JP7366575B2 - Magnetic resonance imaging device, magnetic resonance imaging method, and image processing device - Google Patents

Description

Embodiments of the present invention relate to a magnetic resonance imaging apparatus, a magnetic resonance imaging method, and an image processing apparatus.

In recent years, magnetic resonance imaging (MRI) devices have been used to change the values of imaging parameters such as repetition time (TR), echo time (TE), and inversion time (TI). After generating quantitative parameter maps such as proton density (PD), longitudinal relaxation time (T1), transverse relaxation time (T2), and apparent diffusion coefficient (ADC) from multiple collected images, A technique is known in which a synthetic contrast-enhanced image corresponding to various contrast-enhanced images used in normal diagnosis is generated by calculation using signal values of a quantitative parameter map and values of arbitrary imaging parameters.

Japanese Patent Application Publication No. 2018-033893 Japanese Patent Application Publication No. 2016-123865 US Patent No. 7,340,290 Japanese Patent Application Publication No. 2017-225501

Warntjes JB, et al. , “Rapid magnetic resonance quantification on the brain: optimization for clinical usage”, Magn Reson Med. 2008, 60, p. 320-329 Hagiwara A, et al. , “SyMRI of the Brain: Rapid Quantification of Relaxation Rates and Proton Density, With Synthetic MRI, Automatic Brain Segmentation, and Myelin Measurement”, Investigative Radiology. 2017, 52, p. 647-657 Tanenbaum LN, et al, “Synthetic MRI for Clinical Neuroimaging: Results of the Magnetic Resonance Image Compilation (MAGiC) Prospective, Multicenter, Multireader Trial”, AJNR 2017, 38, p. 1103-1110 Hwang KP, et al., “3D isotropic multi-parameter mapping and synthetic imaging of the brain with 3D-QALAS: Comparison with 2D MAGIC”, ISMRM. 2018, #5627 Gong E, at al., “Improved Synthetic MRI from multi-echo MRI Using Deep Learning”, ISMRM. 2018, #279 Macro Essig, et al., “Assessment of cerebral gliomas by a new dark fluid sequence, high intensity reduction (HIRE): a preliminary study”, JMRI. 2000 May; 11(5): p. 506-517 Salminen LE, Conturo TE, Bolzenius JD, Cabeen RP, Akbudak E, Paul RH, “Reducing CSF Partial Volume Effects to Enhance Diffusion Tensor Imaging Metrics of Brain Microstructure”, Technol Innov. 2016 Apr; 18(1): p. 5-20, Epub 2016

The problem that the present invention seeks to solve is to suppress partial volume effects that occur in quantitative parameter maps due to the mixing of water and tissue components in the signal.

The MRI apparatus according to the embodiment includes an acquisition section, a separation section, and a generation section. The acquisition unit captures images of the same subject in which multiple types of tissues coexist by changing the values of multiple types of imaging parameters. A plurality of images including images taken with a long echo time such that the water component is dominant are collected. The separation unit generates a separated image in which a water component and a tissue component are separated by analyzing at least three images having different echo times among the plurality of images. The generation unit uses the separated images to generate a quantitative parameter map of at least one of longitudinal relaxation time, transverse relaxation time, proton density, and diffusion coefficient.

FIG. 1 is a diagram showing an example of the configuration of an MRI apparatus according to the first embodiment. FIG. 2 is a flowchart showing the procedure of processing executed by the MRI apparatus according to the first embodiment. FIG. 3 is a diagram showing images collected with different TE and b values as two-dimensional data based on the two-component model used in the example of the first embodiment. FIG. 4 is a diagram illustrating an example of suppression of water components in an image and quantitative parameter map generated by the generation function according to the first embodiment. FIG. 5 is a diagram illustrating an example of suppression of water components in a composite contrast-enhanced image generated by the composition function according to the first embodiment. FIG. 6 is a diagram illustrating an example of suppression of a water component by a b value in an image captured by diffusion weighted imaging according to Example 2. FIG. 7 is a diagram showing the relationship between signal intensity and b value in DWI collected by diffusion weighted imaging according to Example 6. FIG. 8 is a diagram illustrating an example of generation of separated images performed by the separation function according to the seventh embodiment. FIG. 9 is a diagram illustrating an example of generation of separated images performed by the separation function according to the eighth embodiment. FIG. 10 is a diagram illustrating an example of a nonlinear function used by the separation function according to the eighth embodiment. FIG. 11 is a diagram illustrating a configuration example of an image processing apparatus according to the second embodiment.

Hereinafter, embodiments of an MRI apparatus, an MRI method, and an image processing apparatus according to the present application will be described in detail with reference to the drawings.

(First embodiment)
FIG. 1 is a diagram showing an example of the configuration of an MRI apparatus according to the first embodiment.

For example, as shown in FIG. 1, the MRI apparatus 100 according to the present embodiment includes a static magnetic field magnet 1, a gradient magnetic field coil 2, a gradient magnetic field power supply 3, a whole body RF coil 4, a local RF coil 5, a transmitting circuit 6, It includes a receiving circuit 7, an RF (Radio Frequency) shield 8, a pedestal 9, a bed 10, an interface 11, a display 12, a storage circuit 13, and processing circuits 14 to 16.

The static magnetic field magnet 1 generates a static magnetic field in the imaging space where the subject S is placed. Specifically, the static magnetic field magnet 1 is formed into a hollow, substantially cylindrical shape (including one whose cross section perpendicular to the central axis has an elliptical shape), and the static magnetic field magnet 1 is formed in an imaging space located on the inner circumference side. Generates a static magnetic field. For example, the static magnetic field magnet 1 includes a cooling container formed in a substantially cylindrical shape and a magnet such as a superconducting magnet immersed in a coolant (for example, liquid helium, etc.) filled in the cooling container. Note that the static magnetic field magnet 1 may be one that generates a static magnetic field using, for example, a permanent magnet.

The gradient magnetic field coil 2 is arranged inside the static magnetic field magnet 1, and generates a gradient magnetic field in the imaging space where the subject S is arranged. Specifically, the gradient magnetic field coil 2 is formed in a hollow, substantially cylindrical shape (including one in which the shape of the cross section perpendicular to the central axis is elliptical), and has an X axis, a Y axis, and a Z axis that are orthogonal to each other. It has an X coil, a Y coil, and a Z coil corresponding to each axis. The X coil, Y coil, and Z coil generate gradient magnetic fields in the imaging space that linearly change along each axis direction based on currents supplied from the gradient magnetic field power supply 3. Here, the Z axis coincides with the cylindrical axis of the gradient magnetic field coil 2 and is set along the magnetic flux of the static magnetic field generated by the static magnetic field magnet 1. Further, the X-axis is set along a horizontal direction perpendicular to the Z-axis, and the Y-axis is set along a vertical direction perpendicular to the Z-axis. Thereby, the X-axis, Y-axis, and Z-axis constitute an apparatus coordinate system specific to the MRI apparatus 100.

The gradient magnetic field power supply 3 supplies a current to the gradient magnetic field coil 2 to generate a gradient magnetic field in the imaging space that linearly changes along each of the readout direction, phase encode direction, and slice direction. Specifically, the gradient magnetic field power supply 3 individually supplies current to the X coil, Y coil, and Z coil of the gradient magnetic field coil 2, thereby generating electric current along each of the readout direction, phase encoding direction, and slice direction that are orthogonal to each other. generates a gradient magnetic field that changes linearly. Note that in the following, a gradient magnetic field along the readout direction is referred to as a readout gradient magnetic field, a gradient magnetic field along the phase encode direction is referred to as a phase encode gradient magnetic field, and a gradient magnetic field along the slice direction is referred to as a slice gradient magnetic field. .

Here, the readout gradient magnetic field, phase encode gradient magnetic field, and slice gradient magnetic field are superimposed on the static magnetic field generated by the static magnetic field magnet 1, thereby imparting spatial position information to the MR signal generated from the subject S. Give. Specifically, the readout gradient magnetic field adds positional information along the readout direction to the MR signal by changing the frequency of the MR signal according to the position in the readout direction. Further, the phase encode gradient magnetic field changes the phase of the MR signal along the phase encode direction, thereby imparting position information along the phase encode direction to the MR signal. Further, the slice gradient magnetic field gives position information along the slice direction to the MR signal. For example, when the imaging region is a slice region, the slice gradient magnetic field is used to determine the direction, thickness, and number of slices, and when the imaging region is a volume region, the slice gradient magnetic field is used to determine the direction, thickness, and number of slices. This is used to change the phase of the MR signal. Thereby, the axis along the readout direction, the axis along the phase encode direction, and the axis along the slice direction constitute a logical coordinate system for defining the slice region or volume region to be imaged.

The whole body RF coil 4 is disposed on the inner circumferential side of the gradient magnetic field coil 2, and applies an RF magnetic field to the subject S placed in the imaging space, and MR generated from the subject S due to the influence of the RF magnetic field. Receive a signal. Specifically, the whole body RF coil 4 is formed in a hollow, substantially cylindrical shape (including one in which the cross section perpendicular to the central axis has an elliptical shape), and receives the RF pulses supplied from the transmitting circuit 6. Based on the signal, an RF magnetic field is applied to the subject S placed in the imaging space located on the inner circumferential side thereof. Further, the whole body RF coil 4 receives an MR signal generated from the subject S under the influence of the RF magnetic field, and outputs the received MR signal to the receiving circuit 7. For example, the whole body RF coil 4 is a QD (quadrature) coil.

