JP2009050738A - Formation method of mri data and imaging apparatus using it - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide the formation method of data (a half Fourier transform MRI) which attempts the shortening of imaging time, accelerates processing after imaging by improving the collecting measure of data and an image reconstruction method, and enables high-speed imaging also in a MRI apparatus in which the operating characteristic of a gradient magnetic field is bad, and an imaging apparatus using it. <P>SOLUTION: The two dimensional array (a line and a row) of data is carried out in k space in which the phase of a spin, which carries out magnetic resonance by adjusting the gradient magnetic field, is encoded in two directions. The formation method of data on the occasion is that the k space is divided into two parts with the central point as a border and the even-numbered line data and the odd-numbered line data are collected for one side and the other side, respectively. The other division data turn up and interpolate the one data with symmetry (point symmetry) in the center of the k space to give all the data in the k space. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

The present invention relates to an imaging apparatus based on the principle of nuclear magnetic resonance (NMR), and more particularly to a half Fourier nuclear magnetic resonance imaging method and an imaging apparatus.
This technology analyzes the internal structure and physical properties of various materials except metals and magnetic materials (nondestructive inspection), fluid visualization, chemical structure and pharmacological effects in pharmacy and chemistry, zoology, biology, medicine, It can be used for analysis of biological tissue structure and function in physiology and the like, biological reaction, measurement of disease inspection / diagnosis (non-invasive inspection / diagnosis), biological measurement during treatment, and the like. In terms of performance, the speed is increased as compared with the prior art, and the imaging time can be shortened (about 1/2).

  An MRI apparatus based on NMR (Magnetic Resonance Imaging) includes a magnet for generating a static magnetic field around a subject to be imaged and a gradient magnetic field generating coil capable of varying the magnetic field strength temporally and spatially. A static magnetic field generating magnet forms a spatially uniform magnetic field Bo, and at the same time, the gradient magnetic field generating coil spatially controls the strength of the magnetic field. The MRI apparatus obtains a signal (echo signal or FID (free induction decay) signal) by changing the intensity of the gradient magnetic field in the static magnetic field Bo to change the phase of the spin derived from the magnetic moment of the nuclear magnetic resonance.

These echo signals and FID (free induction decay) signals are referred to as MRI data, and the MRI data is hereinafter simply abbreviated as “data”.
When this signal is reconstructed by Fourier transform or the like, an image (MRI image) is obtained. In order to encode the spin phase and distribute it in the k-space, a gradient magnetic field is applied in the x, y, and z directions to spatially change the magnetic field strength.

  In the MRI apparatus, in addition to the conventional methods such as the SE (spin echo) method and the IR (inversion recovery) method, there is a demand for a high-speed technique for shortening the imaging time. Various methods have been proposed and patented as high-speed and ultrahigh-speed imaging methods in MRI. These include GRE (gradient echo) method, FE (field echo) method, EPI (echo planar imaging) method, SPI (spiral scan), single fast imaging (SPI) method. There are an excitation SE (Spin Echo) method and random walk magnetic resonance imaging (ultra-fast magnetic resonance imaging apparatus). When imaging 128 × 128 pixels, the imaging time of the two-dimensional image is several seconds to several tens of seconds (GRE method), 0.1 second to several seconds (EPI and SPI), 0.02 to several seconds (Oneshot-SE method), Random walk method).

These conventional methods take time for image processing. In order to solve this problem, an MR imaging method using half Fourier transform has been proposed (see, for example, Patent Document 1).
Japanese Patent Laid-Open No. 11-4818

The above-described conventional MR imaging method, imaging apparatus, and MR imaging method using half Fourier transform are required to have the following points due to their characteristics.
(1) Speeding up imaging:
Reducing the time required for imaging enables measurement and imaging of objects that change at high speed. In addition, when the measurement object is a human or animal, it contributes to the reduction of physical and mental burden.
(2) Ultra-high speed MRI system:
The MR imaging method, the data collection method in the imaging apparatus, and the image reconstruction method are improved and speeded up.
(3) Ultra-high-speed MRI using low-performance gradient magnetic field:
Usually, when performing MRI at ultra-high speed, it is indispensable to use a gradient magnetic field having excellent characteristics (characteristics of rising and falling magnetic fields).
It is an object of the present invention to shorten the imaging time, improve the MRI data collection method and image reconstruction method to speed up imaging, and enable MRI that enables high-speed imaging even with an MRI apparatus having poor gradient magnetic field operating characteristics. It is an object of the present invention to provide a data formation method and an imaging apparatus using the same.

