Data processing device and magnetic resonance imaging device

Abstract

To improve accuracy of analysis using spectral information by precisely identifying a region including a metabolite corresponding to a chemical shift of interest, in an MRS technique.SOLUTION: A data processing device according to an embodiment comprises: an acquisition unit which acquires data for detecting a chemical shift of a substance, the data being based on a magnetic resonance signal collected from a specific region of an analyte; and a calculation unit which calculates a relation between a designated chemical shift designated arbitrarily in a predetermined range of the chemical shift and a displacement amount of a position of the specific region displaced as a result of the designated chemical shift, as position correction data.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The embodiments disclosed in this specification and the drawings relate to a data processing device and a magnetic resonance imaging device. [Background technology]

[0002] When a subject is placed in a strong magnetic field, the nuclear spins of the molecules that make up the subject are aligned, and the nuclear spins precess at the Larmor frequency. When a high-frequency signal at the Larmor frequency is irradiated onto the subject in this state, the nuclear spins of the molecules resonate with the high-frequency signal and are excited. After that, when the molecules return to their original stable state, a magnetic resonance signal (MR (Magnetic Resonance) signal) is generated from the subject. This phenomenon is called magnetic resonance.

[0003] Known techniques that utilize the magnetic resonance phenomenon include magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS).

[0004] MRI is a technology that generates images of a subject using MR signals generated by the subject. In MRI, the frequency and phase of the MR signals are slightly changed for each spatial position in the subject, and images of the subject are generated by correlating positions inside the subject with the intensity of the MR signals.

[0005] On the other hand, MRS is a technology that uses MR signals generated from a subject to detect, analyze, or display the spectrum of a substance in a specific region of the subject. It is known that metabolic substances contained in the subject, such as choline, creatine, and N-acetylaspartic acid (NAA), have slightly different magnetic resonance frequencies depending on the type of substance due to differences in the chemical bonds of various molecules. The deviation of the magnetic resonance frequency of each substance from the magnetic resonance frequency of a specific reference substance (reference magnetic resonance frequency) is called the chemical shift.

[0006] MRS detects the spectrum in a specific region of a subject and correlates it with the chemical shift, thereby enabling estimation of the type and amount of metabolites in that specific region. For example, by setting the lesion site of the subject as a specific region and examining the distribution of peak values ​​of individual metabolites contained in the spectrum in this specific region, useful diagnostic information can be obtained.

[0007] On the other hand, the position of the region where MR signals for detecting the spectrum are collected (i.e., the region of interest (ROI)) depends in principle on the chemical shift. Usually, the region of interest is set by displaying a positioning image on a display and setting the region of interest on this positioning image. However, as described above, the position of the region of interest depends on the chemical shift. In other words, the position of the region of interest corresponding to these will differ depending on the value of the chemical shift or the type of metabolite corresponding to the chemical shift.

[0008] For this reason, it has been difficult to accurately identify the region where the metabolite corresponding to the chemical shift of interest exists. Furthermore, the inability to accurately identify the location has limited the improvement of the accuracy of various analyses and disease diagnosis using spectral information. [Prior art documents] [Patent documents]

[0009] [Patent Document 1] Japanese Patent Application Laid-Open No. 2009-279432 [Non-patent literature]

[0010] [Non-Patent Document 1] Takeshima H. ​​Deep Learning and Its Application to Function Approximation for MR in Medicine: An Overview. Magn Reson Med Sci doi: 10.2463 / mrms.rev.2121-0040 Summary of the Invention [Problem to be solved by the invention]

[0011] One of the problems to be solved by the embodiments disclosed in this specification and the drawings is to enable accurate identification of a region where a metabolite corresponding to a chemical shift of interest exists in MRS technology, thereby improving the accuracy of analysis using spectral information. However, the problems to be solved by the embodiments disclosed in this specification and the drawings are not limited to the above problem. Problems corresponding to the effects of each configuration shown in each embodiment described below can also be positioned as other problems. [Means for solving the problem]

[0012] In one embodiment, a data processing device includes an acquisition unit that acquires data based on magnetic resonance signals collected from a specific region of a subject, the data being for detecting a chemical shift of a substance, and a calculation unit that calculates, as position correction data, the relationship between a designated chemical shift arbitrarily designated within a predetermined range of the chemical shift and the amount of displacement of the position of the specific region that displaces due to the designated chemical shift. [Brief explanation of the drawings]