The local RF coil 5 receives an MR signal generated from the subject S. Specifically, the local RF coil 5 is prepared for each part of the subject S, and when the subject S is imaged, it is placed near the part to be imaged. Then, the local RF coil 5 receives the MR signal generated from the subject S under the influence of the RF magnetic field applied by the whole body RF coil 4, and outputs the received MR signal to the receiving circuit 7. Note that the local RF coil 5 may further have a function of applying an RF magnetic field to the subject S. In that case, the local RF coil 5 is connected to the transmission circuit 6 and applies an RF magnetic field to the subject S based on the RF pulse signal supplied from the transmission circuit 6. For example, the local RF coil 5 is a surface coil or an array coil composed of a plurality of surface coils.

The transmitting circuit 6 outputs to the whole body RF coil 4 an RF pulse signal corresponding to the Larmor frequency specific to the target atomic nucleus placed in the static magnetic field. Specifically, the transmitter circuit 6 includes a pulse generator, an RF generator, a modulator, and an amplifier. The pulse generator generates an RF pulse signal waveform. The RF generator generates an RF signal at a resonant frequency. The modulator generates an RF pulse signal by modulating the amplitude of the RF signal generated by the RF generator with the waveform generated by the pulse generator. The amplifier amplifies the RF pulse signal generated by the modulator and outputs it to the whole body RF coil 4.

The receiving circuit 7 generates MR signal data based on the MR signal output from the whole body RF coil 4 or the local RF coil 5, and outputs the generated MR signal data to the processing circuit 15. For example, the receiving circuit 7 includes a selector, a preamplifier, a phase detector, and an A/D (Analog/Digital) converter. The selector selectively inputs the MR signal output from the whole body RF coil 4 or the local RF coil 5. The preamplifier power amplifies the MR signal output from the selector. The phase detector detects the phase of the MR signal output from the front stage amplifier. The A/D converter generates MR signal data by converting the analog signal output from the phase detector into a digital signal, and outputs the generated MR signal data to the processing circuit 15.

The RF shield 8 is disposed between the gradient magnetic field coil 2 and the whole body RF coil 4 and shields the gradient magnetic field coil 2 from the RF magnetic field generated by the whole body RF coil 4. Specifically, the RF shield 8 is formed into a hollow, substantially cylindrical shape (including one in which the shape of the cross section perpendicular to the central axis of the cylinder is elliptical), and is located on the inner circumferential side of the gradient magnetic field coil 2. It is arranged in the space so as to cover the outer peripheral surface of the whole body RF coil 4.

The pedestal 9 has a hollow bore 9a formed in a substantially cylindrical shape (including one whose cross section perpendicular to the central axis has an elliptical shape), and includes a static magnetic field magnet 1, a gradient magnetic field coil 2, and a whole body RF. It houses a coil 4 and an RF shield 8. Specifically, in the pedestal 9, the whole body RF coil 4 is arranged on the outer circumference side of the bore 9a, the RF shield 8 is arranged on the outer circumference side of the whole body RF coil 4, and the gradient magnetic field coil is arranged on the outer circumference side of the RF shield 8. 2 are arranged, and the static magnetic field magnet 1 is arranged on the outer circumferential side of the gradient magnetic field coil 2, and each of them is accommodated. Here, the space within the bore 9a of the pedestal 9 becomes an imaging space in which the subject S is placed during imaging.

The bed 10 includes a top plate 10a on which the subject S is placed, and when the subject S is imaged, the top plate 10a on which the subject S is placed is moved into the imaging space. For example, the bed 10 is installed such that the longitudinal direction of the top plate 10a is parallel to the central axis of the static magnetic field magnet 1.

Here, an example will be described in which the MRI apparatus 100 has a so-called tunnel-type configuration in which the static magnetic field magnet 1, the gradient magnetic field coil 2, and the whole-body RF coil 4 are each formed in a substantially cylindrical shape. The embodiment is not limited to this. For example, the MRI apparatus 100 has a so-called open type configuration in which a pair of static magnetic field magnets, a pair of gradient magnetic field coils, and a pair of RF coils are arranged to face each other across an imaging space in which the subject S is arranged. You may do so. In this case, the space sandwiched between the pair of static field magnets, the pair of gradient magnetic field coils, and the pair of RF coils corresponds to the bore in the tunnel-type configuration.

The interface 11 receives various instructions and input operations for various information from an operator. Specifically, the interface 11 is connected to the processing circuit 17 , converts an input operation received from an operator into an electrical signal, and outputs the electrical signal to the processing circuit 17 . For example, the interface 11 includes a trackball, a switch button, a mouse, a keyboard, a touch pad for performing input operations by touching the operation surface, and a display screen for setting imaging conditions and a region of interest (ROI). This is realized by a touch screen with an integrated touch pad, a non-contact input circuit using an optical sensor, a voice input circuit, etc. Note that in this specification, the interface 11 is not limited to one that includes physical operation parts such as a mouse and a keyboard. For example, an example of the interface 11 includes an electrical signal processing circuit that receives an electrical signal corresponding to an input operation from an external input device provided separately from the device and outputs this electrical signal to a control circuit.

The display 12 displays various information and various images. Specifically, the display 12 is connected to the processing circuit 17, and converts various information and various image data sent from the processing circuit 17 into electrical signals for display, and outputs the electrical signals. For example, the display 12 is realized by a liquid crystal monitor, a CRT monitor, a touch panel, or the like.

The storage circuit 13 stores various data. Specifically, the storage circuit 13 stores MR signal data and image data. For example, the memory circuit 13 is realized by a semiconductor memory element such as a RAM (Random Access Memory) or a flash memory, a hard disk, an optical disk, or the like.

The processing circuit 14 has a bed control function 14a. The bed control function 14a controls the operation of the bed 10 by outputting a control electrical signal to the bed 10. For example, the bed control function 14a receives an instruction from an operator to move the top plate 10a in the longitudinal direction, vertical direction, or left and right direction via the interface 11, and controls the bed so that the top plate 10a is moved according to the received instruction. 10 is operated.

The processing circuit 15 has a sequence execution function 15a. The sequence execution function 15a executes pulse sequences for various imaging methods by driving the gradient magnetic field power supply 3, the transmission circuit 6, and the reception circuit 7 according to sequence execution data output from the processing circuit 17. Here, the sequence execution data is data representing a pulse sequence, and includes the timing and strength of the current supplied by the gradient magnetic field power source 3 to the gradient magnetic field coil 2, and the RF This information defines the timing of supplying the pulse signal, the strength of the RF pulse to be supplied, the timing at which the receiving circuit 7 samples the MR signal, etc. Then, the sequence execution function 15a receives the MR signal data output from the reception circuit 7 as a result of executing the pulse sequence, and stores it in the storage circuit 13. At this time, the MR signal stored in the storage circuit 13 has positional information along each of the readout direction, phaseout direction, and slice direction due to the readout gradient magnetic field, phase encode gradient magnetic field, and slice gradient magnetic field described above. By being assigned, it is stored as k-space data arranged in two or three dimensions.

The processing circuit 16 has an image generation function 16a. The image generation function 16a generates an image based on the MR signal data stored in the storage circuit 13. Specifically, the image generation function 16a reads the MR signal data collected by the sequence execution function 15a from the storage circuit 13, and performs reconstruction processing such as Fourier transformation on the read MR signal data, thereby generating two-dimensional or Generate a three-dimensional image. The image generation function 16a also stores the generated image in the storage circuit 13, reads out image data from the storage circuit 13, and outputs it to the display 12 in response to a request from the operator.

The processing circuit 17 performs overall control of the MRI apparatus 100 by controlling each component included in the MRI apparatus 100. Specifically, the processing circuit 17 displays a GUI (Graphical User Interface) on the display 12 for receiving various instructions and various information input operations from the operator, and displays the GUI (Graphical User Interface) in response to input operations received via the interface 11. and controls each component included in the MRI apparatus 100. For example, the processing circuit 17 generates sequence execution data based on imaging conditions input by the operator, and outputs the generated sequence execution data to the processing circuit 15, thereby inputting various pulse sequences to the sequence execution function 15a. Let it run. Furthermore, the processing circuit 17 collects various images by controlling the image generation function 16a of the processing circuit 16 to generate images based on MR signal data collected by executing a pulse sequence.

The configuration example of the MRI apparatus 100 according to the present embodiment has been described above. With such a configuration, the MRI apparatus 100 according to the present embodiment has a function of generating quantitative parameter maps such as PD, T1, T2, etc. from a plurality of images acquired by changing the values of a plurality of types of imaging parameters. have. Furthermore, the MRI apparatus 100 according to the present embodiment has a function of calculating and generating a composite contrast-enhanced image corresponding to various contrast-enhanced images using the signal values of the generated quantitative parameter map and the values of arbitrary imaging parameters. have.

For example, a technique called Synthetic MRI (SyMRI) is known as a technique for generating such a synthetic contrast-enhanced image by calculation. This SyMRI has the advantage of being able to obtain quantitative parameter maps that are less dependent on the MRI device model and conditions, and that the overall imaging time, including imaging and processing, is longer than that required when multiple types of contrast-enhanced images are acquired through actual acquisition. This method has advantages such as being able to shorten the time required for image capture, and being able to generate an arbitrary composite contrast-enhanced image using arbitrary imaging parameters, so that an optimal contrast-enhanced image can be generated after the fact. Note that the information on the contrast-enhanced image is reflected in quantitative parameter maps such as T1, T2, PD, etc., but the contrast-enhanced image is generated after the fact because it is commonly used by doctors in normal diagnosis. This is because they are used to contrast-enhanced images.

