Magnetic resonance imaging apparatus and image processing method

Abstract

To perform super-resolution processing on an MRI image by using an NMR signal as a point spread function (PSF).SOLUTION: Image processing of performing processing of increasing the resolution by using a point spread function is performed on a reconstructed image. The point spread function is the signal obtained by acquiring a nuclear magnetic resonance signal from a phantom without applying frequency encoding and phase encoding after irradiating the phantom arranged in an imaging space with a high frequency magnetic field and performing Fourier transformation on the acquired nuclear magnetic resonance signal.SELECTED DRAWING: Figure 4

Description

[Technical field]

[0001] The present invention relates to a magnetic resonance imaging (hereinafter referred to as "MRI") device that measures nuclear magnetic resonance (hereinafter referred to as "NMR") signals from hydrogen, phosphorus, etc. in a subject and images nuclear density distribution, relaxation time distribution, etc., and that is equipped with a function for performing super-resolution processing. [Background technology]

[0002] An MRI device is a device that measures NMR signals generated by the nuclear spins that compose the tissues of a subject, particularly a human body, and creates two-dimensional or three-dimensional images of the morphology and function of the head, abdomen, limbs, etc. During imaging, the NMR signals are given different phase encodings by a gradient magnetic field and are frequency encoded, and are measured as time-series data. The measured NMR signals are reconstructed into an image by two-dimensional or three-dimensional Fourier transformation.

[0003] Since the images obtained by MRI systems are used to determine the presence or absence of disease and its condition, there has always been a demand for higher resolution and higher signal-to-noise ratio. To meet this demand, many technologies have been developed for MRI systems.

[0004] On the other hand, Patent Document 1 discloses a method of making an image degraded by transfer closer to the original image by using a function indicating transfer characteristics. For example, assuming that light points are imaged in the original image, the probability density distribution of the light imaged in the degraded image is expressed by a point spread function (PSF). It is disclosed that the original image can be calculated (estimated) from the degraded image and a previously obtained PSF by the Richardson-Lucy algorithm. The invention of Patent Document 1 also discloses that an optical lens is assumed as the transfer system, and an Optical Transfer Function (OTF) is used instead of a PSF.

[0005] Patent Document 2 also discloses that in a CT image taken by an X-ray computed tomography (CT) device, in order to further increase the resolution limit dependent on the pitch of the detector elements, a point spread function (PSF) is used for processing (super-resolution processing). Specifically, a PSF previously determined for the CT image is deconvoluted (deconvolution operation) to achieve a resolution that exceeds the resolution limit due to the pitch of the detector elements, etc. As the PSF, the distribution (blur) of image data taken by the X-ray CT device of a wire or spherical phantom that is smaller than the pitch of the detector elements of the X-ray CT device is used. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] International Publication No. 2006 / 041126 [Patent Document 2] JP 2005-95329 A Summary of the Invention [Problem to be solved by the invention]

[0007] In Patent Document 2, in the case of an X-ray CT apparatus, a wire or spherical phantom that is smaller than the pitch of the detection elements is imaged by the X-ray CT apparatus to obtain image data, and this image data is used as the PSF.

[0008] However, although Patent Document 2 describes that super-resolution processing using a PSF is also possible in an MRI apparatus, it does not disclose at all a specific method for calculating the PSF in an MRI apparatus.

[0009] When applying the technique of Patent Document 2 to an MRI device, a minute phantom is placed in a static magnetic field space, a high-frequency magnetic field pulse is irradiated to generate nuclear magnetic resonance in the spins of the atomic nuclei of the minute phantom, and a gradient magnetic field is applied in the x-axis direction and the y-axis direction to receive an NMR signal with frequency encoding in the x-axis direction and phase encoding in the y-axis direction. This operation is repeated 256 or 512 times while gradually changing the amount of phase encoding, and the obtained 256 or 512 NMR signals are placed in k-space (measurement space) and subjected to a two-dimensional Fourier transform to reconstruct a two-dimensional image of the minute phantom. The obtained image is used as the PSF. This method requires taking two-dimensional MRI images of the minute phantom over a long period of time.