(1) Image reconstruction method (data collection trajectory and image reconstruction using properties of Fourier transform)
Hermitian property is one of the properties of Fourier transform, which is one of orthogonal transforms. Data (echo signal or FID signal) that is the basis of an image to be reconstructed is observed in a frequency range of ± SW (spectrum distributed in positive and negative frequencies) in k space (frequency space). The observed data has a point-symmetric relationship with respect to the origin of k-space (SW = 0). For this reason, it is not necessary to collect all data (N points × N points) in k-space with ± SW as the frequency distribution, and ½ of the total data collection (N × N / 2), that is, ± SW On the other hand, it can be replaced by collecting either + SW or -SW and assigning the collected data so as to be the target of origin to the area not collected in the k space. Thereby, data is given in the entire k space. This property is Hermitian symmetry. In actual MRI imaging, an image (MRI image) can be reconstructed with only 1/2 data collection for the k-space region. The SNR (signal-to-noise ratio) of the reconstructed image is attenuated to 1 / √2, but the amount of data to be collected is ½, and the imaging time can be shortened to ½ in proportion to this. This feature has already been utilized as the current half-Fourier MRI.
The present invention provides a different image reconstruction method using the Hermitian symmetry. It is characterized in that it can be applied to an existing MRI image reconstruction method using Fourier transform, including a spiral scan method, which has been difficult in the prior art. In particular, this is the first case of introducing the concept of Hermitian subjectivity into the spiral scan method. Even with this method, the imaging time can be reduced to ½.

(2) Phase Correction Method In the proposed method, signals are collected every other row (odd or even number) in the k space, and the collected signals are folded and filled with data. Due to the non-uniformity of the static magnetic field and the non-linearity of the gradient magnetic field, there may be a difference in the spin phase between the folded and filled data. When this phase difference occurs, artifacts (shading irregularities and false images) occur in the reconstructed image. For this reason, when a phase difference occurs, an average value of the phases of the odd-numbered data and the even-numbered data is calculated and given as the data phase, or a process of matching with the phase of any data is performed. .

Specifically, the following means are adopted.
(1) The data formation method is a data formation method obtained by encoding a gradient field in a static magnetic field and encoding it in a two-dimensional direction, and the data is divided into two with the center of k-space as the boundary, One is encoded to collect even-numbered row data and the other is collected to collect odd-numbered row data, and the other divided data is generated by interpolating the one data symmetrically at the center and interpolating. (2) In the data formation method described in (1) above, the data is divided into two parts in positive and negative areas in the k space.
(3) The data formation method described in (1) or (2) is characterized in that the data collection method is applied to an MRI image reconstruction method using Fourier transform.

(4) A half Fourier nuclear magnetic resonance imaging apparatus includes an excitation / detection device that detects an echo signal or a free induction decay signal from an imaging target constituting an MRI image, data based on the echo signal or the free induction decay signal, and A half Fourier nuclear magnetic resonance imaging apparatus comprising: an image reconstruction device that forms a k-space in which data is two-dimensionally arranged, and Fourier transforms the k-space data to form an MRI image, wherein the image reconstruction device 4. The data formation according to claim 1, wherein when a gradient magnetic field is applied to encode spin phases in two directions (x, y) and two-dimensionally array in k-space. The method is characterized in that the method is applied by performing symmetrical interpolation at the center point.