In addition, in recent years, MR Finger Printing (MR Finger Printing), which combines high-speed incoherent collection using k-space spiral trajectory of SSFP (Steady State Free Precession) sequences and dictionary matching to quickly generate quantitative parameter maps from minimal data. A method of generating contrast-enhanced images using a method called MRF is also attracting attention as SyMRI in a broader sense.

Furthermore, a quantitative parameter map of the ADC is generated from images taken by changing the b value representing the intensity of signal attenuation due to phase dispersion due to diffusion using diffusion weighted imaging (DWI), and A method of generating a computational diffusion weighted image using a parameter map and values of arbitrary imaging parameters is also known. In DWI, suppression of water, ie, CSF (Cerebral Spinal Fluid), is important in quantifying parameters such as ADC and FA, or in improving the accuracy of tractography for extracting nerve fibers. For this reason, imaging methods for imaging DWI include methods that use FLAIR (Fluid Attenuated Inversion Recovery) during data collection, set the b value in two or multiple stages, and analyze water components and tissue components analytically from data from multiple directions. There are known methods for separating these.

Here, the above-mentioned methods have in common that the generation of a quantitative parameter map precedes the generation of a composite contrast-enhanced image, and in all methods, a partial volume effect (PVE) is applied to the quantitative parameter map. As a result, the synthesized contrast-enhanced image obtained using these deviates from the original quantitative value due to the mixing of different tissues within the same voxel. Artifacts may occur.

For example, in a T2-FLAIR image of the head generated by SyMRI, high signals that occur near the ventricles, cerebral sulci of the cortex, and brain tissue appear as artifacts that do not occur in T2-FLAIR images obtained by actual acquisition. There are many reports that it is an obstacle to diagnosis. Such high signal artifacts are thought to be caused by PVE caused by mixing of CSF and tissue within the voxel.

In other words, in FLAIR images obtained by actual acquisition, the longitudinal magnetization Mz of water (CSF) is set to zero by an IR (Inversion Recovery) pulse, and then the signal is collected as the transverse magnetization Mxy, so the longitudinal magnetization of the water is completely zero. If so, even if a water component is included, in principle the signal will only be a tissue component. Therefore, PVE does not occur in the quantitative parameter map. In contrast, in SyMRI, quantitative parameter maps are usually generated from signals collected when the longitudinal magnetization of water is not zero, resulting in PVE due to mixing of water and tissue components in the signal. . As a result, in Synthetic FLAIR images generated based on quantitative parameters, the influence of longitudinal magnetization of water cannot be avoided. In particular, since T2 of water is long compared to tissue, by setting TE long, water The more emphasized the Synthetic FLAIR image, the more prominent the high signal artifacts due to PVE will be.

For this reason, the MRI apparatus 100 according to the present embodiment is configured to be able to suppress PVE that occurs in the quantitative parameter map due to the mixing of the water component and tissue component in the signal.

Specifically, in this embodiment, the processing circuit 17 shown in FIG. 1 has a collection function 17a, a separation function 17b, a generation function 17c, and a composition function 17d. Here, the collection function 17a is an example of a collection unit. Furthermore, the separation function 17b is an example of a separation unit. Further, the generation function 17c is an example of a generation unit. Furthermore, the composition function 17d is an example of a composition section.

Note that the processing circuits 14 to 17 described above are realized by, for example, a processor. In this case, the processing functions of each processing circuit are stored in the storage circuit 13 in the form of a computer-executable program, for example. Each processing circuit reads each program from the storage circuit 13 and executes it, thereby realizing a processing function corresponding to each program. In other words, each processing circuit in a state where each program has been read has each function shown in each processing circuit in FIG. Although each processing circuit has been described here as being realized by a single processor, each processing circuit is configured by combining multiple independent processors, and each processor executes a program to perform each processing function. It may be possible to realize this. Further, the processing functions of each processing circuit may be appropriately distributed or integrated into a single processing circuit or a plurality of processing circuits. In addition, in the example shown in FIG. 1, the single memory circuit 13 has been described as storing programs corresponding to each processing function, but it is also possible to arrange a plurality of memory circuits in a distributed manner so that the processing circuit has individual memory. A configuration may also be used in which a corresponding program is read out from the circuit.

FIG. 2 is a flowchart showing the procedure of processing executed by the MRI apparatus 100 according to the first embodiment.

For example, as shown in FIG. 2, in the present embodiment, the acquisition function 17a captures images of the same subject in which multiple types of tissues coexist by changing the values of multiple types of imaging parameters. A plurality of images including an image taken with a TE close to T2 and an image taken with a long TE such that the water component is dominant in the signal are collected (step S01). At this time, the collection function 17a collects the plurality of images by controlling the sequence execution function 15a of the processing circuit 15 and the image generation function 16a of the processing circuit 16. The processing in step S01 is realized, for example, by the processing circuit 17 reading out a predetermined program corresponding to the collection function 17a from the storage circuit 13 and executing it.

Further, the separation function 17b generates at least two separated images in which the water component and the tissue component are separated by analyzing at least three images having different TEs among the plurality of images collected by the collection function 17a. (Step S02). At this time, the separation function 17b reads out the plurality of images generated by the image generation function 16a from the storage circuit 13 and analyzes them. The process of step S02 is realized, for example, by the processing circuit 17 reading out a predetermined program corresponding to the separation function 17b from the storage circuit 13 and executing it.

Furthermore, the generation function 17c generates at least one type of quantitative parameter map among T1, T2, PD, and ADC using at least two separated images generated by the separation function 17b (step S03). The processing in step S03 is realized, for example, by the processing circuit 17 reading out a predetermined program corresponding to the generation function 17c from the storage circuit 13 and executing it.

Further, the synthesis function 17d uses at least one type of quantitative parameter map generated by the generation function 17c to generate a synthesized contrast-enhanced image corresponding to an arbitrary imaging parameter by calculation (step S04). At this time, the composition function 17d displays the generated composite contrast-enhanced image on the display 12. The process of step S04 is realized, for example, by the processing circuit 17 reading out a predetermined program corresponding to the composition function 17d from the storage circuit 13 and executing it.

According to such a configuration, by generating a quantitative parameter map using a separated image in which water components and tissue components in a signal are separated, a quantitative parameter map is generated by mixing the water component and tissue component in the signal. PVE that occurs can be suppressed. Furthermore, by generating a composite contrast-enhanced image using a quantitative parameter map in which PVE is suppressed, it is possible to suppress high-signal artifacts caused in the composite contrast-enhanced image due to PVE.

Hereinafter, specific examples of the first embodiment described above will be described. In the following examples, the quantitative maps of T1, T2, PD, and ADC are referred to as T1 map, T2 map, PD map, and ADC map.

First, the MR signal model used in the following examples will be explained. Here, the MR signal model is a model that defines the relationship between the overall signal strength of the signal and the signal strength of each of the water component and tissue component within the signal.

In the following examples, it is assumed that one unit voxel corresponds to one MR signal, and an MR signal model in which the component in a unit voxel is single is called a single-compartment model. An MR signal model in which two components, a water component and a tissue component, are included in a voxel is called a two-comparment model. In addition, here, for each model, an MR signal model obtained by the spin echo (SE) method and an MR signal model obtained by the inversion recovery (IR) method are defined.

For example, regarding a one-component model, if the MR signal obtained by the SE method is S _SE and the MR signal obtained by IR is S _IR , then the one-component model of the SE method and the one-component model of the IR method are as follows ( Expression 1) and (Formula 2).

S _SE (TR, TE, b, PD, T1, T2, ADC) = PD * DecayT1 _SE (TR, T1) * DecayT2 (TE, T2) * DecayADC(b, ADC) ... (Formula 1)
S _IR (TR, TI, TE, b, PD, T1, T2, ADC) = PD * DecayT1 _IR (TR, TI, T1) * DecayT2 (TE, T2) * DecayADC (b, ADC) ... (Eq. 2)

Here, TR is the repetition time, TE is the echo time, TI is the inversion time, and b is the b value. Further, TI is longitudinal relaxation time, T2 is transverse relaxation time, PD is proton density, and ADC is diffusion coefficient.

In addition, DecayT1 _SE (TR, T1) is the T1 attenuation in the SE method, DecayT1 _IR (TR, TI, T1) is the T1 attenuation in the IR method, DecayT2 (TE, T2) is the T2 attenuation common to the SE method and the IR method, DecayADC (b, ADC) is ADC attenuation common to the SE method and the IR method, and is expressed by the following (Formula 3) to (Formula 6), respectively.

DecayT1 _SE (TR, T1) = 1-2*exp[-(TR-TE/2)/T1]-exp[-TR/T1]...(Formula 3)
DecayT1 _IR (TR, TI, T1) = 1-2*exp(-TI/T1)+2*exp[-(TR-TE/2)/T1]-exp(-TR/T1)...(Formula 4 )
DecayT2 (TE, T2) = exp (-TE/T2) ... (Formula 5)
DecayADC (b, ADC) = exp (-b*ADC) ... (Formula 6)

Further, for example, regarding the two-component model, if the volume of the water component in a unit voxel is Vw and the volume of the tissue component is _Vt , the volume V of the unit voxel is expressed by the following equation (7). Furthermore, if the PD of the water component is PDw and the PD of the tissue component is _PDt , then the average proton density PD within a unit voxel is expressed by the following (Equation 8).