[0010] An object of the present invention is to perform super-resolution processing of MRI images using NMR signals as point spread functions (PSFs). [Means for solving the problem]

[0011] In order to solve the above problems, the present invention comprises a static magnetic field generating device which applies a static magnetic field to an imaging space, a radio frequency magnetic field irradiating device which applies a radio frequency magnetic field to the spins of atomic nuclei of a subject placed in the imaging space, a gradient magnetic field generating device which applies gradient magnetic fields in two predetermined directions to impart frequency encoding and phase encoding to the spins of atomic nuclei, a receiving device which receives nuclear magnetic resonance signals generated by the spins of the atomic nuclei, an image reconstruction unit which performs a two-dimensional Fourier transform on a plurality of nuclear magnetic resonance signals placed in the measurement space to reconstruct an image of the subject, and an image processing unit which performs processing to enhance the resolution of the reconstructed image using a point spread function.

[0012] The point spread function is a signal obtained by Fourier transforming the nuclear magnetic resonance signal obtained by irradiating a phantom placed in the imaging space with the high-frequency magnetic field from the high-frequency magnetic field irradiation device, and then acquiring the nuclear magnetic resonance signal from the phantom by the receiving device without imparting the frequency encoding and phase encoding. Effect of the Invention

[0013] According to the present invention, it is possible to generate stable, high-resolution MRI images using NMR signals as PSFs. [Brief description of the drawings]

[0014] [Figure 1] 1 is a block diagram showing the overall configuration of an MRI apparatus according to a first embodiment of the present invention. [Diagram 2] FIG. 2 is a functional block diagram of a central processing unit 110 of the MRI apparatus according to the first embodiment. [Diagram 3] 4 is a flowchart showing the operation of the MRI apparatus according to the first embodiment. [Figure 4] 4 is a flowchart showing the operation of the MRI apparatus according to the first embodiment. [Diagram 5] 4 is a flowchart showing the operation of the MRI apparatus according to the first embodiment. [Figure 6] 4 is a flowchart showing the operation of the MRI apparatus according to the first embodiment. [Figure 7] FIG. 2 is a diagram for explaining how acquired NMR signals are arranged in k-space in the MRI apparatus of the first embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] An embodiment of the present invention will be described with reference to the drawings.

[0016] In this embodiment, a process for enhancing the resolution of an image reconstructed by an MRI apparatus is performed using a point spread function (PSF).The point spread function (PSF) is obtained by irradiating a phantom placed in an imaging space with a high-frequency magnetic field, acquiring a nuclear magnetic resonance signal from the phantom without applying frequency encoding and phase encoding, and performing a Fourier transform on the acquired nuclear magnetic resonance signal.

[0017] This allows the MRI device to obtain the PSF directly from the NMR signal and perform super-resolution processing.

[0018] When acquiring the NMR signal, since position encoding (frequency encoding and phase encoding) is not performed, if it is an ideal device without reflections or the like in the signal transmission system of the NMR signal of the MRI apparatus, the obtained NMR signal will be a signal corresponding to an optically ideal point image. Therefore, in an actual MRI apparatus, the distortion included in the NMR signal without position encoding (frequency encoding and phase encoding) is caused by reflections or the like in the signal transmission system of the NMR signal. Thus, by using the NMR signal without position encoding as the PSF, a more accurate PSF can be obtained than using a phantom image reconstructed by two-dimensional Fourier transform.

[0019] By performing super-resolution processing on the MRI image using this PSF, a high-resolution MRI image with high accuracy can be obtained.

[0020] Also, when acquiring the PSF, since it is not necessary to reconstruct the two-dimensional image of the phantom by two-dimensional Fourier transform, the PSF can be acquired in a short time.

[0021] Hereinafter, the MRI apparatus of this embodiment will be specifically described.

[0022] <<<First Embodiment>>> <Configuration of MRI Apparatus> First, the configuration of the MRI apparatus 100 of the first embodiment will be described with reference to FIG. 1.

[0023] The MRI apparatus 100 includes a static magnetic field generator 130, a gradient magnetic field generator 132 that applies a gradient magnetic field pulse to the subject 10, a high-frequency magnetic field irradiation device 140 that irradiates the subject 10 with a high-frequency magnetic field pulse (hereinafter referred to as an RF pulse), a receiver 150 that receives an echo signal, which is an NMR signal, from the subject 10, a processing device 160 including a central processing unit (hereinafter referred to as a CPU) 110, a sequencer 120, an operation device 170 for performing various operations related to data input, imaging, etc., and a bed 30.

[0024] The static magnetic field generating device 130 applies a uniform magnetic field to the imaging space 20 .