[0013] [Figure 1] 1 is a configuration diagram showing an example of the overall configuration of a magnetic resonance imaging apparatus according to an embodiment; [Figure 2] FIG. 1 is a block diagram showing an example of the configuration of an MRI / MRS processing device and a data processing device. [Figure 3] 4 is a flowchart showing an example of the operation of the data processing device according to the first embodiment. [Figure 4]1 is a sequence diagram showing an example of an MRS pulse sequence. [Figure 5] (a) is a diagram showing an example of an image in which a rectangular frame indicating the region of the reference ROI is superimposed on a positioning image, and (b) is a diagram showing an example of a spectral image generated by MRS. [Figure 6] A diagram explaining the difference between the reference ROI and the effective ROI in one dimension in the X direction. [Figure 7] 10 is a diagram illustrating the relationship between a two-dimensional reference ROI and an effective ROI in the X and Y directions; [Figure 8] FIG. 1 shows an example of an MR image and a spectral image. [Figure 9] 10 is a flowchart showing an example of the operation of a data processing device according to a second embodiment. [Figure 10] FIG. 1 shows an example of a pulse sequence for multivoxel MRSI. [Figure 11] FIG. 1 shows an example of a conventional display of MRSI data generated by reconstruction. [Figure 12] An example of a display of an MR image and a spectral image with a voxel frame overlaid. [Figure 13] A diagram showing a display method for displaying MRSI data as an intensity map (chemical shift image). [Figure 14] FIG. 10 is a diagram showing a display method according to a modified example of the second embodiment. [Figure 15] FIG. 1A is a diagram showing the position of a reference ROI in an MRS when OVS is not applied, and FIG. 1B is a diagram showing an example of a reference ROI and an OVS region in an MRS when OVS is applied. [Figure 16] 10A and 10B are diagrams illustrating examples of a display image and a display image of a spectral image according to the third embodiment. [Figure 17] 13A to 13C are diagrams illustrating examples of display images of an intensity map and an effective OVS region in the third embodiment. [Figure 18] 1A to 1C are diagrams illustrating different shapes of OVS regions. [Figure 19] FIG. 10 is a block diagram showing an example of the arrangement of an MRI / MRS processing apparatus and a data processing apparatus according to a fourth embodiment. [Figure 20] 10 is a flowchart showing an example of the operation of the MRI / MRS processing device and data processing device according to the fourth embodiment. [Figure 21] FIG. 10 is a diagram showing an example of an operation for designating a chemical shift of interest and extracting spectral data in a predetermined range centered on each chemical shift of interest. [Figure 22] FIG. 10 is a diagram showing a partial MR image in which a region of interest in a partial MR image is specified within an MR image and the center position is corrected using position correction data based on a specified chemical shift. [Figure 23] FIG. 10 is a diagram showing a processing concept for extracting a position-corrected partial MR image region from an MR image to generate a partial MR image. [Figure 24] FIG. 10 is a diagram showing an example of an operation in which extracted partial spectral data and partial MR image data are input into a trained model and analysis results are obtained from the trained model. DETAILED DESCRIPTION OF THE INVENTION

[0014] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A data processing device and a magnetic resonance imaging device according to embodiments of the present invention will now be described with reference to the accompanying drawings.

[0015] (Magnetic Resonance Imaging Device) 1 is a block diagram showing the overall configuration of a magnetic resonance imaging apparatus 1 according to an embodiment. The magnetic resonance imaging apparatus 1 of the embodiment includes a magnet gantry 100, a control cabinet 300, an MRI / MRS processing device 400, a bed 500, and a data processing device 600.

[0016] The magnetic gantry 100 has a static magnetic field magnet 10, a gradient magnetic field coil 11, an RF coil 12, etc., and these components are housed in a cylindrical housing. The bed 500 has a bed body 50 and a tabletop 51. The magnetic resonance imaging apparatus 1 also has a local coil 20 disposed close to the subject.

[0017] The control cabinet 300 includes gradient magnetic field power supplies 31 (for the X axis 31x, for the Y axis 31y, and for the Z axis 31z), an RF receiver 32, an RF transmitter 33, and a sequence controller .

[0018] The static magnetic field magnet 10 of the magnetic gantry 100 has a roughly cylindrical shape and generates a static magnetic field within a bore (i.e., the space inside the cylinder of the static magnetic field magnet 10), which is the imaging region of a subject (e.g., a patient). The static magnetic field magnet 10 incorporates a superconducting coil, which is cooled to an extremely low temperature by liquid helium. In the excitation mode, the static magnetic field magnet 10 generates a static magnetic field by applying a current supplied from a static magnetic field power supply (not shown) to the superconducting coil. After that, when the static magnetic field magnet 10 transitions to the persistent current mode, the static magnetic field power supply is disconnected. Once transitioned to the persistent current mode, the static magnetic field magnet 10 continues to generate a strong static magnetic field for a long period of time, for example, for more than one year. The static magnetic field magnet 10 may also be configured as a permanent magnet.

[0019] The gradient magnetic field coil 11 also has a roughly cylindrical shape and is fixed inside the static magnetic field magnet 10. This gradient magnetic field coil 11 applies gradient magnetic fields to the subject in the X-axis, Y-axis, and Z-axis directions by currents supplied from gradient magnetic field power supplies (31x, 31y, 31z).

[0020] The bed body 50 of the bed 500 has a top plate 51 that can be moved up and down, and the subject placed on the top plate 51 is moved to a predetermined height before imaging. Then, during imaging, the top plate 51 is moved horizontally to move the subject into the bore.

[0021] The RF coil 12 is also called a WB (Whole Body) coil or a birdcage coil. The RF coil 12 is fixed in a roughly cylindrical shape so as to surround the subject inside the gradient magnetic field coil 11. The RF coil 12 transmits RF pulses transmitted from the RF transmitter 33 toward the subject, and also receives magnetic resonance signals emitted from the subject due to excitation of hydrogen nuclei.

[0022] The local coil 20 receives magnetic resonance signals emitted from the subject at a position close to the subject. The local coil 20 is composed of, for example, a plurality of element coils. There are various types of local coils 20 depending on the imaging region of the subject, such as for the head, chest, spine, lower limbs, or whole body, but Fig. 1 shows a local coil 20 for the chest.

[0023] The RF transmitter 33 transmits RF pulses to the RF coil 12 based on instructions from the sequence controller 34. On the other hand, the RF receiver 32 detects magnetic resonance signals received by the RF coil 12 and the local coil 20, and sends raw data obtained by digitizing the detected magnetic resonance signals to the sequence controller 34.

[0024] The sequence controller 34 scans the subject by driving the gradient magnetic field power supply 31, the RF transmitter 33, and the RF receiver 32 under the control of the MRI / MRS processing device 400. After performing the scan and receiving raw data from the RF receiver 32, the sequence controller 34 sends the raw data to the MRI / MRS processing device 400.