(1) Reduction of imaging time (speeding up of MRI imaging):
An MRI image is reconstructed by collecting only half of the total data in k-space. Since the imaging time depends on the amount of data to be collected, according to the present invention, the imaging speed can be doubled (imaging time is ½). Since the data amount is halved compared to the conventional method, the SN ratio is reduced to 1 / √2 times. This method can be applied to all MRI imaging methods using Fourier transformation as the principle of image reconstruction. On the other hand, when data of a specific spatial frequency component is collected using the keyhole technique, the spatial frequency of the image is distributed concentrically around the origin in the k space. For this reason, when the keyhole technique is used for the purpose of speeding up, the half Fourier spiral MR imaging is useful.

(2) Correction of magnetic field inhomogeneity (corresponding to magnet having magnetic field inhomogeneity):
In the MRI apparatus, when the static magnetic field Bo fluctuates with time, or when the intensity of the magnetic field varies spatially without uniformity, the phase of the signal to be measured changes according to the fluctuation. On the other hand, the same applies when the spatial linearity and temporal stability of the gradient magnetic field are low. These changes appear in the reconstructed image as image distortion and deformation, and irregular shading changes. In the case of conventional half Fourier MR imaging that collects data in only a half region of k-space for speeding up, it is susceptible to these variations due to data collection in only a half region of k-space.
In order to achieve speedup using the Hermite objectivity of Fourier transform even in an MRI apparatus with poor magnetic field uniformity and temporal stability in this way, the entire k-space region can be realized using this method. By collecting data across the board, it is easy to correct changes caused by these. As a result, it is possible to reduce image distortion and irregular shading due to magnetic field inhomogeneity.

The present invention will be described with reference to the drawings.
FIG. 1 shows a block diagram of a nuclear magnetic resonance imaging apparatus of the present invention.
The magnetic resonance imaging apparatus of the present invention is based on NMR (nuclear magnetic resonance). Outside the excitation / detection coil 10, static magnetic field generating magnets 11a and 11b for generating a temporally and spatially constant magnetic field on both sides of the imaging target (or measurement substance), temporally and spatially Are provided with gradient magnetic field generating coils 12a and 12b for varying the magnetic field strength. The static magnetic field generating magnets 11a and 11b are connected to a static magnetic field generating magnet / shim coil power source 13 and a control device 14 for controlling the strength of the magnetic field. The gradient magnetic field generating coils 12a and 12b are connected to a gradient magnetic field generating coil power supply 16 and a controller 17 for controlling the strength of the magnetic field. Connected to the excitation / detection coil 10 is an image processing device including a receiver 19 that receives a signal corresponding to nuclear magnetic resonance, an image reconstruction device 20 that reconstructs an image, and an image display device 21.

Further, an RF (radio frequency) wave having a frequency equal to the resonance frequency is applied to the excitation / detection coil 10 in a pulsed manner in order to resonate the nuclear magnetism of the nucleus of the object (or measurement substance) to be imaged. A pulse generator 18 is connected. When the RF pulse of the RF pulse generator 18 is applied to the excitation / detection coil 10, the imaging target induces nuclear magnetic resonance, and a detection signal (FID [free induction decay] signal or Echo signal). This signal is collected by a receiver 19 connected to the excitation / detection coil 10. Thereafter, the image is reconstructed by the image reconstruction device 20.
FIG. 2 is an explanatory diagram showing the locus of data collection in k-space for two-dimensional imaging based on the principle of the present invention. The data collection trajectory shown in FIG. 2 is a two-dimensional SE (spin echo) method, a GRE (gradient echo) method, a FastSE (fast spin echo) method, and the like. Applicable to MRI technique to collect.

Since the object of imaging in the present invention is a two-dimensional coordinate system, the k-space is a space that gives a two-dimensional spatial frequency distribution of the image. The center of the k space has a spatial frequency of 0 and a signal intensity of ± ω on the left and right and top and bottom. That is, the signal distributed at the position of (+ n, + m) in the k space gives the frequency ω _{+ n, + m,} and the signal intensity at that frequency becomes the value (intensity) at this point.
In MRI, a signal (spin echo or FID) is observed using any one of imaging methods such as SE method, GRE method, EPI method (FIG. 7 described later) and spiral method (FIG. 11 described later). An example of the signal is shown in FIG. The horizontal axis represents time (generally observed at several tens of milliseconds), and the vertical axis represents signal intensity. Depending on the imaging method in MRI, signals are collected based on the trajectories of FIG. 3, FIG. 5, FIG. 8, FIG. 12, FIG.
For example, the signal of FIG. 3 is observed on one of a plurality of horizontal lines shown in FIG. In the case of reconstructing a 256 × 256 pixel two-dimensional image, since 256 signals are observed, 256 trajectories are taken. A two-dimensional MRI image is reconstructed by sampling 256 points for each observed signal and performing a two-dimensional Fourier transform on the obtained 256 × 256 data.