[0025] The bed carries the subject 10, and at least the part of the subject 10 to be imaged is placed in the imaging space 20.

[0026] The gradient magnetic field generating device 132, the high frequency magnetic field irradiation device 140, and the receiving device 150 constitute an imaging section 300. The imaging section 300 applies an RF pulse and a gradient magnetic field pulse to a subject 10 placed in an imaging space 20, and acquires an NMR signal generated from the subject.

[0027] The central processing unit 110, as shown in the functional block diagram of FIG. 2, functions as a control unit 200 and controls the sequencer 120 so as to cause the imaging unit 300 to repeatedly apply radio frequency magnetic field pulses and gradient magnetic field pulses and acquire nuclear magnetic resonance signals at predetermined timings in accordance with a predetermined imaging pulse sequence.

[0028] The central processing unit 110 also functions as an image reconstruction unit 210, which processes the number of NMR signals acquired by the imaging unit 300 required for image reconstruction, and generates a tomographic image (MRI image) of the subject 10.

[0029] Furthermore, the central processing unit 110 also functions as an image processing unit 220, and performs processing to increase the resolution of the reconstructed MRI image of the subject 10 using a point spread function. For this purpose, the image processing unit 220 includes a PSF acquisition unit 221 that acquires the point spread function, a PSF storage unit 222 that stores the acquired point spread function, and a super-resolution processing unit 223 that processes the MRI image using the PSF. The PSF storage unit 222 may be a memory built into the central processing unit 110, or may be realized by the RAM 168, the optical disk 162, or the like.

[0030] Further, each part will be described in detail. Although the description of the specific configuration of the static magnetic field generating device 130 is omitted, the static magnetic field generating device 130 generates an extremely uniform static magnetic field in the space around the subject 10 in a direction perpendicular to the body axis in the case of a vertical magnetic field type, and in the direction of the body axis in the case of a horizontal magnetic field type. In order to generate the static magnetic field, the static magnetic field generating device 130 has a permanent magnet, a resistive magnet, or a superconducting magnet arranged around the subject 10 as a static magnetic field generating source.

[0031] The radio frequency magnetic field irradiation device 140 excites the nuclear spins of atoms constituting the biological tissue of the subject 10 by irradiating the subject 10 with an RF pulse, thereby causing nuclear magnetic resonance. For example, the radio frequency magnetic field irradiation device 140 includes a radio frequency oscillator 142, a modulator 144, a radio frequency amplifier 146, and a radio frequency coil 148 on the transmitting side that operates as a transmitting coil. The radio frequency pulse output from the radio frequency oscillator 142 is amplitude-modulated by the modulator 144 at a timing instructed by the sequencer 120, and the amplitude-modulated radio frequency pulse is amplified by the radio frequency amplifier 146. The amplified radio frequency pulse is supplied to the radio frequency coil 148 disposed close to the subject 10. As a result, the radio frequency coil 148 irradiates the subject 10 with the RF pulse.

[0032] The gradient magnetic field generating device 132 has a gradient magnetic field coil 134 wound in three axial directions of the X-axis, Y-axis, and Z-axis, which are a coordinate system of the MRI device 100, for example, a stationary coordinate system, and a gradient magnetic field power supply 136 that supplies a driving current for generating a gradient magnetic field to each gradient magnetic field coil. The gradient magnetic field power supply 136 operates according to an instruction from the sequencer 120 and supplies a driving current to the gradient magnetic field coil 134 in the three axial directions of the X-axis, Y-axis, and Z-axis. As a result, the gradient magnetic field coil 134 generates gradient magnetic fields Gx, Gy, and Gz in the three axial directions of the X-axis, Y-axis, and Z-axis. The generated gradient magnetic fields Gx, Gy, and Gz are applied to the subject 10 in the imaging space 20. During imaging, a gradient magnetic field pulse (Gz) for determining a slice plane in a direction perpendicular to the imaging cross section (slice plane) is applied according to the imaging pulse sequence. A gradient magnetic field pulse (Gx) that imparts frequency encoding to the nuclear spins and a gradient magnetic field pulse (Gy) that imparts phase encoding to the nuclear spins are applied in the remaining two directions, X and Y, which are perpendicular to the slice plane and perpendicular to each other. This causes position information in each direction to be encoded in the echo signal (NMR signal) emitted from the nuclear spins.