[0025] Of the components shown in FIG. 1 , the magnet gantry 100, the control cabinet 300, and the bed 500 are referred to as an imaging unit (or scanner). The MRI / MRS processing device 400 controls the imaging unit to perform magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS). The data processing device 600 performs various data processing operations on MR image data, MRS data, and MRSI (Magnetic Resonance Spectroscopic Imaging) data output from the MRI / MRS processing device 400. The operations of the MRI / MRS processing device 400 and the data processing device 600 will be described in more detail below.

[0026] 2 is a block diagram showing an example of the configuration of the MRI / MRS processing device 400 and the data processing device 600. The MRI / MRS processing device 400 is configured to include, for example, a memory circuitry 41, a display 42, an input interface 43, and a processing circuitry 40. Similarly, the data processing device 600 is also configured to include, for example, a memory circuit 61, a display 62, an input interface 63, and a processing circuit 60.

[0027] The storage circuits 41 and 61 are storage media including external storage devices such as ROM (Read Only Memory), RAM (Random Access Memory), HDD (Hard Disk Drive), optical disk device, etc. The storage circuits 41 and 61 store various information and data, as well as various programs executed by the processors provided in the processing circuits 40 and 60.

[0028] The displays 42, 62 are display devices such as a liquid crystal display panel, a plasma display panel, an organic EL panel, etc. The input interfaces 43, 63 are, for example, a mouse, a keyboard, a trackball, a touch panel, etc., and include various devices that allow the operator to input various information and data.

[0029] The processing circuits 40, 60 are circuits equipped with, for example, a CPU or a dedicated or general-purpose processor. The processor executes various programs stored in the storage circuits 41, 61 to realize the various functions described below. The processing circuits 40, 60 may be configured with hardware such as an FPGA (field programmable gate array) or an ASIC (application specific integrated circuit). The various functions described below can also be realized by such hardware. The processing circuits 40, 60 can also realize the various functions described below by combining software processing by a processor and a program with hardware processing.

[0030] The processing circuitry 40 of the MRI / MRS processing device 400 realizes the following functions: an imaging condition setting function F40, an MRI function F41, an MRS function F42, and an MRSI function F43. MRSI is also called CSI (Chemical Shift Imaging).

[0031] The imaging condition setting function F40 sets imaging conditions such as pulse sequences for MRI, MRS, and MRSI, which have been selected or set via the input interface 43, in the sequence controller .

[0032] The MRI function F41 reconstructs MR signals acquired from a subject by executing an MRI pulse sequence to generate MR image data. The generated MR image data is displayed on the display 42 and sent to the data processing device 600.

[0033] The MRS function F42 reconstructs MR signals acquired from a subject by executing an MRS pulse sequence to generate MRS data. The MRS data is, for example, spectral data obtained by Fourier transforming MR signals acquired from a specific region of the subject. The generated MRS data is displayed on the display 42 and sent to the data processing device 600. The specific region is also called a single voxel, a region of interest, or a ROI (Region of Interest).

[0034] The MRSI function F43 reconstructs MR signals acquired from a subject by executing a pulse sequence for MRSI to generate MRSI data. The MRSI data is, for example, a collection of spectral data for multiple voxels obtained by Fourier transforming MR signals acquired from multiple regions of the subject. The generated MRSI data is displayed on the display 42 and sent to the data processing device 600. The multiple regions are also called multi-voxels.

[0035] Meanwhile, the processing circuitry 60 of the data processing device 600 realizes an acquisition function F60, a UI (user interface) function F61, a display control function F62, and a correction function F63. The correction function F63 has, as its internal functions, a position correction data calculation function F64, an effective ROI calculation function F65, and a voxel position correction function F66.

[0036] The acquisition function F60 acquires data based on magnetic resonance signals collected from a specific region of the subject, for detecting a chemical shift of a substance, and also acquires position information regarding a reference region of interest set as the specific region.

[0037] The correction function F62 calculates position correction data representing the relationship between a designated chemical shift, which is arbitrarily designated within a predetermined range of chemical shifts, and the amount of displacement of a specific region caused by the designated chemical shift, and then calculates an effective region of interest, the position of which changes from the reference region of interest in response to the designated chemical shift, based on the position correction data. Although the MRI / MRS processing device 400 and the data processing device 600 are shown as separate devices in FIG. 1, they may be configured as a single device.

[0038] (First embodiment) Hereinafter, a more detailed operation of the data processing device 600 according to the first embodiment will be described with reference to the flowchart of Fig. 3 and the diagrams of Fig. 4 to Fig. 8. In the flowchart of Fig. 3, the processes from step ST100 to step ST104 are performed by, for example, the MRI / MRS processing device 400, and the processes from step ST200 to step ST206 are performed by, for example, the data processing device 600.

[0039] In step ST100, imaging is performed to generate a positioning image. The positioning image here is an MR image for setting a specific region from which MR signals for detecting a chemical shift in the subject by MRS are acquired, i.e., a region of interest. For example, when a region of interest is set in the subject's head, all or part of an axial, coronal, and sagittal image of the head may be used as the positioning image. For example, a coronal image of the head as shown in FIG. 5(a) may be used as the positioning image.

[0040] In step ST101, the generated positioning image is displayed on the display 42. Then, in step ST102, a specific region is set on the positioning image. This region may be hereinafter referred to as the reference region of interest (or reference ROI). Figure 5(a) shows an example of an image in which a rectangular frame indicating the region of the reference ROI is superimposed on the positioning image.

[0041] In step ST103, imaging conditions including a pulse sequence for MRS are set. FIG. 4 is a sequence diagram showing an example of a pulse sequence for MRS. In this pulse sequence, an excitation pulse with a flip angle of 90° is followed by two refocus pulses with a flip angle of 180°. Then, spin echoes after the second refocus pulse are collected as MR signals for generating an MRS spectrum. The excitation pulse and the two refocus pulses are applied with gradient magnetic fields Gz, Gy, and Gx for region selection, respectively.