V = V _w + V _t = 1 (Formula 7)
PD=V _w *PD _w +V _t *PD _t ... (Formula 8)

Then, S _SEw is the MR signal of the water component obtained by the SE method, S SEt is the MR signal of the tissue component obtained by the SE method, S _{IRw is} the MR signal of the water component obtained by the IR method, and S _IRw is the MR signal of the water component obtained by the IR method. When the component MR signal is S _IRt , the two-component model of the SE method and the two-component model of the IR method are expressed by the following (Formula 9) and (Formula 10), respectively.

S _SE (TR, TE, b, PD, T1, T2, ADC) = S _SEw (TR, TE, b, PD _w , T1 _w , T2 _w , ADC _w ) + S _SEt (TR, TE, b, PD _t , T1 _t , T2 _t , ADC _t ) ... (Formula 9)
S _IR (TR, TI, TE, b, PD, T1, T2, ADC) = S _IRw (TR, TI, TE, b, PD _w , T1 _w , T2 _w , ADC _w ) + S _IRt (TR, TI, TE, b, PD _t , T1 _t , T2 _t , ADC _t ) ... (Formula 10)

Here, T1 _w is TI of water, T2 _w is T2 of water, and ADC _w is ADC of water. Further, T1 _t is the T1 of the tissue, T2 _t is the T2 of the tissue, and ADC _t is the ADC of the tissue.

In this way, the MR signal model used in the following examples defines the MR signal as the product of PD and three types of signal attenuation terms (T1 attenuation, T2 attenuation, ADC attenuation). Here, PD, T1, and T2 of the water component and the tissue component have values specific to each component.

In addition, in the above example, we decided to define MR signal models for the SE method and the IR method, but by setting TI=TR, the T1 attenuation (DecayT1) becomes the same for the SE method and the IR method. , each can be unified with a common model. Furthermore, in calculating T1 attenuation (DecayT1), TE may be approximated as 0.

FIG. 3 is a diagram showing images collected with different TE and b values as two-dimensional data based on the two-component model used in the example of the first embodiment.

Here, FIG. 3 shows the signal strength S(TE, b) when an image is captured by the SE method with TR and TI constant (that is, longitudinal magnetization Mz is constant). In addition, in FIG. 3, the dashed line indicates the signal strength S _w of the water component, the dashed double line indicates the signal strength S _t of the tissue component, and the dashed line indicates the signal strength S w of the water component. The overall signal strength S, which is the sum of the components and tissue components, is shown.

For example, in the following example, S _SE1 (TE1, b0), S _SE2 (TE2, b0), S _SE3 (TE3, b0), S _DWI1 (TE1, b1), S _DWI2 (TE2, b1) shown in FIG. ), S _DWI3 (TE3, b1) and S _DWI4 (TE2, b2), seven images are used corresponding to the seven acquired sample points. Here, the conditions for TE, b value, and TI at each collection sampling point are 0<TE1<TE2<TE3, 0≈b0<b1<b2, and 0<TI1, respectively.

Although not shown in FIG. 3, in the following embodiments, images collected by varying the TI using the IR method are also used in order to calculate T1 as a quantitative parameter. For example, in the following example, two images are used corresponding to two collected sample points, S _IR1 (TI1, TE1, b0) and S _IR2 (TI1, TE3, b0).

Normally, MR signals are observed as the sum (average) of two components within a voxel, and the ratio of each component is unknown, so they cannot be separated. However, if an acquisition method that suppresses one component is used, If possible, it is possible to separate the two components. For example, if only the water component within a voxel can be imaged, only the signal of the water component can be observed even if the ratio of the water component within the voxel is partial. Furthermore, by subtracting the water component signal from a signal containing both water and tissue components, a tissue component signal can also be obtained. That is, a signal in which a water component and a tissue component are mixed can be separated into two components.

The basic idea of the examples described below is that the water component is suppressed using several types of contrast-enhanced images such as PD-weighted images, T1-weighted images, T2-weighted images, and diffusion-weighted images as source images. Generate the image as a separated image, use the separated image to generate quantitative parameter maps such as T1, T2, PD, ADC, etc. before and after suppression of the water component, and further use these quantitative parameters in combination. By doing so, the signal of the water component in a composite contrast-enhanced image such as an SE image or a FLAIR image generated by SyMRI or the like is suppressed.

(Example 1)
First, Example 1 will be described. In this embodiment, a water component and a tissue component are separated by utilizing the difference in T2 between water and tissue. More specifically, in this example, the water component and the tissue component are separated by utilizing the fact that the T2 of water is sufficiently longer than the T2 of tissue.

That is, in this embodiment, the acquisition function 17a captures images of the same subject with different TE and TI, thereby distinguishing between images captured at a TE close to T2 of the tissue and images in which the water component is dominant in the signal. A plurality of images including images taken with a long TE such that

For example, the acquisition function 17a captures images of the subject by changing the TE using the SE method, thereby capturing the images of the object S _SE1 (TE1, b0), S _SE2 (TE2, b0), and S _SE3 (TE3, Collect three images corresponding to the three collected sample points of b0). As the pulse sequence of the SE method used here, for example, an SE type EPI (Echo Planar Imaging) sequence, an FSE (Fast Spin Echo) sequence, or a GE (Gradient echo) type SSFP is used. Further, as a filling method for k-space data, a normal Cartesian, a high-speed spiral, and a radial are used. The following explanation will be based on a sequence using parameters in the SE system, but in SSFP, which is often used in MRF, flip angle (FA) is added as a parameter, and by appropriately setting it including TR and TE. It can do the same thing as the SE system, which can emphasize water and tissue.

Furthermore, the acquisition function 17a captures images of the subject by changing TE and TI using the IR method, thereby obtaining the above-mentioned S _IR1 (TR1, TI1, TE1, b0) and S _IR2 (TR1, TI1, TE3, b0). ) to collect two images corresponding to the two collected sample points.

In this embodiment, the images corresponding to the points S _SE1 (TE1, b0), S _SE2 (TE2, b0), and S _SE3 (TE3, b0) are converted to S _SE1 (TR1, TE1, b0), respectively. Expressed as S _SE2 (TR1, TE2 _, b0) and S _SE3 (TR1, TE3, b0), corresponding to each point of S IR1 (TR1, TI1, TE1, b0) and S _IR2 (TR1, TI1, TE3, b0) The images are represented as S _IR1 (TR1, TI1, TE1, b0) and S _IR2 (TR1, TI1, TE3, b0), respectively.

Here, the shortest value that can be set in the MRI apparatus 100 is used for TE1. As a result, S _SE1 (TR1, TE1, b0) becomes a PD-enhanced image. Further, for TE2, a value longer than TE1 and close to T2 of the tissue is used. As a result, S _SE2 (TR1, TE2, b0) becomes a T2 weighted image. Furthermore, a value that is sufficiently longer than TE1 and TE2 is used for TE3. As a result, S _SE3 (TR1, TE3, b0) becomes an image in which the water component is dominant. Furthermore, a value close to T1 of the tissue is used for TI1. Note that the minimum value that can be set in the MRI apparatus 100 (for example, b0<1 [s/mm ² ]) is used for b0. Regarding b0, increasing b0 to a certain extent improves the effect of suppressing water components in each image, but conversely the SNR (Signal-to-Noise Ratio) of the signal strength decreases, so it is difficult to suppress water components based on T2. In the embodiment, it is desirable to set it to the minimum value.

For example, when the strength of the static magnetic field is 3T, the following values are used for the values of each imaging parameter.

TR1=4000 to 8000 [ms], TE1=25 [ms], TE2=100 [ms], TE3=300 to 500 [ms], TI1=1000 [ms], b0=0 [s/mm ² ], b1 =500 [s/mm ² ], b2=1000 [s/mm ² ]

In this embodiment, the separation function 17b selects a long image in which the water component is dominant from an image with a short TE among at least three images with different TEs among the plurality of images collected by the acquisition function 17a. By weighting and subtracting the TE images, an image in which the water component is suppressed is generated as a separated image.

For example, the separation function 17b weights S _SE1 (TR1, TE1, b0) to S _SE3 (TR1, TE3, b0) with a weighting coefficient α and differentiates the water component using the following (Equation 11). A suppressed image S _t (TR1, TE1, b0) is generated. In addition, the separation function 17b weights S _SE2 (TR1, TE2, b0) to S _SE3 (TR1, TE3, b0) with a weighting coefficient α and differentiates the water component using the following (Equation 12). A suppressed image S _t (TR1, TE2, b0) is generated. Note that S _t means that the water component is suppressed, resulting in an image of tissue components.

S _t (TR1, TE1, b0) = S _SE1 (TR1, TE1, b0) - α*S _SE3 (TR1, TE3, b0) ... (Formula 11)
S _t (TR1, TE2, b0) = S _SE1 (TR1, TE2, b0) - α*S _SE3 (TR1, TE3, b0) ... (Formula 12)

Here, if S _SE3 contains only a water component and the setting of the weighting coefficient α is appropriate, S _SE1 (TR1, TE1, b0) is a PD-weighted image and S _SE2 (TR1, TE2, b0) is a T2-weighted image. ), even if the ratio of the water component to the tissue component in the voxel where PVE occurs is unknown, only the tissue component contained in the voxel remains due to the difference.