[0033] The receiving device 150 receives and processes the emitted NMR signal. The receiving device 150 has a receiving side radio frequency coil 152 that operates as a receiving coil, a signal amplifier 154 that amplifies the received NMR signal, a quadrature phase detector 156, and an A / D converter 158 that converts an analog signal into a digital signal. The radio frequency coil 152 detects the NMR signal emitted by the subject, and the signal amplifier 154 amplifies the detected NMR signal. The quadrature phase detector 156 splits the NMR signal into two orthogonal signals at a timing instructed by the sequencer 120. The split two signals are each converted into a digital quantity by the A / D converter 158 and sent to the central processing unit 110.

[0034] The transmitting side radio frequency coil 148 and the gradient magnetic field coil 134 are disposed in the static magnetic field space of the static magnetic field generating device 130 in which the subject 10 is disposed. In terms of their orientation, they are disposed facing the subject 10 in the case of a vertical magnetic field system, and are installed so as to surround the subject 10 in the case of a horizontal magnetic field system. The receiving side radio frequency coil 152 is disposed so as to face or surround the subject 10.

[0035] The sequencer 120, under the control of the central processing unit 110, outputs control signals (commands) at predetermined timing to the high frequency magnetic field irradiation device 140, the gradient magnetic field generator 132, and the receiver 150. This causes RF pulses and gradient magnetic field pulses to be applied to the subject at timing according to a predetermined imaging pulse sequence, and causes the imaging section 300 to repeatedly acquire NMR signals while changing the amount of phase encoding.

[0036] In addition to the central processing unit 110, the processing device 160 has an external storage device such as an optical disk 162 or a magnetic disk 164 for storing information, a ROM 166 for storing necessary data and programs for processing, a RAM 168 for performing temporary storage for processing, and a display 169 such as a CRT. With this, the processing device 160 performs various data processing and displays and stores the processing results. For example, when the processing results received and processed by the receiving device 150 are input from the receiving device 150 to the processing device 160, the central processing device 110 of the processing device 160 performs signal processing, image reconstruction, image processing, and other processing. The resulting tomographic image of the subject 10 is displayed on the display 169 and is recorded on the optical disk 162, the magnetic disk 164, or the like of the external storage device as necessary. Although not shown, it is also possible to print or transmit it to another system.

[0037] The operation device 170 includes a pointing device 174 such as a trackball or a mouse, and a keyboard 176. The operator inputs various control information of the MRI device 100 and control information of the processing performed by the processing device 160 via the operation device 170. The operation device 170 is arranged close to the display 169, and the operator can interactively operate through the operation device 170 while viewing the display on the display 169. Note that the operation device 170 is not limited to this, and may include, for example, a touch panel provided on the display surface of the display 169. The operation device 170 is provided in an operation room separated from the main body of the MRI device 100. A part of the operation device 170 is further provided on the main body of the MRI device 100 and the bed 30. Thus, the operator can perform necessary operations near the subject 10.

[0038] The nuclide to be imaged of the subject 10 in the MRI device 100 is, for example, the hydrogen atomic nucleus, that is, the proton, which is the main constituent substance of the subject, as is widespread in clinical practice. The MRI device 100 images information regarding the spatial distribution of the proton density and the spatial distribution of the relaxation time of the excited state. Thereby, the MRI device 100 can two-dimensionally or three-dimensionally image and reconstruct the form or function of the subject 10, such as the head, abdomen, or limbs, and display the reconstructed image on the display 169. The reconstructed image is stored in the optical disk 162 or the magnetic disk 164 as needed. Also, the image is printed based on an operation or transmitted to other necessary systems.

[0039] <Imaging operation of the MRI device> The imaging pulse sequence is stored in advance for each of a plurality of imaging methods in a memory built in the central processing unit 110, the magnetic disk 164, the ROM 166, etc. The operator selects an imaging method via the operation device 170 and further inputs imaging conditions (repetition time TR, etc.). Thereby, the central processing unit 110 generates an imaging pulse sequence according to the set imaging conditions.

[0040] Here, the central processing unit 110 is a processor (computer) such as a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit), and by reading and executing a program previously stored in the ROM 166, the functions of each part in FIG. 2 are realized by software. In this way, the processes shown in the flows of FIGS. 3 to 5 are executed. It is also possible to realize some or all of the functions in FIG. 2 by hardware. For example, some or all of the central processing unit 110 may be configured using a custom IC such as an ASIC (Application Specific Integrated Circuit) or a programmable IC such as an FPGA (Field-Programmable Gate Array), and the programmable IC may be designed as a circuit to perform the processes in FIGS. 3 and 4.