[0042] In step ST104, the above-described pulse sequence is applied to the subject to perform MRS imaging. Then, MR signals acquired by this imaging are subjected to processing such as Fourier transform to generate a spectrum of the MR signals as MRS data. The generated MRS data (i.e., MRS spectrum) is sent from the MRI / MRS processing device 400 to the data processing device 600. In step ST200, the acquisition function F60 of the data processing device 600 acquires MRS data and position information of the reference ROI from the MRI / MRS processing device 400.

[0043] In step ST201, a spectral image IM02 based on the MRS data is displayed on the display 62 of the data processing device 600. FIG.

[0044] Additionally, in step ST201, an MRI image IM01 on which a reference ROI is superimposed, as shown in Fig. 5(a), is displayed on the display 62 of the data processing device 600. The MR image IM01 displayed on the display 62 may be acquired from the MRI / MRS processing device 400 in step ST200, for example. The MR image IM01 may be the positioning image described above, or may be an MR image including a diagnostic target region that is captured separately from the positioning image. In the spectral image IM02 illustrated in FIG. 5(b), for example, the horizontal axis represents frequency corresponding to chemical shift (ppm), and the vertical axis represents signal intensity, for example.

[0045] The chemical shift varies depending on the metabolic substance, such as Lac (lactic acid), NAA (N-acetylaspartic acid), Cr (creatine), and Cho (choline), and the chemical shift value of each metabolic substance is known.

[0046] On the other hand, it is known that the distribution state of each metabolite and the absolute or relative amount of each metabolite vary depending on the location within the subject (e.g., within the brain). It is also known that the distribution state of each metabolite differs between diseased sites such as tumors and normal sites, and furthermore, even within the same tumor site, the distribution state of each metabolite differs depending on the grade of the tumor.

[0047] Therefore, information about a specific metabolite detected as an MRS spectrum and the specific location in the subject where that metabolite is present is extremely important when diagnosing the presence or absence of a disease in the subject and the location of the disease.

[0048] However, as will be explained below, it is known that the data collection region, i.e., the region of interest (ROI), changes depending on the chemical shift value (i.e., the frequency value of the MRS spectrum). Note that, hereinafter, the region of interest that changes depending on the chemical shift value will be referred to as the effective region of interest (effective ROI), and will be distinguished from the reference region of interest (reference ROI), the position of which is fixed and set in step ST102.

[0049] 6 is a diagram explaining the difference between the reference ROI and the effective ROI one-dimensionally in the X direction. The horizontal axis of FIG. 6 is the position in the X direction, and the vertical axis is the magnetic resonance frequency f. The magnetic resonance frequency f is expressed, for example, by the following (Equation 1). f=f0+(γ / 2π)·Gx·X-δ f (Formula 1)

[0050] where f0 is the center frequency of the excitation pulse, Gx is the gradient field strength in the x direction, X is the position in the x direction, and δ f is the chemical shift frequency corresponding to the chemical shift δ (ppm), and γ is a constant (gyromagnetic ratio).

[0051] The proportional line shown by the solid line in Fig. 6 corresponds to the case where the chemical shift is zero, that is, the proportional relationship between the frequency f of a specific substance (reference substance) that serves as the reference for the chemical shift and the position X. In this case, the position and width Lx of the bold solid double-headed arrow in the figure correspond to the position and width in the x direction of the reference ROI described above.

[0052] On the other hand, the proportional line shown by the dashed line in Fig. 6 corresponds to the case where the chemical shift is non-zero, i.e., the proportional relationship between the frequency f of metabolic substances other than the reference substance, such as lactate, NAA, and choline, and the position X. In this case, the position and width Lx of the thick dashed double-headed arrow in the figure correspond to the position and width in the x direction of the above-mentioned effective ROI.

[0053] The position shift of the ROI shown in Fig. 6 also occurs in the y and z directions. Fig. 7 is a diagram illustrating the relationship between the two-dimensional reference ROI and the effective ROI in the x and y directions as viewed on the xy plane. As shown in Fig. 7, due to chemical shift, the reference ROI shown by the dashed line shifts by Dx in the x direction and Dy in the y direction to the position of the effective ROI shown by the solid line. The three-dimensional position shift amounts Dx, Dy, and Dz are expressed, for example, by the following (Equation 2), (Equation 3), and (Equation 4). Dx=K δ / Gx (Equation 2) Dy=K δ / Gy (Equation 3) Dz=K δ / Gz (Equation 4) Here, Gx, Gy, and Gz are the gradient magnetic field intensities in the x, y, and z directions, respectively, δ is the chemical shift (ppm), and K is a constant.

[0054] As can be seen from the above explanation, the magnetic resonance frequency of a substance of interest shifts slightly from the magnetic resonance frequency of a reference substance depending on the chemical shift of the substance, and as a result, the position of the region selectively excited by the excitation pulse and gradient magnetic field (i.e., region of interest (ROI)) effectively shifts depending on the chemical shift.

[0055] As mentioned above, information about a particular metabolite and the specific location within the subject where that metabolite is present is very important when diagnosing the presence or absence of a disease in the subject and the location of the disease. However, it is inconvenient if the position of the effective region of interest shifts in response to the chemical shift.

[0056] The data processing device 600 according to the first embodiment is provided with a user interface function F61 and a display control function F62 to solve such problems. Fig. 8 shows an example of an MR image IM01 and a spectroscopic image IM02 realized by the user interface function F61 and the display control function F62.

[0057] Returning to Fig. 3, in step ST202 in Fig. 3, a chemical shift is designated on the display. Specifically, as shown in Fig. 8(b), a vertical line marker for designating a specific chemical shift δ is displayed on the spectral image displayed on the display 62. The shape of the marker is not particularly limited, and may be a symbol of any shape.