The method of the present invention is a method for reconstructing an image by halving the trajectory in the k space. Since the trajectory is halved, the imaging time is halved and high-speed imaging is possible. In addition, the method of the present invention can be applied to most of the proposed MRI imaging (all MRI imaging methods using Fourier transform for image reconstruction). As an example of image reconstruction, description will be made using an image reconstructing method (FIG. 4) for data (FIG. 3) collected based on the trajectory of FIG. 2 in k-space.
FIG. 4A shows a trajectory of data actually collected in the k space. When the MRI image to be reconstructed is N × N pixels (N = 128, 256, 512, etc.), N / 2 signals are collected. FIG. 4 (b) is a result of transposing FIG. 4 (a) with point symmetry, and is merely transposed in the same coordinate system. For example, the data point at the upper left in FIG. It is located in the lower right in FIG. Next, the process of superimposing FIG. 4A and FIG. 4B is performed. The result is shown in FIG. FIG. 4A includes N / 2 × N data, and FIG. 4B also includes N / 2 × N data. Since they do not overlap each other, (c) obtained by superimposing them is N / 2 × N + N / 2 × N, and therefore N × N data.

Since Fourier transformation is frequently used for image reconstruction methods in MRI, this processing can be applied to all of these image reconstruction methods. This transposition is made possible by the symmetry of the frequency distribution of the Fourier transform, the principle of which is known as a characteristic of the Fourier transform.
The method of the present invention is characterized by collecting data distributed over the entire area of k-space and halving the imaging time. Since data is collected in the entire area of k-space, the phase change of the observation signal caused by the non-uniformity can be corrected even with an MRI apparatus in which the intensity of the static magnetic field Bo created by the magnet is spatially non-uniform. . Also, it can be applied to the spiral scan method. On the other hand, since the data collection amount is half as a disadvantage, the S / N ratio of the reconstructed MRI image is theoretically reduced to 1 / √2 times. At this time, if there is a phase difference between the trajectory data of FIG. 4A and the trajectory data of FIG. 4B, the phase is corrected. The image is obtained by Fourier transforming the trajectory data shown in FIG. O represents the origin (the same applies to the following drawings).

In particular,
(1) By applying a gradient magnetic field, spin phases are encoded in two directions (x, y), and two-dimensionally arranged in k-space. At that time, the k-space is divided into two with the center as a boundary, one of which is even-numbered row array data and the other of which is odd-numbered row array data. The other divided data is characterized in that the one data is interpolated symmetrically (point symmetry) at the center point of k-space to obtain data in k-space.
(2) The data formation method according to (1) is characterized in that the data is divided into two divided into positive and negative phase regions in the k space.
(3) Applying the data collection method to all MRI image reconstruction methods (magnetic resonance imaging) using Fourier transform such as conventional SE method, GRE method, FastSE method, EPI method, spiral scan method, etc. Features.

(4) A half-Fourier nuclear magnetic resonance imaging apparatus is an excitation / detection device that detects an echo signal or a free induction attenuation signal from an imaging target that is an MR image signal, and data based on the echo signal or the free induction attenuation signal. A nuclear magnetic resonance imaging apparatus comprising an image reconstruction device that forms a two-dimensionally arranged k-space and Fourier transforms the k-space data to form an MRI image, wherein the image reconstruction device includes a gradient magnetic field When the data is encoded in the x and y directions to form a two-dimensional k-space, the data forming method according to any one of claims 1 to 3 is applied to perform symmetrical interpolation at the center point. It is characterized by creating.