Note that when it is required to separate the water component and the tissue component more accurately, the separation function 17b uses the assumed or measured T2 of the water to select a long TE in which the water component is dominant. An image in which the water component is suppressed may be generated as a separated image by generating an image of the water component corresponding to the short TE from the image and subtracting this image from the image of the short TE.

For example, the separation function 17b generates a water component image S _w (TR1, TE1, b0) from S _SE3 (TR1, TE3, b0) using the following (Equation 13). In addition, the separation function 17b subtracts S _w (TR1, TE1, b0) from S _SE1 (TR1, TE1, b0) using the following (Equation 14), thereby obtaining an image S _t ( TR1, TE1, b0) are generated.

S _w (TR1, TE1, b0) = S _SE3 (TR1, TE3, b0) * exp [(TE3-TE1)/T2 _w ] ... (Formula 13)
S _t (TR1, TE1, b0) = S _SE1 (TR1, TE1, b0) - S _w (TR1, TE1, b0) ... (Formula 14)

Further, the separation function 17b generates a water component image S _w (TR1, TE2, b0) from S _SE3 (TR1, TE3, b0) using the following (Equation 15). Furthermore, the separation function 17b subtracts S _w (TR1, TE2, b0) from S _SE2 (TR1, TE2, b0) using the following (Equation 16), thereby obtaining an image S _t ( TR1, TE2, b0) are generated.

S _w (TR1, TE2, b0) = S _SE3 (TR1, TE3, b0) *exp [(TE3-TE2)/T2 _w ] ... (Formula 15)
S _t (TR1, TE2, b0) = S _SE2 (TR1, TE2, b0) - S _w (TR1, TE2, b0) ... (Formula 16)

Here, T2 _W included in (Equation 13) and (Equation 15) is set to the assumed or measured T2 of water. For example, T2 _W is set to a water T2 value assumed using a theoretical value (for example, T2 _W =4000 [ms]). Alternatively, in T2 _W , the water component in each of S _SE1 (TR1, TE1, b0) and S _SE2 (TR1, TE2, b0) is close to 100% (e.g., the ventricle, etc.) according to the following (Equation 17). The value of T2 calculated from the signal strength of the voxel) is set.

T2 _w = (TE2-TE1)/ln[S _w (TR1, TE1, b0)/S _w (TR1, TE2, b0)] ... (Formula 17)

In this embodiment, the generation function 17c generates the entire T2 map, T1 map, and PD map including water components and tissue components using the plurality of images collected by the collection function 17a. Furthermore, the generation function 17c generates a T2 map, a T1 map, and a PD map in which water components are suppressed, using the separated images generated by the separation function 17b.

For example, the generation function 17c generates the entire T2 map T2 _orig using S _SE1 (TR1, TE1, b0) and S _SE2 (TR1, TE2, b0) according to the following (Equation 18). In addition, the generation function 17c generates a T2 map T2 _t in which the water component is suppressed using S _t (TR1, TE1, b0) and S _t (TR1, TE2, b0) according to the following (Equation 19). do. Note that T2 _t means that the T2 map is a tissue component due to suppression of water components.

T2 _orig = (TE2-TE1)/ln[S _SE1 (TR1, TE1, b0)/S _SE2 (TR1, TE2, b0)] ... (Formula 18)
T2 _t = (TE2-TE1)/ln[S _t (TR1, TE1, b0)/S _t (TR1, TE2, b0)] ... (Formula 19)

Furthermore, the generation function 17c generates the entire T1 map T1 _orig using S _SE1 (TR1, TE1, b0) and S _IR1 (TR1, TI1, TE1, b0) according to the following (Equation 20). In addition, the generation function 17c uses S _t (TR1, TE1, b0) and S _t (TR1, TI1, TE1, b0) according to the following (Equation 21) to create a T1 map T1 _t in which the water component is suppressed. generate. Note that T1 _t means that the water component is suppressed, resulting in a T1 map of tissue components.

T1 _orig = TI1/ln [{1-S _SE1 (TR1, TE1, b0)/S _IR1 (TR1, TI1, TE1, b0)}/2] ... (Formula 20)
T1 _t = TI1/ln [{1-S _t (TR1, TE1, b0)/S _t (TR1, TI1, TE1, b0)}/2] ... (Formula 21)

Here, S _t (TR1, TI1, TE1, b0) is a weighting coefficient of S _IR2 (TR1, TI1, TE3, b0) from S _IR1 (TR1, TI1, TE1, b0) using the following (Equation 22). It is obtained by weighting with α and calculating the difference.

S _t (TR1, TI1, TE1, b0) = S _IR1 (TR1, TI1, TE1, b0) - α*S _IR2 (TR1, TI1, TE3, b0) ... (Formula 22)

Furthermore, the generation function 17c generates the entire PD map PD _orig using S _SE1 (TR1, TE1, b0), T2 _orig , T1 _orig , and ADC _orig according to the following (Equation 23). Furthermore, the generation function 17c generates a PD map PD _t in which the water component is suppressed using S _t (TR1, TE1, b0), T2 _t , T1 _t and ADC _t according to the following (Equation 24). . Note that PD _t means that the PD map is a tissue component due to the suppression of water components.

PD _orig = S _SE1 (TR1, TE1, b0) / DecayT2 (TE1, T2 _orig ) / DecayT1 (TR1, T1 _orig ) / DecayADC (b0, ADC _orig ) ... (Formula 23)
PD _t = S _t (TR1, TE1, b0) / DecayT2 (TE1, T2t) / DecayT1 (TR1, T1t) / DecayADC (b0, ADCt) ... (Formula 24)

Here, DecayT2 (TE1, T2 _orig ) and DecayT2 (TE, T2 _t ) are each determined by the above-mentioned (Equation 5). Furthermore, DecayT1 (TE1, T1 _orig ) and DecayT1 (TE, T1 _t ) are each determined by the above-mentioned (Equation 3). Note that DecayADC (b0, ADC _orig ) and DecayADC (b0, ADC _t ) are obtained from the overall ADC map ADC _orig and the ADC map ADC _t in which the water component is suppressed, respectively, according to the above-mentioned (Equation 6). However, in this embodiment, since ADC _orig and ADC _t are not calculated, they are each set to 1.

FIG. 4 is a diagram illustrating an example of suppression of water components in the image and quantitative parameter map generated by the generation function 17c according to the first embodiment.

Here, the graph in (a) of FIG. 4 shows the signal strength versus V _w in the image S orig captured by _{the SE method at TE=25 [ms] and the image S t in} which the water component has been suppressed by the generation function 17c. It shows change. In addition, the graph in FIG. 4(b) shows the change in signal intensity with respect to _Vw in the image S _orig captured by the SE method with TE=100 [ms] and the image S _t in which the water component is suppressed by the generation function 17c. It shows.

Furthermore, the graph in FIG. 4C shows the change in signal intensity with respect to V _w in the entire PD map PD _orig generated by the generation function 17c and the PD map PD _t in which the water component is suppressed. Further, the graph in FIG. 4(d) shows the change in signal intensity with respect to V _w in the entire T1 map T1 _orig generated by the generation function 17c and the T1 map T1 _t in which the water component is suppressed. Further, the graph in FIG. 4(e) shows the change in signal intensity with respect to V _w in the entire T2 map T2 _orig generated by the generation function 17c and the T2 map T2 _t in which the water component is suppressed.

In addition, in (a) to (e) of FIG. 4, each image and It shows the change in signal strength in each map.

For example, as shown in FIGS. 4(a) to 4(e), in both images and maps in which the water component is suppressed, the weighting coefficient α is The larger the signal, the more the water component is suppressed and the signal approaches a state where only tissue components are present.

In this embodiment, the synthesis function 17d combines the entire T1 map T1 _orig , T2 map T2 _orig , and PD map PD _orig generated by the generation function 17c with the T1 maps T1 _t and T2 in which the water component is suppressed. The map T2 _t and the PD map PD _t are used in combination to generate a composite contrast-enhanced image of the SE method and a composite contrast-enhanced image of the FLAIR method corresponding to arbitrary imaging parameters.

Specifically, the synthesis function 17d substitutes TR and TE input by the operator into TR and TE of (Formula 1) representing a one-component model of the SE method, and further substitutes T1 of (Formula 1). , by substituting one of T1 _orig and T1 _t , one of T2 _orig and T2 _t , and one of PD _orig and PD _t for T2 and PD, respectively, to create an SE in which the water component is suppressed. The method generates a synthetic contrast-enhanced image. Note that since ADC _orig and ADC _t are not calculated in this embodiment, DecayADC (b, ADC) in (Formula 1) is set to 1.

In addition, the synthesis function 17d substitutes TR, TE, and TI input by the operator into TR, TE, and TI of (Formula 2) representing a one-component model of the IR method, and further substitutes the TR, TE, and TI input by the operator, respectively. The water component was suppressed by substituting either one of T1 _orig and T1 _t , one of T2 _orig and T2 _t , and one of PD _orig and PD _t for T1, T2 and PD, respectively. Generate a composite contrast-enhanced image using the FLAIR method. Note that in this embodiment, since ADC _orig and ADC _t are not calculated, 1 is substituted for DecayADC (b, ADC) in (Formula 2).