[0041] The imaging operation of the MRI apparatus of this embodiment will be described in detail below with reference to the flows of FIG. 3 and FIG.

[0042] (Step 301) First, the PSF acquisition section 221 of the image processing section 220 executes preliminary shooting to acquire a PSF, and stores the acquired PSF in the PSF storage section 222. Step 301 will be explained in more detail below with respect to steps 401 to 405.

[0043] ((Step 401)) Specifically, first, the operator places the subject 10 in the imaging space 20 .

[0044] ((Step 402)) The PSF acquisition unit 221 controls the sequencer 120 to irradiate the phantom with a radio frequency magnetic field pulse from the radio frequency magnetic field irradiation device 140, thereby exciting nuclear magnetic spins in the phantom.

[0045] ((Step 403)) Without applying a gradient magnetic field for frequency encoding and a gradient magnetic field for phase encoding, the PSF acquisition section 221 causes the receiving device 150 to receive an NMR signal emitted by the nuclear magnetic spins of the subject at a predetermined timing.

[0046] Since no frequency encoding is performed, if the MRI device is an ideal device with no reflections in the NMR signal transmission system, the NMR signal does not spread in the frequency direction. Therefore, the NMR signal is a single-frequency signal or a DC component, as shown in FIG. 6(b-2). This does not depend on the size of the subject 10, and whether the subject 10 is a tiny object (close to a Δ function) like phantom A in FIG. 6(b-1) or a non-tiny object like phantom B, the obtained NMR signal will be a single-frequency signal or a DC component as long as the MRI device is an ideal device.

[0047] In reality, in an MRI device, there are reflections, loops, phase shifts, etc. in the NMR signal transmission system, so the NMR signal in FIG. 6(b-2) has distortion components and is not a completely single frequency or DC component.

[0048] ((Step 404)) The PSF acquisition unit 221 performs a Fourier transform on the NMR signal of FIG. 6(b-2) obtained without frequency encoding or phase encoding.

[0049] Because the NMR signal in Fig. 6(b-2) has distortion, the NMR signal after Fourier transform is not a perfect single peak (Δ function) but has a waveform with a spread, as shown in Fig. 6(b-3). This spread represents reflections, wraparound, phase shifts, etc. of the signal transmission system of the MRI device's NMR signal.

[0050] ((Step 405)) The PSF acquisition unit 221 stores the NMR signal of FIG. 6(b-3) in the PSF storage unit 222 as a PSF.

[0051] This makes it possible to obtain a highly accurate PSF for the MRI system. The PSF for the MRI system differs from the PSF for other optical systems in that the PSF is not one-dimensional data of only the intensity signal, but is complex data of real and imaginary parts (or intensity and phase).

[0052] (Step 302) Next, the control unit 200 generates an imaging sequence of the imaging conditions received from the operator, sets the generated imaging sequence in the sequencer 120, and causes the imaging unit 300 to perform actual imaging of the subject .

[0053] Specifically, the control unit 200 excites nuclear magnetic spins by irradiating a radio frequency magnetic field pulse from the radio frequency magnetic field irradiation device 140 to the subject 10 placed in the imaging space 20 as shown in FIG. 6(a-1) in accordance with an imaging sequence, phase encodes the subject 10 by applying a Y-direction gradient magnetic field pulse Gy from the gradient magnetic field generator 132, and receives the NMR signal at the receiving device 150 while applying an X-direction gradient magnetic field pulse Gx.

[0054] As a result, when the subject 10 is placed as shown in Fig. 6(a-1), an NMR signal having a waveform like a sinc function is received as shown in Fig. 6(a-2). The receiver 150 acquires one line of NMR signals per AD conversion. In reality, one line of NMR signals is QD detected, so complex data of real and imaginary parts is obtained.

[0055] The control unit 200 repeats the imaging sequence while changing the amount of phase encoding, and the receiving device 150 sequentially receives 256 or 512 NMR signals with different amounts of phase encoding.

[0056] (Step 302) The image reconstruction unit 210 arranges the obtained NMR signals in a measurement space (k-space) with a frequency encoding direction kx and a phase encoding direction ky, as shown in Fig. 7. That is, one-dimensional NMR data with different phase encoding is stacked to fill the k-space in the two-dimensional space of the frequency encoding direction and the phase encoding direction.