[0058] In step ST203, position correction data is calculated based on the specified chemical shift 6. Specifically, a position correction data calculation function F64 of the correction function F63 calculates position shifts Dx, Dy, and Dz as position correction data based on (Equation 2), (Equation 3), (Equation 4), etc.

[0059] Next, in step ST204, the effective ROI is calculated based on the calculated position correction data. Specifically, the position of the effective ROI is calculated by adding or subtracting the position correction data Dx, Dy, and Dz to or from the position information of the reference ROI to calculate the coordinates of each vertex of a cube that indicates the area of ​​the effective ROI.

[0060] Then, in step ST205, the calculated effective ROI is displayed on the display 62. Specifically, as shown in Fig. 8(a), the area of ​​the effective ROI indicated by the solid-line rectangle is displayed superimposed on the MR image. Note that the dashed-line rectangle indicates the position of the reference ROI. The chemical shift can be easily specified by using a user interface such as the slide bar SB shown at the bottom of FIG. 8(b).

[0061] In step ST206, it is determined whether the designated chemical shift has been changed, and if it has been changed, the process returns to step ST202. By using a user interface such as the slide bar SB, the user can easily change the position of the designated chemical shift in real time while viewing the spectral image displayed on the display 62.

[0062] The position correction data calculation function F64 calculates the position of the effective ROI corresponding to the changed designated chemical shift in real time, and the display control function F62 moves the position of the effective ROI displayed on the MR image in real time in conjunction with the change in the designated chemical shift.

[0063] As described above, the data processing device 600 according to the first embodiment allows a user to easily recognize a chemical shift and an effective ROI region (i.e., a region where a metabolite corresponding to the chemical shift exists). As a result, for example, highly accurate disease diagnosis becomes possible.

[0064] (Second embodiment) The first embodiment described above corresponds to a method for acquiring MR data for MRS by a so-called single-voxel method, whereas the second embodiment corresponds to a method for acquiring MR data for MRSI (or CSI) by a multi-voxel method.

[0065] 9 is a flowchart showing an example of the operation of the data processing device 600 according to the second embodiment. Steps ST300 to ST302 are the same as those of the first embodiment. In step ST303, imaging conditions such as a pulse sequence for MRSI (or CSI) are set. Then, in step ST304, imaging of MRSI (or CSI) is performed according to the set imaging conditions. In step ST304, MR data for MRSI collected in imaging of MRSI (or CSI) is reconstructed to generate MRSI data. Furthermore, the generated MRSI data may be displayed on the display 42.

[0066] Fig. 10 shows an example of a pulse sequence for multi-voxel MRSI (or CSI). The difference from the pulse sequence for single voxel (Fig. 4) is that phase-encoding gradient magnetic fields Gpx and Gpy are applied between the 90° pulse (excitation pulse) and the first 180° pulse (refocus pulse) to add position information. By performing, for example, a three-dimensional Fourier transform on the MR signals S (Gpx, Gp, t) for MRSI acquired by such a pulse sequence, it is possible to reconstruct the spectrum for each voxel position (X, Y) as MRSI data.

[0067] 11(a) and (b) show examples of conventional displays of MRSI data generated by reconstruction. Image display IM03 shown in FIG. 11(a) is an example display in which multiple voxels acquired in MRSI imaging are displayed in a grid-like frame and overlaid on an MR image. On the other hand, image display IM04 shown in FIG. 11(b) is an example display in which spectral images calculated for each voxel in FIG. 11(a) are arranged corresponding to the position of each voxel.

[0068] In the following description using Figure 11 etc., the number of voxels is set to a small number such as 4 x 3 in order to simplify the illustration, but the number of voxels is not limited to this. For example, the number of voxels may be 64 x 64. The processes from step ST300 to step ST304 are performed by the MRI / MRS processing device 400, for example.

[0069] In step ST400, the acquisition function F60 of the data processing device 600 acquires MRSI data (or CSI data) and position information of the reference ROI from the MRI / MRS processing device 400. The reference ROI in the second embodiment may be the entire region including all voxels, or may be the region of an individual voxel. Hereinafter, the region of an individual voxel will be treated as the reference ROI. That is, the reference ROI in the first embodiment will be treated as the reference voxel in the second embodiment.

[0070] In step ST401, an MR image IM05 with a voxel frame superimposed thereon, as shown in Fig. 12(a), and a spectroscopic image IM06, as shown in Fig. 12(b), are displayed on the display 62 of the data processing device 600 based on the MRSI data. Each voxel frame shown in Fig. 12(a) corresponds to the reference voxel described above, and is displayed by, for example, a dashed line.

[0071] In the next step ST402, a voxel and a chemical shift are designated on the display 62. Specifically, as shown in FIG. 12(b), when the position of a voxel for which a chemical shift is to be designated is clicked with, for example, a mouse, the spectral image of the voxel is enlarged and displayed. Thereafter, as in the first embodiment, a specific chemical shift (or a specific metabolite) is designated using a user interface such as a slide bar SB and a marker located below the enlarged spectral image.

[0072] In step ST403, the position correction data calculation function F64 of the correction function F63 calculates the position shifts Dx, Dy, and Dz corresponding to the specified chemical shift as position correction data based on (Equation 2), (Equation 3), (Equation 4), etc.

[0073] In step ST404, the position of the effective voxel is calculated based on the calculated position correction data. Specifically, the position of the effective voxel is calculated by adding or subtracting the position correction data Dx, Dy, and Dz to or from the position information of the reference voxel to calculate the coordinates of each vertex of a cube that indicates the area of ​​the effective voxel.