For example, the synthesis function 17d uses the entire T1 map T1 _orig , the T2 map T2 _t in which the water component is suppressed, and the PD map PD _t in which the water component is suppressed, to create the SE method in which the water component is suppressed. A composite contrast-enhanced image is generated.

Further, for example, the synthesis function 17d uses the entire T1 map T1 _orig , the entire PD map PD _orig , and the T2 map T2 _t in which the water component is suppressed to synthesize the FLAIR method in which the water component is suppressed. Generate a contrast-enhanced image.

FIG. 5 is a diagram illustrating an example of suppression of water components in a composite contrast-enhanced image generated by the composite function 17d according to the first embodiment.

Here, the graphs in FIGS. 5(a) and (e) show the change in signal intensity with respect to TE in the contrast-enhanced images of the SE method and the IR (FLAIR) method, respectively, collected by ideal real acquisition. There is. In addition, (b) and (f) of FIG. 5 show that the water component generated using the entire PD map PD _orig , the entire T1 map T1 _orig , and the entire T2 map T2 _orig is suppressed, respectively. It shows the change in signal intensity with respect to TE in a composite contrast-enhanced image of the conventional SE method and IR (FLAIR) method.

Moreover, (c) and (g) of FIG. 5 are the SE method generated using the entire PD map PD _orig , the entire T1 map T1 _orig , and the T2 map T2 _t in which the water component is suppressed, respectively. It shows changes in signal intensity with respect to TE in a composite contrast-enhanced image obtained by the and IR (FLAIR) methods. In addition, (d) and (h) of FIG. 5 are generated using the PD map PD _t in which the water component is suppressed, the entire T1 map T1 _orig , and the T2 map T2 _t in which the water component is suppressed, respectively. It shows the change in signal intensity with respect to TE in a composite contrast-enhanced image of the SE method and IR (FLAIR) method.

In addition, in (a) to (h) of FIG. 5, V _w is set to V _w =0 [%], V _w =25 [%], V _w =50 [%], V _w =75 [%], and , V _w =100 [%], and the change in signal intensity in each image is shown.

For example, as shown in FIG. 5(e), in a contrast-enhanced image obtained by the IR (FLAIR) method, the signal intensity is proportional to V _t =1−V _w . Furthermore, as shown in FIG. 5(f), in the composite contrast-enhanced image of the conventional IR (FLAIR) method, as the TE becomes longer, the signal intensity of each _Vw reverses and decreases. On the other hand, as shown in (g) and (h) of FIG. 5, in the composite contrast-enhanced image of the IR (FLAIR) method in which the water component is suppressed by the composition function 17d according to this embodiment, No inversion phenomenon of signal intensity occurs, and the signal intensity is approximately the same as that of the contrast-enhanced image obtained by the IR (FLAIR) method that was actually collected.

Furthermore, as shown in FIGS. 5(c) and 5(d), for example, in the SE method composite contrast-enhanced image in which the water component is suppressed by the compositing function 17d according to the present embodiment, FIG. Compared to the synthesized contrast-enhanced image of the IR (FLAIR) method in which the water component shown in the figure is suppressed, the order of the signal intensity with respect to V _w is the same as the contrast-enhanced image of the IR (FLAIR) method collected by actual collection. , and the SNR is further improved. Here, although there is noise propagation from the quantitative parameter map in synthetic contrast-enhanced images other than those actually acquired, the background noise is constant between images obtained by calculation, so the signal intensity can be regarded as relative SNR.

Here, we have explained an example where S _IR2 (TR1, TI1, TE3, b0) is collected, but S _IR2 (TR1, TI1, TE3, b0) is a T1 _t map in which the water component is suppressed. Since it is only used when generating the T1 t map, it is not necessary to collect it if the T1 _t map is unnecessary, as in the case where the contrast-enhanced image is FLAIR or SE as in this embodiment.

The first embodiment described above separates the water component and tissue component using only the difference in T2, and therefore can be easily applied to conventional methods such as the Fast SE method.

Furthermore, in the first embodiment described above, the collection function 17a may collect images of the subject using a multi-echo method. In actual collection using the IR (FLAIR) method, since the tissue is not sufficiently saturated with TI, the longitudinal magnetization of the tissue cannot be made the same as in the SE method, so water components and tissue components cannot be separated by difference. Separate acquisitions are required depending on the presence or absence of pulses, which takes time. On the other hand, in actual acquisition using the SE method, if a multi-echo method using the same excitation is used, misregistration due to extension of imaging time or body movement hardly occurs.

Furthermore, in the first embodiment described above, the acquisition function 17a may apply a gradient magnetic field for Gradient Moment Nulling (GMN) (gradient magnetic field for flow rephase) to capture an image of the subject. This makes it possible to suppress artifacts and signal degradation due to movement of CSF and blood flow.

In addition, in the above-mentioned Example 1, the water at each TE is separated from the TE where the tissue component is approximately zero, using a simple separation method based on a weighted difference using the weighting coefficient α and the assumed or measured water T2. A highly accurate separation method of estimating and subtracting components can be appropriately selected and used.

In addition, in Example 1 described above, when it is required to completely reduce the water component to zero, by setting a threshold value on the image of the water tissue, it is possible to distinguish between voxels where the water component is 100% and other tissue parts. voxels may be separated and a quantitative parameter map may be generated separately from each voxel. In that case, voxels in which the two components coexist will overlap, but such voxels may be color-coded and displayed in an overlapping manner. In that case, the mixed voxels are displayed in a color according to the mixing ratio. For example, in the case of PD, if the tissue component is red and the water component is light blue, if V _w =50%, the color will be purple. Alternatively, in order to avoid overlapping voxels, the threshold value of the water component signal may be set to 50%, and the quantitative parameter map of the component with the larger amount may be obtained and synthesized.

In addition, in Example 1 described above, a method for suppressing the water component by subtracting long TE images in which the water component is dominant was explained. If images are not required, collect images using the IR (FLAIR) method from the beginning, that is, with a combination of TR and TI that makes the longitudinal magnetization of water zero, with two or more different TEs, and then combine those images. A T2 map may be generated using the T2 map, and a composite contrast-enhanced image may be generated. In this case, if you make TR long enough to generate an image with TE=0, you can also obtain a PD map with suppressed water components, and in addition to the calculated FLAIR image, you can also obtain a calculated PD map with suppressed water components. SE images are also obtained.

(Example 2)
Next, Example 2 will be explained. This embodiment uses the difference in ADC between water and tissue to separate water components and tissue components contained in MR signals. More specifically, this embodiment uses the fact that the ADC of water is larger than the ADC of tissue to separate the water component and tissue component contained in the MR signal.

That is, in this embodiment, the collection function 17a collects an image in which the water component is suppressed by capturing an image of the subject using a non-zero b value by diffusion weighted imaging that captures DWI.

In diffusion-weighted imaging, a follow-encoding gradient magnetic field called MPG (Motion Probing gradient) is applied during data acquisition, and this MPG creates a contrast in the intensity of the MR signal depending on the difference in ADC in each tissue in the living body. , DWI with emphasized differences in ADC is imaged. Here, it is known that the ADC of water is about three times larger than the ADC of tissue. Specifically, water has an ADC of 3*10 ^-3 [mm ² /s], normal tissue has an ADC of 1*10 -3 [mm 2 /s], and tumors etc. have an ADC of 1*10 ^-3 [mm ² /s]. becomes smaller. In diffusion weighted imaging, the larger the b value representing phase dispersion due to diffusion, the more suppressed water components are and the more dominant tissue components can be obtained.

FIG. 6 is a diagram illustrating an example of suppression of a water component by a b value in an image captured by diffusion weighted imaging according to Example 2.

Here, the graph in (a) of FIG. 6 shows the change in signal intensity with respect to the b value in a PD-enhanced image captured with TR=4000 [ms] and TE=35 [ms]. Further, the graph in FIG. 6(b) shows the change in signal intensity with respect to the b value in a PD emphasized image captured with TR=4000 [ms] and TE=100 [ms].

For example, as shown in FIGS. 6A and 6B, as the b value increases, the water component is suppressed, and the ratio S _t /S _orig of the tissue component to the whole increases. For example, if b is set to about 300 [s/mm ² ], the water component can be considerably suppressed.

For example, as shown in FIG. 6(a), in an image with TE=35 [ms], the signal strength S _w of the water component is approximately 55% at b=250 [s/mm ² ], and approximately 55% at b=500 [s/mm 2 ]; s/mm ² ] to about 25%. For example, as shown in FIG. 6(b), in an image with TE=100 [ms], the signal strength S _w of the water component is about 55% at b=250 [s/mm ² ], and b= It is reduced to about 40% at 500 [s/mm ² ]. That is, if TE is shortened, the relative proportion of the water component at b=0 becomes smaller, so the effect of suppressing the water component becomes higher.

For example, the acquisition function 17a captures an image of the subject using a b value that is not 0 while changing the TE by diffusion weighted imaging of the SE method, thereby obtaining the S _DWI1 (TE1, b1) and S _DWI2 shown in FIG. 3. Two images corresponding to the two collected sample points (TE2, b1) are collected.