[0057] The image reconstruction unit 210 reconstructs a two-dimensional MRI image (xy real space) by performing a two-dimensional Fourier transform on the NMR signal in the k-space. This makes it possible to obtain an MRI image close to the actual object (subject 10) as shown in FIG. 6(a-3).

[0058] The image reconstruction unit 210 stores the obtained MRI image obtained during the actual imaging in the RAM 168 or the like.

[0059] (Step 304) The super-resolution processing unit 223 of the image processing unit 220 performs super-resolution processing on the MRI image obtained in step 303 .

[0060] The detailed process of step 304 will be described with reference to steps 501 to 506 in the flow of FIG.

[0061] ((Step 501)) The super-resolution processor 223 reads out the PSF stored in the PSF storage unit 222 in step 405 above, and sets it as the initial value of the PSF.

[0062] ((Step 502)) The super-resolution processor 223 reads out the actual MRI image (hereinafter also referred to as the degraded image) stored in the RAM 168 or the like.

[0063] ((Step 503)) The super-resolution processor 223 performs a deconvolution operation on the read PSF and the actual MRI image, thereby increasing the resolution of the degraded image (MRI image).

[0064] ((Step 504)) The super-resolution processing unit 223 obtains the difference between the degraded image (MRI image) and the image with high resolution obtained in step 503 by calculation.

[0065] ((Step 505)) If the difference calculated in step 504 is equal to or greater than a predetermined threshold value, the super-resolution processor 223 proceeds to step 506 .

[0066] ((Step 506)) The super-resolution processor 223 corrects the PSF (the initial value of the PSF in step 501 in the first step) based on the difference calculated in step 504, and returns to step 503 to perform deconvolution calculation of the PSF and the degraded image again.

[0067] Steps 503 to 506 are repeated until the difference between the degraded image (MRI image) and the image with high resolution obtained in step 503 becomes smaller than the threshold value. This improves the accuracy of the PSF with each repetition, making it possible to obtain an image with higher resolution.

[0068] (Step 305) The super-resolution processor 223 displays on the display 169 a high-resolution image obtained by repeating steps 503 to 506 .

[0069] As described above, according to the first embodiment, a nuclear magnetic resonance signal is acquired from a subject without applying frequency encoding and phase encoding, and a signal obtained by Fourier transforming the acquired nuclear magnetic resonance signal is used as a PSF, so that a PSF that well reflects the characteristics of the MRI apparatus can be obtained in a short time.

[0070] Therefore, by using this PSF to perform super-resolution processing on an MRI image, a more accurate high-resolution MRI image can be obtained.

[0071] <<<Second embodiment>>> In the first embodiment, the subject 10 on which the actual imaging is performed is also used to acquire the PSF in step 301. However, considering the size of the subject 10, the PSF has a width in the frequency direction due to the influence of the static magnetic field uniformity. Therefore, in the second embodiment, in order to avoid the influence of the non-uniformity of the static magnetic field strength, a sufficiently small phantom, which will be described later, is used to execute step 301 (steps 401 to 405) of the first embodiment to acquire the PSF. The device configuration of the MRI device of the second embodiment and the operation of each unit are the same as those of the first embodiment, except that a sufficiently small phantom is used, and therefore description thereof will be omitted.

[0072] The PSF signal of the second embodiment will be further explained. In the configuration of this embodiment, in which an NMR signal is acquired without adding position encoding (frequency encoding and phase encoding) and used as a PSF, if the magnetic field strength at the position where the object is placed is not completely uniform, the frequency of the NMR signal obtained in FIG. 6(b-2) has a width that depends on the magnetic field inhomogeneity. Since the magnetic field homogeneity differs depending on the position, it is not appropriate to apply the obtained PSF to the entire image.

[0073] Therefore, if an appropriate phantom that is large enough to obtain a signal and small enough that magnetic field inhomogeneity can be ignored is placed in the imaging space, and then the NMR signal is obtained without position encoding and used as the PSF, an appropriate PSF that is not affected by magnetic field inhomogeneity and reflects the signal transmission system of the MRI device can be obtained.

[0074] In addition, as the phantom becomes smaller, the signal strength decreases in proportion to its volume, while the signal-to-noise ratio increases in proportion to the static magnetic field strength. Therefore, it is not possible to generalize what a sufficiently small phantom is, but for example, using a phantom of about 3 cm square in a 1.5T device is sufficient to obtain NMR signals for PSF. Also, a phantom of about 3 cm square can fully avoid the effects of magnetic field inhomogeneity.

