing shown in FIG. 7 is performed using these three pieces of image data to obtain a water / fat separated image. This is displayed on the monitor of the MRI apparatus.

In parallel with this image processing, the next measurement 704, 705
, And when the measurement 704 is completed, image processing is performed using the image data obtained by the measurement 704 and the image data obtained by the previous measurements 702 and 703, and the second set of water
Obtain a fat separation image. This image is displayed on the
Replace with the fat image and update the image. Hereinafter, similarly, 1
Each time one measurement is completed, image processing is performed by combining the image data obtained by the two immediately preceding measurements to obtain a set of water / fat images. In this way, as in the case of the two-point Dixon method, a new image can be created and updated at substantially the same time interval as one measurement interval.

As described above, as a first embodiment of the present invention, two or three pieces of image data having different echo times are continuously obtained, and each time one piece of image data is obtained, the image data is combined with the immediately preceding image data. The dynamic imaging method for image processing has been described. However, when a water / fat separated image is obtained by calculating three image data, image data that is not directly reflected on the image is intermittently acquired, and the imaging time and image The update interval can be shortened. Less than,
Such an embodiment will be described.

FIG. 9 shows a second embodiment of the present invention in which a part of image data is measured intermittently. Also in this figure, white rectangles 701, 704, and 707 'indicate measurement for acquiring image data for the first echo time TE1 (measurement of TE1), and hatched rectangles 702 and 705 indicate the measurement for the second echo time TE2. Indicates the measurement (TE2 measurement) to acquire image data,
A hatched rectangle 703 indicates measurement for acquiring image data for the third echo time TE3 (measurement of TE3). The horizontal axis is the time axis.

As described in the image processing in the three-point Dixon method, the image data obtained by the measurement of TE3 is used only for the purpose of obtaining the inhomogeneous distribution of the static magnetic field. Because of this, the magnet
If the static magnetic field in the space in 302 is stable and does not fluctuate much over time, the measurement of TE3 can be thinned out appropriately. In this embodiment, the measurement of TE1 and the measurement of TE2 are sequentially repeated, but the measurement of TE3 is not repeated at the same frequency, but is performed once every two repetitions of the measurement of TE1 and TE2.

After the first measurements 701 to 703 are completed, image processing 710 is performed using the three pieces of image data obtained by these measurements.
In the subsequent measurements 704 and 705, each time one measurement is completed, image processing 711 and 712 are performed by combining image data obtained in the previous measurement, as in the embodiment of FIG.
Subsequent to measurement 705, measurement of TE1 is performed instead of measurement of TE3.

In the image processing 713 ', a water / fat separation image is created using the image data obtained by the measurement 707' and the TE2 image data obtained by the immediately preceding measurement 705. The static magnetic field inhomogeneity map data for correcting the phase of the TE2 image data may be created by combining the data acquired by the measurement 703 and the measurement 707 ′, or the image processing immediately before
The one created in 712 may be used as it is.

Thereafter, each time the image data of TE1, TE2, and TE3 is continuously obtained, the static magnetic field inhomogeneity map data is updated using the image data of TE3 in the image processing executed at that time. Until the TE3 image data is obtained, a water / fat separation image is created by image processing using the updated static magnetic field inhomogeneity map data. As a result, the overall imaging time can be reduced, and the real-time performance of image display can be improved. If there is little or negligible temporal fluctuation of the static magnetic field, TE3
Can be made less frequently than that shown in the figure, and it is possible to perform only the first one.

According to this embodiment, by making the measurement of TE3 intermittent, three repetitions are regarded as one unit.
In dynamic imaging to which the point Dixon method is applied, it is possible to reduce the imaging time and the image update interval.

Although the dynamic imaging method for obtaining a water / fat separated image using two or three image data having different TEs has been described above, the present invention uses four or more image data having different TEs. The same applies to the case where water / fat separation image processing is performed by calculation.

[0053]

According to the present invention, in the dynamic imaging for continuously obtaining water / fat separated images by repeatedly performing an imaging method for obtaining a set of water / fat separated images by arithmetic processing of several image data, A set of water / fat separated images can be obtained and updated at the timing of acquiring one piece of image data. Thereby, the real-time property of the water fat separation image display can be improved.

[Brief description of the drawings]

FIG. 1 is a diagram showing an example of a data acquisition time chart in a two-point Dixon method.

FIG. 2 is a diagram showing another example of a data acquisition time chart in the two-point Dixon method.

FIG. 3 is a diagram showing an outline of an MRI apparatus to which the present invention is applied.

FIG. 4 is a flowchart showing an example of image processing in the two-point Dixon method.

FIG. 5 is a diagram showing an example of dynamic imaging according to the present invention to which the two-point Dixon method is applied.

FIG. 6 is a diagram showing an example of a data acquisition time chart in the three-point Dixon method.

FIG. 7 is a flowchart showing an example of image processing in the three-point Dixon method.

FIG. 8 is a diagram showing an example of dynamic imaging according to the first embodiment of the present invention to which the three-point Dixon method is applied.

FIG. 9 is a diagram showing an example of dynamic imaging according to a second embodiment of the present invention to which a three-point Dixon method is applied.

[Explanation of symbols]

301: subject, 302: static magnetic field magnet, 303: gradient coil, 304: RF coil, 305: RF probe, 306: signal detector, 307: signal Processing unit, 308 ・ ・ ・ Display unit, 311 ・ ・ ・ Control unit

Claims (2)

[Claims] 1. A magnetic resonance imaging method for acquiring a plurality of image data for each echo time from a plurality of nuclear magnetic resonance signals having different echo times, and obtaining a water / fat separation image by an operation between these images, A series of measurements for acquiring a plurality of image data is repeated, and after the first measurement, every time one image data is acquired, a plurality of image data obtained by combining the previously acquired image data are obtained. And displaying a set of water / fat separated images at substantially the same timing as when acquiring one piece of image data. 2. In the measurement for acquiring a plurality of image data, at least two image data among image data having different echo times are repeatedly acquired, and image data other than the two image data is intermittently acquired. The magnetic resonance imaging method according to claim 1, wherein the acquisition is performed at a time.