[0074] Then, in step ST405, the calculated positions of the effective voxels are displayed on the display 62. Specifically, as shown in Fig. 12(a), the positions of the effective voxels corresponding to the designated voxels are displayed as solid-line rectangles and are superimposed on the MR image. As mentioned above, the dashed-line rectangles indicate the positions of the reference voxels.

[0075] In step ST406, it is determined whether or not to change the voxel or chemical shift. If a change is to be made, the process returns to step ST402. A new voxel can be specified by, for example, clicking on the desired voxel on the display screen of FIG. 12(b). Similarly to the first embodiment, the chemical shift can be easily changed by using a user interface such as the slide bar SB shown below the enlarged spectral image in FIG. 12(b).

[0076] (Modification of the second embodiment) The method of displaying MRSI data is not limited to the display method in which a plurality of spectral images are arranged in a grid pattern as shown in FIGS. Figure 13 shows an example of displaying MRSI data as an intensity map (i.e., a chemical shift image). In this display method, a specific chemical shift or a specific metabolite is specified. Then, from the reconstructed spectrum at each voxel position, the spectral intensity of the frequency corresponding to the specified chemical shift or metabolite is extracted, and the extracted spectral intensity is mapped as the signal intensity of each voxel.

[0077] 13 shows an example of three intensity maps corresponding to three metabolites, Cho, NAA, and Lac. Fig. 13 shows an example in which a grid-like intensity map corresponding to 5 × 4 voxels is displayed superimposed on an MR image. The number of voxels is not limited to this example, and may be, for example, 64 × 64.

[0078] Fig. 13 shows a conventional display example in which position correction according to chemical shift is not performed. On the other hand, Fig. 14 shows a display method according to a modification of the second embodiment. In the modification of the second embodiment, the position of the intensity map is corrected and displayed according to the chemical shift or the type of metabolite.

[0079] As a first method of position correction, similarly to the first and second embodiments described above, position shift amounts Dx, Dy, and Dz corresponding to chemical shifts are calculated based on (Equation 2), (Equation 3), and (Equation 4). Then, by using the calculated position shift amounts to shift each voxel position of the intensity map in real space, a position-corrected intensity map (i.e., a position-corrected chemical shift image) can be generated.

[0080] As a second method of position correction, phase correction data having the calculated position shift amount as a phase term is calculated, and the MR data expressed in k-space is corrected in k-space using this phase correction data. Then, by performing reconstruction processing using a Fourier transform or the like on the corrected k-space data, a position-corrected intensity map (i.e., a position-corrected chemical shift image) can be generated.

[0081] (Third embodiment) As described above, the position of the region of interest (ROI), which is the target region for data collection in single-voxel MRS or multi-voxel MRSI (or CSI), moves according to the chemical shift. Therefore, in the first and second embodiments, in addition to the position of the ROI set by the user (i.e., the reference ROI), the position of the effective ROI (i.e., the effective ROI) in which a metabolite having a specified chemical shift is estimated to actually exist is displayed on the display 62 or the like.

[0082] On the other hand, in order to suppress contamination of the region of interest in MRS or MRSI from surrounding tissues (e.g., surrounding adipose tissue), multiple saturation bands are sometimes set around the region of interest. This method of suppressing contamination from the surroundings of the region of interest is called OVS (outer volume suppression).

[0083] Fig. 15(a) is a diagram showing the position of the reference ROI in the MRS when OVS is not applied, and is the same as Fig. 5(a). In contrast, Fig. 15(b) is a diagram showing an example of the reference ROI and OVS region in the MRS when OVS is applied.

[0084] In the example shown in Figure 15(b), the OVS region is set by four strip-shaped regions surrounding the reference ROI. The OVS region is set, for example, by exciting four slabs (thick slices) with region-selective RF pulses (hereinafter referred to as OVS pulses). After applying the OVS pulses, for example, by applying a dephasing gradient magnetic field to disperse the phases of spins in the OVS region, the adverse effects of the OVS region can be suppressed.

[0085] The OVS pulse also undergoes a position shift according to the chemical shift, using the mechanism explained using Figure 6. However, since the OVS pulse and the excitation pulse (90° pulse) and refocusing pulse (180° pulse) shown in Figure 4 etc. have different properties such as frequency characteristics, the amount of movement of the position of the OVS region does not necessarily match the amount of movement of the position of the region of interest in MRS or MRSI.

[0086] Therefore, it is useful to provide users with information indicating the change in the position of the OVS region in addition to information indicating the change in the position of the ROI in MRS or MRSI. Hereinafter, the OVS region set by the user will be called the reference OVS region, and the OVS region that changes according to the specified chemical shift will be called the effective OVS region.

[0087] FIG. 16 corresponds to FIG. 8 in the description of the first embodiment. FIG. 16(a) illustrates a display image IM01 on the display 62, in which the effective OVS region and the reference OVS region are superimposed on the MR image in addition to the effective ROI and the reference ROI. Similarly to FIG. 8(b), FIG. 16(b) illustrates a display image IM02 of a spectrogram, which includes a user interface such as a slide bar SB. A chemical shift can be specified by operating the slide bar SB, and the positions of the effective ROI and the effective OVS region, which are independently calculated based on the specified chemical shift, move in conjunction with the chemical shift specification. As a result, the user can easily grasp the positions of the effective ROI and the effective OVS region.

[0088] Fig. 17 is a diagram corresponding to Fig. 14 in the explanation of the modified example of the second embodiment. In the modified example of the second embodiment, as shown in Fig. 14, the position of the intensity map is corrected and displayed according to the chemical shift or the type of metabolite.