[0075] <<<Third embodiment>>> In the second embodiment, the influence of static magnetic field inhomogeneity was avoided by using a small phantom, whereas in the third embodiment, when a radio frequency magnetic field pulse is applied to the subject 10 or the phantom in step 402, the influence of static magnetic field inhomogeneity is similarly avoided and the PSF is obtained by selectively exciting only a very narrow range of nuclear spins.

[0076] Specifically, methods for selectively exciting atomic nuclear spins include using a presaturation pulse to suppress signals from non-selected areas, and using a two-dimensional selective RF pulse to excite only the desired areas.

[0077] In the first to third embodiments, the case where the NMR signal is one-dimensional data has been described, but the present invention can also be applied to two-dimensional and three-dimensional MRI images.

[0078] In addition, the algorithm for performing super-resolution processing on an image using the PSF is not limited to the above-mentioned algorithm, and various known algorithms can be used. In other words, there is no particular restriction on how to use the PSF, and high-resolution processing can be performed using a more accurate method. [Explanation of symbols]

[0079] 10 Subject 20 Imaging Space 30 berths 100 MRI machine 110 Central Processing Unit 120 Sequencer 130 Static Magnetic Field Generator 132 Gradient Magnetic Field Generator 134 Gradient Magnetic Field Coil 136 Gradient magnetic field power supply 140 High-frequency magnetic field irradiation device 142 High Frequency Oscillator 144 Modulator 146 High Frequency Amplifier 148 High Frequency Coil 150 Receiving device 152 High Frequency Coil 154 Signal Amplifier 156 Quadrature Detector 158 Converter 160 Processing Equipment 162 Optical Discs 164 Magnetic Disk 169 Display 170 Operating device 174 Pointing Device 176 Keyboard 200 Control section 210 Image reconstruction unit 220 Image Processing Unit 221 PSF acquisition department 222 PSF storage section 223 Super-resolution processing unit 300 Filming Department

Claims

1. the imaging apparatus includes a static magnetic field generating device which applies a static magnetic field to an imaging space, a high frequency magnetic field irradiating device which irradiates a high frequency magnetic field to the spins of atomic nuclei of a subject placed in the imaging space, a gradient magnetic field generating device which applies gradient magnetic fields in two predetermined directions in order to impart frequency encoding and phase encoding to the spins of the atomic nuclei, a receiving device which receives nuclear magnetic resonance signals generated by the spins of the atomic nuclei, an image reconstructing unit which performs a two-dimensional Fourier transform on a plurality of the nuclear magnetic resonance signals placed in a measurement space to reconstruct an image of the subject, and an image processing unit which performs processing to enhance the resolution of the reconstructed image by using a point spread function; the image processing unit includes an acquisition unit that acquires the point spread function, and a storage unit that stores the acquired point spread function, the acquisition unit, after irradiating the high-frequency magnetic field from the high-frequency magnetic field irradiation device to a phantom or subject placed in the imaging space, causes the receiving device to acquire the nuclear magnetic resonance signal from the phantom or subject without imparting the frequency encoding and phase encoding, performs a Fourier transform on the acquired nuclear magnetic resonance signal, and stores the obtained signal in the memory unit as the point spread function.

2. 2. The magnetic resonance imaging apparatus according to claim 1, wherein the image processing unit generates an image with improved resolution by performing a deconvolution operation on the image of the subject and the point spread function.

3. 2. The magnetic resonance imaging apparatus according to claim 1, wherein the phantom is equal to or greater than 1 cc and less than 30 cc.

4. 2. The magnetic resonance imaging apparatus according to claim 1, wherein the acquisition section selectively excites only atomic nuclear spins in a partial range of the phantom or the subject by the high frequency magnetic field.

5. 1. An image processing method for enhancing resolution of an image captured by a magnetic resonance imaging apparatus by using a point spread function that has been calculated in advance, comprising the steps of: an image processing method, characterized in that the point spread function is a signal obtained by Fourier transforming a nuclear magnetic resonance signal from a phantom or a subject placed in an imaging space of the magnetic resonance imaging apparatus after irradiating the phantom or the subject with a radio frequency magnetic field and then acquiring the nuclear magnetic resonance signal from the phantom or the subject without applying frequency encoding and phase encoding.

Images