In this example, the images corresponding to each point of S _DWI1 (TE1, b1) and S _DWI2 (TE2, b1) are converted into S _DWI1 (TR1, TE1, b1) and S _DWI2 (TR1, TE2, b1), respectively. ).

Here, the shortest value that can be set in the MRI apparatus 100 is used for TE1. Further, a value longer than TE1 is used for TE2. A value larger than 0 is used for b1. As a result, S _DWI1 (TR1, TE1, b1) and S _DWI2 (TR1, TE2, b1) each become images in which the water component is suppressed compared to when the images are captured with the b value set to 0.

In this embodiment, the generation function 17c uses S _DWI1 (TR1, TE1, b1) and S _DWI2 (TR1, TE2, b1) according to the following (Equation 25) to generate T2 in which the water component is suppressed. Generate map T2 _t .

T2 _t = (TE2-TE1)/ln[S _DWI1 (TR1, TE1, b1)/S _DWI2 (TR1, TE2, b1)] ... (Formula 25)

In the second embodiment described above, since a difference is also used, the image captured by diffusion weighted imaging may be an isotropic DWI that eliminates the influence of the directionality of diffusion.

Furthermore, in the above-mentioned Example 2, when diffusion weighted imaging is used together with b=100 to 200 during acquisition, the same water component suppression effect can be obtained even with a short TE compared to Example 1 (b=0). It will be done.

In addition, in the method using the b value of diffusion weighted imaging as in Example 2 described above, since a gradient magnetic field for MPG is applied to emphasize movement up to TE, the larger the b value, the more Although the minimum TE becomes longer and the SNR of the quantitative parameter map decreases, it is possible to obtain the T2 map as long as the TEs are even slightly apart.

In addition, an advantage of the method using the b value of diffusion weighted imaging as in Example 2 described above is that an image dominated by tissue components is directly obtained, rather than an image of water components. Compared to the case where only the difference in TE is used, the SNR deteriorates less due to the difference.

Furthermore, if only the method of Example 1 is used, motion artifacts due to the movement of the CSF may enter the parenchyma, causing errors in subsequent quantitative parameter maps, etc., but when using diffusion weighted imaging as in Example 2, Since the CSF signal is suppressed on the original image, it is possible to suppress motion artifacts.

(Example 3)
Next, Example 3 will be explained. This example utilizes the difference in T2 between water and tissue as in Example 1, and utilizes the difference in ADC between water and tissue as in Example 2 to generate an MR signal. It separates the water component and tissue component contained.

That is, in this embodiment, the collection function 17a captures images of the same subject using a b value that is not 0 while changing the TE and TI, so that the water component is detected at a TE close to T2 of the tissue. A plurality of images including an image captured using a b value such that the water component is suppressed, and an image captured using a b value such that the water component is suppressed with a long TE such that the water component is dominant. Collect.

For example, the acquisition function 17a captures an image of the subject using a b value that is not 0 while changing the TE by diffusion weighted imaging _of the SE method _. Three images are collected corresponding to the three collected sample points: (TE2, b1) and S _DWI3 (TE3, b1).

In this embodiment, the images corresponding to the points S _DWI1 (TE1, b1), S _DWI2 (TE2, b1), and S _DWI3 (TE3, b1) are converted to S _DWI1 (TR1, TE1, b1), respectively. They are expressed as S _DWI2 (TR1, TE2, b1) and S _DWI3 (TR1, TE3, b1).

Here, for TE1, the shortest value that can be set in the MRI apparatus 100 is used, for TE2, a value longer than TE1 and close to T2 of the tissue is used, and for TE3, a value that is sufficiently longer than TE1 and TE2 is used. value is used. As a result, S _DWI3 (TR1, TE3, b1) becomes an image in which the water component is dominant. A value larger than 0 is used for b1. As a result, S _DWI1 (TR1, TE1, b1) and S _DWI2 (TR1, TE2, b1) each become images in which the water component is suppressed compared to when the images are captured with the b value set to 0.

In this embodiment, the separation function 17b uses S _DWI1 (TR1, TE1, b1) and S _DWI3 (TR1, TE3, b1) according to the following (Equation 26) to create an image in which the water component is suppressed. Generate S _t (TR1, TE1, b1). Furthermore, the separation function 17b uses S _DWI2 (TR1, TE2, b1) and S _DWI3 (TR1, TE3, b1) to generate an image S _t (TR1, TE2,b1) is generated.

S _t (TR1, TE1, b1) = S _DWI1 (TR1, TE1, b1) - α*S _DWI3 (TR1, TE3, b1) ... (Formula 26)
S _t (TR1, TE2, b1) = S _DWI1 (TR1, TE2, b1) - α*S _DWI3 (TR1, TE3, b1) ... (Formula 27)

Note that when it is required to separate the water component and tissue component more accurately, the separation function 17b converts the water component image S from S _DWI3 (TR1, TE3, b1) using the following (Equation 28). Generate _w (TR1, TE1, b1). In addition, the separation function 17b subtracts S _w (TR1, TE1, b1) from S _DWI1 (TR1, TE1, b1) using the following (Equation 29), so that the image S _t ( TR1, TE1, b1) are generated.

S _w (TR1, TE1, b1) = S _DWI3 (TR1, TE3, b1) * exp [(TE3-TE1)/T2 _w ] ... (Formula 28)
S _t (TR1, TE1, b1) = S _DWI1 (TR1, TE1, b1) - S _w (TR1, TE1, b1) ... (Formula 29)

Further, the separation function 17b generates a water component image S _w (TR1, TE2, b1) from S _DWI3 (TR1, TE3, b1) using the following (Equation 30). In addition, the separation function 17b subtracts S _w (TR1, TE2, b1) from S _DWI2 (TR1, TE2, b1) using the following (Equation 31), thereby obtaining an image S _t ( TR1, TE2, b1) are generated.

S _w (TR1, TE2, b1) = S _DWI3 (TR1, TE3, b1) * exp [(TE3-TE2)/T2 _w ] ... (Formula 30)
S _t (TR1, TE2, b1) = S _DWI2 (TR1, TE2, b1) - S _w (TR1, TE2, b1) ... (Formula 31)

Here, T2 _W included in (Equation 28) and (Equation 30) is set to the assumed or measured T2 of water, as in the first embodiment.

In this embodiment, the generation function 17c uses S _t (TR1, TE1, b1) and S _t (TR1, TE2, b1) according to the following (Equation 32) to generate T2 in which the water component is suppressed. Generate map T2 _t .

T2 _t = (TE2-TE1)/ln[S _t (TR1, TE1, b1)/S _t (TR1, TE2, b1)] ... (Formula 32)

The method of suppressing the water component using diffusion weighted imaging as in the third embodiment described above can also be combined with the method of suppressing the water component using the difference in T2 as in the first embodiment. In other words, with the method using diffusion weighted imaging, it is also possible to use data collected using both the b value and TE, and each has a smaller value than when using the b value or TE alone. However, the effect of suppressing water components can be improved. For example, if the original image is collected at three stages of TE while using b1 (b1 to 200 [s/mm ² ]) of a certain size, and the image with a long TE is subtracted from the image with a short TE, Example 1 Since the CSF signal itself can be suppressed compared to the case where b0≈0 is used, the CSF suppression effect becomes larger and motion artifacts are also reduced.Alternatively, if the same suppression effect is acceptable, the water component is It is also possible to obtain the effect that the TE (TE3) of a dominant image can be shortened.

(Example 4)
Next, Example 4 will be described. In this embodiment, an ADC map in which the water component is suppressed is generated by capturing images of the same subject with different TE and b values.

That is, in this embodiment, the collection function 17a uses diffusion weighted imaging to image DWI to identify a first TE close to T2 of the tissue, a long second TE in which the water component is dominant, and a first TE. and a second b value larger than the first b value such that the water component is suppressed compared to the first b value. A first image taken with the TE and the first b value, a second image taken with the second TE and the first b value, and a third image taken with the first TE and the second b value. Collect images of.

For example, the acquisition function 17a captures images of the subject by changing the TE and b values by diffusion weighted imaging using the SE method, thereby capturing images of the subject such as S _SE2 (TE2, b0) and S _SE3 (TE3, b0) shown in FIG. ) and S _DWI4 (TE2, b2).

In this example, the images corresponding to the points S _SE2 (TE2, b0), S _SE3 (TE3, b0), and S _DWI4 (TE2, b2) are converted to S _SE2 (TR1, TE2, b0), respectively. They are expressed as S _SE3 (TR1, TE3, b0) and S _DWI4 (TR1, TE2, b2).

Here, a value close to T2 of the tissue is used for TE2, and a value that is sufficiently longer than TE2 is used for TE3. As a result, S _SE3 (TR1, TE3, b0) becomes an image in which the water component is dominant. A value larger than b0 is used for b2. As a result, S _DWI4 (TR1, TE2, b2) becomes an image in which the water component is suppressed compared to S _SE3 (TR1, TE3, b0) and S _SE2 (TR1, TE2, b0). Note that the minimum value that can be set in the MRI apparatus 100 (for example, b0<1 [s/mm ² ]) is used for b0. Regarding b0, if it is increased to a certain extent, the effect of suppressing the water component in each image will improve, but on the contrary, the signal strength and SNR of ADC will decrease. Therefore, in this example, in which the water component is suppressed based on T2, It is preferable to use a value.