[0089] In contrast to this, in the third embodiment, in addition to the intensity map whose position is corrected according to the chemical shift, an effective OVS region whose position is corrected according to the chemical shift is displayed, as shown in Fig. 17. The effective OVS region is a reference OVS region (not shown) set on the periphery of the region of interest of the multi-voxel MRSI, whose position is moved according to the chemical shift.

[0090] The shape of the OVS region is not limited to the shapes exemplified in FIG. 15(b) and FIG. 17. The shape of the OVS region can be determined by arranging multiple strip regions in any direction. FIG. 18 is a diagram illustrating an example of an effective OVS region corresponding to an OVS region having a shape different from that of FIG. 17. In this way, the shape of the OVS region can be determined by arranging multiple strip regions not only in the vertical direction or horizontal direction, but also in an oblique direction with any inclination angle. As a result, an OVS region can be set to surround the periphery of even a region of interest having a more complex shape without any gaps.

[0091] (Fourth embodiment) 19 is a block diagram showing an example of the configuration of an MRI / MRS processing device 400 and a data processing device 600 according to the fourth embodiment. The difference from the first and second embodiments is that the processing circuitry 60 of the data processing device 600 has an analysis function F67, and the other configurations are the same as those of the first and second embodiments.

[0092] The analysis function F67 of the data processing device 600 according to the fourth embodiment inputs an MRI image generated by MRI imaging and a spectrum generated by MRS imaging into a trained model generated by machine learning, and realizes a function of outputting a predetermined analysis result from the trained model, such as a classifier. For example, the analysis function F67 outputs a disease identification result, such as the presence or absence of a disease such as a tumor in a subject, the grade of the tumor, or the identification of the disease site, from the trained model as the predetermined analysis result. FIG. 20 is a flowchart showing an example of the operation of the MRI / MRS processing device 400 and the data processing device 600 according to the fourth embodiment.

[0093] In step ST500, imaging conditions such as a pulse sequence for MRI are set, and in step ST501, MRI imaging is performed according to the set imaging conditions. Then, an MR image is generated from the MR signals acquired by the MRI imaging.

[0094] Meanwhile, in step ST502, imaging conditions such as a pulse sequence for MRS are set, and in step ST503, MRS imaging is performed according to the set imaging conditions. Then, an MRS spectrum is generated from MR signals acquired by MRS imaging. Steps ST500 to ST503 are processes performed by the MRI / MRS processing device 400.

[0095] In steps ST600 and ST601, the acquisition function F60 of the data processing device 600 acquires the MR image and MRS spectrum generated by the MRI / MRS processing device 400.

[0096] In the next step ST602, one or more chemical shifts are designated as chemical shifts of interest, and in step ST603, spectral data in a predetermined range centered on each chemical shift of interest is extracted.

[0097] 21 is a diagram showing the concept of the processing of steps ST602 and ST603. In step ST602, for example, a spectral image as shown in FIG. 21(a) is displayed on the display 63, and chemical shifts of interest are designated using a user interface, for example, a user interface such as the slide bar SB shown in FIG. 8(b). In the example shown in FIGS. 21(a) and 21(b), a chemical shift of interest δ1 corresponding to Cho (choline), a chemical shift of interest δ2 corresponding to NAA (N-acetylaspartic acid), and a chemical shift of interest δ3 corresponding to Lac (lactic acid) are designated.

[0098] Then, in step ST603, as shown in FIG. 21(b), the spectral data in a predetermined range centered on three chemical shifts of interest δ1, δ2, and δ3 is extracted to generate three partial spectral data. In step ST602, instead of specifying a chemical shift of interest, a metabolite of interest itself may be specified.

[0099] In step ST604, a region of a partial MR image of interest is designated in the MR image. Then, in step ST605, position correction data is calculated based on the designated chemical shift. Furthermore, in step ST606, a partial MR image whose center position has been corrected based on the position correction data is cut out from the MR image.

[0100] 22 and 23 are diagrams illustrating the concept of the processing from step ST604 to step ST606. The dashed rectangular frame in the MR image in FIG. 22(a) indicates the region of the partial MR image before correction. The three solid rectangular frames indicate the region of the partial MR image after correction, in which the respective positions have been corrected using position correction data based on the chemical shifts of the three metabolites of interest, Cho, NAA, and Lac. Note that it is assumed that the pixel size of the MR image is smaller than the specific region or the size of the partial MR image.

[0101] As in the first embodiment, position correction is performed by calculating the position shifts Dx, Dy, and Dz corresponding to the specified chemical shift δ as position correction data based on (Equation 2), (Equation 3), (Equation 4), etc., and correcting the position of the partial MR image using this position correction data.

[0102] FIG. 23 is a diagram showing a processing concept for extracting position-corrected partial MR image regions from an MR image and generating three partial MR images corresponding to three chemical shifts of interest δ1, δ2, and δ3.

[0103] In step ST607, the partial spectrum data extracted in step ST603 and the partial MR image data extracted in diagram 606 are input to the trained model. Then, in step ST608, the analysis result by the trained model is acquired.

[0104] 24 is a diagram showing the processing concept of steps ST607 and ST608. The trained model is, for example, a classifier generated by machine learning. The trained model is, for example, a deep neural network (DNN) generated by deep learning (DL), but is not limited to this. For example, the trained model may be a classifier or classifier such as a support vector machine or a random forest.

[0105] As described above, the input to the trained model is a pair of partial spectral data and partial MR image data. The partial spectral data input to the trained model is data in which a predetermined range of spectrum is cut out centered on the chemical shift of a metabolite of interest. The partial MR image data input to the trained model is data of a partial MR image corresponding to a region in which the positional shift caused by the chemical shift has been corrected.