FIG. 5 is an explanatory diagram showing a data collection trajectory for a two-dimensional EPI (echo planar) method based on the principle of the present invention. A data point sequence on the line is collected so that the region shown by the solid line in the figure is drawn along the line from the upper left (or from the lower right can be reversed). The conventional EPI method has taken a locus including a broken line region. According to the data collection method of the present invention, the collected data need only be a solid line portion (half area in k-space takes even-numbered row data, and the other half area takes odd-numbered trajectory). The amount of processing is ½ the processing time of all the conventional collections. As a result, the imaging time is reduced to ½.
FIG. 6 is an explanatory diagram showing a filling method for filling uncollected data in k-space (data necessary for image reconstruction) based on the principle of the present invention in the EPI method given in FIG. As in FIG. 4, the uncollected data points in the half of the image are filled by inverting the already collected data at the point-symmetrical positions with respect to the origin 0 as it is. Image (a) and image (b) are formed as described above. Thereafter, the filled image (a) and image (b) are synthesized to obtain an image (c).

  FIG. 7 is a waveform diagram showing a pulse sequence that enables formation of the collection trajectory of FIG. The horizontal axis of FIG. 7 is time TR, RF is an excitation signal by a radio frequency 90 ° RF pulse, Gs is a gradient magnetic field for slice selection, Gp is a gradient magnetic field for phase encoding of resonating spins, and Gr is for frequency encoding. Give a gradient magnetic field. Each is applied individually at the time shown. As a result, a signal is observed at the time indicated by the double-ended arrows in the figure. Since the observed signal exists in the k space, the spin phase changes depending on the strengths of the gradient magnetic fields Gp and Gr. By matching Gp and Gr with two orthogonal directions in the Euclidean space, a two-dimensional distribution of spins based on the two axes of Gp and Gr is obtained. The phase of the spin is proportional to the strength of the gradient magnetic fields (Gp and Gr). Therefore, by making the gradient magnetic field spatially linear (linear), the spin phase in the k-space shows a linear phase distribution. For this reason, an image (MRI image) is obtained by Fourier-transforming the observation signal collected in the k space.

FIG. 8 is an explanatory diagram showing a conventional spiral scan (SPI: Spiral MR Imaging) method in which a k-space is collected in a spiral shape. A trajectory is drawn spirally from the center of k-space to the outside or from the outside to the center, and points on the trajectory are collected as data.
FIG. 9 is an explanatory view showing a filling method for filling uncollected data in k-space (data necessary for image reconstruction) based on the principle of the present invention in the spiral scan method given in FIG.

Like FIG. 4 for FIG. 2 and FIG. 6 for FIG. 5, the data collected in FIG. 9A is transposed point-symmetrically as shown in FIG. 9B, and FIG. 9A and FIG. By combining the data of 9 (b), data is filled as shown in FIG. 9 (c).
FIG. 10 is an explanatory diagram showing the distribution of collected data in the k space, which is the processing result shown in FIG. 9C. The solid line is the locus of the collected data in FIG. 9A, the dotted line is the filled locus in FIG. 9B, and the locus obtained by combining the solid locus and the dotted locus in FIG. This is the final data collection trajectory. Data points on the line in the figure have been collected.

FIG. 11 is a waveform diagram showing a pulse sequence that enables the imaging method that takes a spiral locus shown in FIGS. As in FIG. 7, this is a time range indicated by an arrow by applying an RF wave for resonance (90 ° pulse) and a gradient magnetic field (Gs, G ₁ , G ₂ ) in the time series shown in the figure. A signal is observed at. Gs is a gradient magnetic field for determining the position of the slice cross section, and G ₁ and G ₂ are gradient magnetic fields applied to take the locus shown in FIG. 8 in the k space. The figure shows an imaging method that reconstructs an image by observing an FID (free induction decay) signal. After the 90 ° RF pulse in the figure, a 180 ° RF pulse is applied after half the TE (echo time). Echo signals can be observed, and diffusion-weighted imaging (displacement of magnetic resonance elements due to diffusion or reflux in materials) can be performed by applying MPG (Motion Probing Gradients) between 90 ° and 180 °. enable. Further, the figure shows an imaging method for drawing a locus from the center of the k space in FIG. 8 toward the outside. However, as described above, when the locus is drawn from the outside to the inside, the gradient magnetic field (G ₁ , G ₂ ) over time. In other words, the initial value is greatly increased and gradually attenuated in the change over time in the figure. The signal is observed during the same period as shown.