[0106] In the example shown in Figure 24, the following pairs are input into the trained model: (1) a pair of partial spectral data centered on the chemical shift δ1 of Cho and partial MR image data whose position has been corrected based on the chemical shift δ1 of Cho; (2) a pair of partial spectral data centered on the chemical shift δ2 of NAA and partial MR image data whose position has been corrected based on the chemical shift δ2 of NAA; and (3) a pair of partial spectral data centered on the chemical shift δ3 of Lac and partial MR image data whose position has been corrected based on the chemical shift δ3 of Lac. 24, three pairs are input to the trained model, but the number of pairs to be input may be one, or three or more. The analysis results output from the trained model are, for example, disease identification results such as the presence or absence of a disease such as a tumor in the subject, the grade of the tumor, or the identification of the diseased site.

[0107] If partial MR image data from an area where the positional shift caused by chemical shift has not been corrected is input into a trained model, the relationship between the metabolic substance of interest and the area in which it is present becomes unclear, and it may not be possible to obtain highly accurate analysis results.

[0108] In contrast, in the data processing device 600 according to the fourth embodiment, data of a partial MR image corresponding to a region where the positional deviation caused by chemical shift has been corrected is input as MR image data to be input to the trained model together with the partial spectral data. In other words, the metabolite of interest and the region where the metabolite actually exists are input to the trained model. As a result, it is possible to improve the accuracy of analysis results such as disease identification results.

[0109] As described above, the data processing device of the embodiment enables accurate identification of a region where a metabolite corresponding to a chemical shift of interest exists in MRS technology, thereby improving the accuracy of analysis using spectral information.

[0110] Although several embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included within the scope and spirit of the invention, as well as within the scope of the invention described in the claims and their equivalents. [Explanation of symbols]

[0111] 1. Magnetic resonance imaging device 40 Processing circuit 60 Processing circuit 400 MRI / MRS Processing Units 600 Data processing device F41 MRI function F42 MRS function F43 MRSI (CSI) function F61 User Interface Function F62 Display Control Function F63 correction function F64 Position correction data calculation function F65 Effective ROI Calculation Function F66 Voxel position correction function

Claims

1. an acquisition unit that acquires data based on magnetic resonance signals collected from a specific region of a subject, the data being for detecting a chemical shift of a substance; a calculation unit that calculates, based on a designated chemical shift arbitrarily designated within a predetermined range of the chemical shift, an amount of displacement of the position of the specific region that displaces in a specific direction due to the designated chemical shift, as position correction data of the designated chemical shift every time the designated chemical shift is changed; A data processing device comprising:

2. the acquisition unit further acquires position information regarding a reference region of interest set as the specific region; the calculation unit calculates an effective region of interest, the position of which changes from the reference region of interest in accordance with the specified chemical shift, based on the position correction data; 2. The data processing device according to claim 1.

3. The display and a display control unit that displays, on the display, a magnetic resonance image of the subject and a spectral image in which one axis represents the chemical shift and the other axis represents signal intensity; a user interface for specifying the specified chemical shift in the spectral image displayed on the display; Equipped with the calculation unit calculates a position of the effective region of interest corresponding to the specified chemical shift; the display control unit displays the effective region of interest superimposed on the magnetic resonance image; 3. The data processing device according to claim 2.

4. the user interface specifies a position of the specified chemical shift in the spectral image in a variably manner; the calculation unit calculates in real time the position of the effective region of interest corresponding to the designated chemical shift that is changeably designated on the user interface; the display control unit moves the position of the effective region of interest displayed on the magnetic resonance image in real time in conjunction with a change in the designated chemical shift designated in the spectroscopic image.

4. The data processing device according to claim 3.

5. the display control unit further superimposes the reference region of interest on the magnetic resonance image and displays it; 4. The data processing device according to claim 3.

6. the display control unit displays a linear or predetermined shaped symbol indicating the position of the specified chemical shift superimposed on the spectral image.

4. The data processing device according to claim 3.

7. a reconstruction unit that generates a chemical shift image by calculating a spectrum of the magnetic resonance signal for each of a plurality of voxels based on the acquired data, the reconstruction unit corrects voxel positions of the chemical shift image using the position correction data; 2. The data processing device according to claim 1.

8. the reconstruction unit corrects the data in k-space using phase correction data based on the position correction data, and performs reconstruction processing on the corrected data to generate the chemical shift image.

8. A data processing device according to claim 7.

9. An acquisition unit that acquires data based on magnetic resonance signals collected from a specific region of a subject, the data being for detecting a chemical shift of a substance; a calculation unit that calculates, based on a designated chemical shift arbitrarily designated within a predetermined range of the chemical shift, a displacement amount of the position of the specific region that displaces in a specific direction due to the designated chemical shift as position correction data of the designated chemical shift; an analysis unit that outputs a predetermined analysis result, the analysis unit receives as input (a) partial spectral data obtained by extracting a predetermined range centered on the designated chemical shift from spectral data of the magnetic resonance signal, and (b) a partial magnetic resonance image obtained by extracting a part of a magnetic resonance image generated by reconstructing magnetic resonance signals collected from a region including the specific region, the central position of the partial magnetic resonance image being corrected using the position correction data corresponding to the designated chemical shift. Data processing device.

10. the pixel size of the magnetic resonance image is smaller than the size of the specific region; 10. The data processing device according to claim 9.

11. the designated chemical shifts are a plurality of designated chemical shifts respectively associated with a plurality of metabolites; the partial magnetic resonance images are a plurality of partial magnetic resonance images respectively corrected with the position correction data corresponding to the plurality of specified chemical shifts; 10. The data processing device according to claim 9.

12. the analysis unit is configured as a classifier using a DNN (Deep Neural Network), The predetermined analysis result includes a disease identification result of the subject.

12. A data processing apparatus according to claim 11.

13. A magnetic resonance imaging apparatus comprising the data processing device according to any one of claims 1 to 12.

Images