  FIG. 12 is an explanatory diagram showing an example of application of the imaging method of the present invention. FIG. 12 is a high-speed imaging that uses the keyhole method together with the present invention, and an object that changes over time can be imaged at an ultra-high speed. FIG. 12A shows a trajectory at the time of initial data collection, and collects data for the entire region using a conventional spiral scan method. FIG. 12B shows the data collection trajectory for the second and subsequent times, and collects only a specific frequency region. FIG. 12C shows a data collection trajectory for image reconstruction based on the present invention, and the data of FIG. 12B is transposed and filled with data based on the present invention. FIG. 12D shows a data collection trajectory to which the keyhole method is applied, and the data of the region (center part in the figure) that is missing in the result of FIG. 12C is filled with the collected data of FIG. To obtain the final data in k-space. Thereafter, an image (MRI image) is reconstructed by Fourier transform. An image equivalent to that in FIG. 12A is reconstructed only by collecting data in FIG. Only a specific region (FIG. 12B) that changes in the k-space is collected by the method shown in the present invention. Since the data collection time is proportional to the amount of data to be collected, the imaging time varies depending on the reduction amount of the collected data. Examples of applications include imaging of brain functions found in functional MRI (Functional MRI), movement of the heart and extremities.

FIG. 13 is an explanatory view showing another application example of the imaging method given in the present invention. In FIG. 13, the trajectory in the k space can be changed according to the frequency distribution in which the object changes, and when the object changes uniformly in the real space where the object exists, as shown in FIG. As shown in FIG. 13 (b), the collection trajectory can be changed as shown in FIG. 13 (c). Specifically, the gradient magnetic field application signals G ₁ and G ₂ shown in FIG. 11 are changed so as to take an arbitrary locus in FIGS. 13A to 13C in the k space.

FIG. 14 is an explanatory view showing an embodiment of the present invention. This figure is an image obtained by introducing the concept of the present invention to the EPI method, and reconstructing the data collected in FIG. 5 based on the pulse sequence in FIG. 7 and the data processing in FIG. FIG. 14A shows the result of reconstructing an image (two-dimensional image, 128 × 128 pixels, imaging time of about 100 msec) based on the conventional EPI method. On the other hand, FIG. 14B shows the result of the conventional half-Fourier EPI method (collecting only data of a little over half [about half +3] in the k space) for the purpose of shortening the imaging time. Is reduced to about ½. FIG. 14C is an image to which the half Fourier EPI method based on the present invention is applied. This method completes imaging in a half time shorter than the conventional method. In addition, since data of the entire area of k-space is collected, it is highly resistant to temporal and spatial inhomogeneities such as a static magnetic field (equivalent to the conventional EPI method (a) for collecting all data in k-space), and imaging time Is shortened.
(Difference from conventional technology)

FIG. 15 is an explanatory diagram for explaining an EPI pulse sequence in half Fourier imaging. In the conventional EPI method, after exciting spins using RF pulses, the gradient magnetic field is varied as shown in FIG. 15A with respect to the two-dimensional surface to be imaged. In the EPI method, there is half-Fourier merging using Hermite objectivity in Fourier transform, which can also be applied to the EPI method. The locus of data collection in k space at that time is shown in FIG. The purpose of the half Fourier technique in MRI is to increase the imaging speed. Data was collected only in an area of a little over half of the k-space, and the image was reconstructed by filling it with other areas. The problem with this method is that only half of the region is collected, so it is difficult to correct the phase of the collected data, and it can be pointed out that the method is vulnerable to magnetic field inhomogeneities. The present invention (c) has characteristics that overcome these problems.
FIG. 16 is an explanatory diagram showing a locus in the k space for the pulse sequence shown in FIG. In each of (a) full encoding method, (b) half encoding method, and (c) proposed encoding method, the trajectory for data collection is as shown in the figure.
FIG. 17 is an explanatory diagram showing a gradient magnetic field application method in the spiral scan method. In order to draw a spiral data collection trajectory in k-space, the gradient magnetic field takes the illustrated sine wave.

  This technology analyzes the internal structure and physical properties of various materials except metals and magnetic materials (nondestructive inspection), fluid visualization, chemical structure and pharmacological effects in pharmacy and chemistry, zoology, biology, medicine, It can be used for analysis of tissue structure and function of a living body in physiology and the like, biological reaction, measurement of disease inspection / diagnosis (non-invasive inspection), and the like.

1 shows a block configuration diagram of a nuclear magnetic resonance imaging apparatus of the present invention. FIG. It is explanatory drawing which shows the locus | trajectory of the data collection in k space with respect to the two-dimensional imaging based on the principle of invention. The waveform figure of the signal observed by MRI is shown. It is explanatory drawing which shows the filling method of the data with respect to the locus | trajectory of FIG. It is explanatory drawing which shows the collection locus | trajectory of the data for the two-dimensional EPI (echo planar) method based on the principle of this invention. FIG. 6 is an explanatory diagram showing a filling method for filling uncollected data (data necessary for image reconstruction) in k space based on the principle of the present invention in the EPI method given in FIG. 5. FIG. 6 is a waveform diagram showing a pulse sequence that enables formation of the collection trajectory of FIG. 5. It is explanatory drawing which shows the conventional spiral scan (SPI: Spiral MR Imaging) method which takes a collection locus | trajectory spirally in k space. FIG. 9 is an explanatory diagram showing a filling method for filling uncollected data (data necessary for image reconstruction) in k space based on the principle of the present invention in the spiral scan method given in FIG. 8. It is explanatory drawing which shows the collection locus | trajectory in k space given by FIG.9 (c). FIG. 11 is a waveform diagram showing a pulse sequence that enables the imaging method that takes a spiral locus shown in FIGS. 8 to 10. It is explanatory drawing which shows an example of the application of the imaging method of this invention. It is explanatory drawing which shows the other application example of the imaging method given by this invention. It is explanatory drawing which shows the Example of this invention. It is explanatory drawing explaining the pulse series with respect to an EPI method. It is explanatory drawing which shows the locus | trajectory in k space in an EPI method. It is explanatory drawing explaining the application method of the gradient magnetic field in a spiral scan method.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Excitation / detection coil 11a, 11b Static magnetic field generation magnet 12a, 12b Gradient magnetic field generation coil 13 Static magnetic field magnet shim coil power supply 14, 17 Control device 16 Gradient magnetic field generation coil power supply 18 RF pulse generator 19 Receiver 20 Image reconstruction apparatus 21 Image display device

Claims (4)

A method of forming MRI data composed of a two-dimensional matrix that obtains MRI data by varying a gradient magnetic field in a static magnetic field and encoding it in a two-dimensional direction in a static magnetic field. A method for forming MRI data, wherein the data is encoded so as to be collected in a spiral shape around the center. 2. The method of forming MRI data according to claim 1, wherein the spirally collected data is filled by being inverted to a point-symmetrical position that is not collected from the origin. 3. The MRI image reconstruction method according to claim 1, wherein the MRI data formation method is applied to an MRI image reconstruction method using Fourier transform. An excitation / detection device for detecting an echo signal or a free induction decay signal from an imaging target constituting an MRI image, and a k-space in which MRI data and the MRI data are two-dimensionally arranged based on the echo signal or the free induction decay signal A half Fourier nuclear magnetic resonance imaging apparatus comprising: an image reconstruction apparatus that forms and forms an MRI image by Fourier transforming the MRI data in the k space;
4. The image reconstruction device according to claim 1, wherein the image reconstruction device encodes a spin phase in two directions by applying a gradient magnetic field and two-dimensionally arranges MRI data in k space. 5. A half Fourier nuclear magnetic resonance imaging apparatus to which an MRI image reconstruction method is applied.