For example, by setting TE3 to about 300 [ms], the water component becomes dominant in S _SE3 (TR1, TE3, b0). Further, by setting b2>1000 [s/mm ² ], the water component in S _DWI4 (TR1, TE2, b2) can be made approximately zero.

In this embodiment, the separation function 17b weights and subtracts the second image from the first image collected by the collection function 17a, so that the water component corresponding to the first b value is suppressed. The separated image is generated as a separated image.

That is, in this embodiment, since the water component is approximately zero in S _DWI4 (TR1, TE2, b2), the water component is suppressed only in S _SE2 (TR1, TE2, b0).

For example, the separation function 17b weights S _SE2 (TR1, TE2, b0) to S _SE3 (TR1, TE3, b0) with a weighting coefficient α and differentiates the water component using the following (Equation 33). A suppressed image S _t (TR1, TE2, b0) is generated.

S _t (TR1, TE2, b0) = S _SE2 (TR1, TE2, b0) - α*S _SE3 (TR1, TE3, b0) ... (Formula 33)

In this embodiment, the generation function 17c generates the entire ADC map including water components and tissue components using the plurality of images collected by the collection function 17a. Furthermore, the generation function 17c generates an ADC map in which the water component is suppressed, using the third image collected by the collection function 17a and the separated image generated by the separation function 17b.

For example, the generation function 17c generates the entire ADC map ADC _orig using S _SE2 (TR1, TE2, b2) and S _DWI4 (TR1, TE2, b2) according to the following (Equation 34). In addition, the generation function 17c uses S _DWI4 (TR1, TE2, b2) and S _t (TR1, TE1, b0) generated by the separation function 17b to generate a water component according to the following (Equation 35). Generate a suppressed ADC map ADC _t . Note that ADC _t means that the ADC map is a tissue component due to the suppression of water components.

ADC _orig = ln [S _SE2 (TR1, TE2, b0) / S _DWI4 (TR1, TE2, b2)] / (b2 - b0) ... (Formula 34)
ADC _t = /ln [S _t (TR1, TE2, b0) / S _DWI4 (TR1, TE2, b2)] / (b2 - b0) ... (Formula 35)

In the fourth embodiment described above, since the image of b2 is not subtracted from the water component image, there is no reduction in SNR. Furthermore, since the b0 image has a difference from the water component image, the SNR decreases; however, by increasing the average of only the b0 image, the SNR decrease can be minimized. Further, even when acquiring b values in multiple stages, if b2 is set to a value at which the water signal is approximately zero (b2>1000 [s/mm ² ]), no difference is required.

(Example 5)
Next, Example 5 will be described. In this example, as in Example 4, an ADC map in which the water component is suppressed is generated by capturing images of the same subject with different TE and b values, but for each b value. By generating and subtracting water component images, the accuracy of water component suppression is further improved.

That is, in this example, the collection function 17a uses diffusion weighted imaging to image DWI, as in Example 4, to collect the first TE close to T2 of the tissue and the long TE in which the water component is predominant. An image of the subject is captured using a TE of 2, a first b value, and a second b value larger than the first b value such that water components are suppressed compared to the first b value. By doing so, the first image captured with the first TE and the first b value, the second image captured with the second TE and the first b value, and the first image captured with the first TE and the second b value are A third image captured using the b value is collected.

Further, as in the fourth embodiment, the separation function 17b weights and subtracts the second image from the first image collected by the collection function 17a, so that the water component corresponding to the first b value is obtained. Generate the suppressed image as a separated image.

In this embodiment, the generation function 17c generates an image of the water component corresponding to the second b value from the second image collected by the collection function 17a using the assumed or measured water diffusion coefficient. By subtracting this image from the third image, an image in which the water component corresponding to the second b value is suppressed is generated, and this image and the separated image generated by the separation function 17b are separated. to generate an ADC map in which the water component is suppressed.

For example, the generation function 17c generates the water component image S _w (TR1, TE2, b0) by weighting S _SE3 (TR1, TE3, b0) with the weighting coefficient α according to the following (Equation 36), and then , a water component image S _w (TR1, TE2, b2) is generated from S _w (TR1, TE2, b0) using the following (Equation 37).

S _w (TR1, TE2, b0) = α*S _SE3 (TR1, TE3, b0) ... (Formula 36)
S _w (TR1, TE2, b2) = S _w (TR1, TE2, b0) * exp [- (b2 - b0) * ADC _w ] ... (Formula 37)

Here, ADC _W included in (Equation 36) is set to the assumed or measured ADC of water. For example, ADC _W is set to a water ADC value assumed using a theoretical value (for example, ADC _W =3.0*10 ⁻³ [mm ² /s]). Alternatively, for ADC _W , voxels ( _{e.g. ,} ventricles, The ADC value calculated from the signal strength S of the voxel) is set.

ADC _w = ln [S _w (TR1, TE1, b0) / S _w (TR1, TE2, b2)] / (b2 - b0) ... (Formula 38)

Thereafter, the generation function 17c subtracts S _w (TR1, TE2, b2) from S _DWI4 (TR1, TE2, b2) using the following (Equation 39), thereby creating an image S _t ( TR1, TE2, b2) are generated.

S _t (TR1, TE2, b2) = S _DWI4 (TR1, TE2, b2) - S _w (TR1, TE2, b2) ... (Formula 39)

Then, the generation function 17c generates the entire ADC map ADC _orig using S _SE2 (TR1, TE2, b2) and S _DWI4 (TR1, TE2, b2) according to the following (Equation 40). In addition, the generation function 17c uses S _t (TR1, TE2, b2) and S _t (TR1, TE2, b0) generated by the separation function 17b to generate a water component using (Equation 41) below. Generate a suppressed ADC map ADC _t .

ADC _orig = ln[S _SE2 (TR1, TE2, b0)/S _DWI4 (TR1, TE2, b2)]/(b2-b0) ... (Formula 40)
ADC _t = ln [S _t (TR1, TE2, b0) / S _t (TR1, TE2, b2)] / (b2 - b0) ... (Formula 41)

Note that here, ADC _t was determined by a two-point method using b0 and b2, but the present embodiment is not limited to this. For example, n = 0, 1, 2, . . . N ADC _t may be determined by least squares approximation using an N-point method using a plurality of bns.

In that case, the collection function 17a changes the b value to n=0, 1, 2,...N to obtain S _SE2 (TR1, TE2, b0) and S _SE3 (TR1, TE3, b0). and multiple S _DWI (TR1, TE2, bn) are collected.

In addition, after the generation function 17c generates the water component image S _w (TR1, TE2, b0) by weighting S _SE3 (TR1, TE3, b0) with the weighting coefficient α according to the following (Equation 42), , a water component image S _w (TR1, TE2, bn) is generated from S _w (TR1, TE2, b0) for each bn using the following (Equation 43).

S _w (TR1, TE2, b0) = α*S _SE3 (TR1, TE3, b0) ... (Formula 42)
S _w (TR1, TE2, bn) = S _w (TR1, TE2, b0) * exp [- (bn - b0) * ADC _w ] ... (Formula 43)

Here, ADC _W included in (Equation 41) is set to the assumed or measured ADC of water, as in the above example.

After that, the generation function 17c subtracts S _w (TR1, TE2, bn) from S _DWI (TR1, TE2, bn) for each bn using the following (Equation 44), thereby suppressing the water component. An image S _t (TR1, TE2, bn) is generated.

S _t (TR1, TE2, bn) = S _DWI (TR1, TE2, bn) - S _w (TR1, TE2, bn) ... (Formula 44)

Then, the generation function 17c performs least squares approximation using S _SE2 (TR1, TE2, b2) and S _DWI (TR1, TE2, bn) collected for each bn, according to the following (Equation 45). Thus, the entire ADC map ADC _orig is generated. Furthermore, the generation function 17c uses the following (Equation 46) to calculate S _t (TR1, TE2, bn) generated for each bn and S _t (TR1, TE2, b0) generated by the separation function 17b. By performing approximation using the least squares method, an ADC map ADC _t in which the water component is suppressed is generated.

ADC _orig = ln[S _SE2 (TR1, TE2, b0)/S _DWI4 (TR1, TE2, b2)]/(b2-b0) ... (Formula 45)
ADC _t = ln [S _t (TR1, TE2, b0) / S _t (TR1, TE2, bn)] / (bn - b0) ... (Formula 46)

In Example 5 described above, unlike Example 4, it is possible to suppress the water component regardless of the magnitude of the b value.

In addition, as long as the b value is small to a certain extent, the water component is suppressed by the method of Example 5, and when the b value becomes large to a certain extent, the water component can be ignored, so the same method as in Example 4 is used. , it is also possible to use the original signal strength without performing the difference. In that case, for example, b=1000 [s/mm ² ] is set as the boundary for switching the method.

In addition, DWI of the brain is often hampered by the quantitative nature of CSF, etc., but if accuracy is desired for the entire image, including the CSF part, voxels that are 100% water can be used. It may be excluded in advance and combined with the quantitative parameter of the water component generated without separately suppressing the water component. In addition, the parameters of diffusion weighted imaging are not limited to ADC, but also FA (Fractional Anisotropy) map, which is a DTI (Diffusion Weighted Image) acquired by changing MPG directions in multiple directions, and color encoded DTI map (RGB color by diffusion direction). If a DWI signal in which the water component is suppressed is applied to display DTI map) and tractography, it is expected that accuracy will be improved in areas where PVE due to CSF is large, such as the cortex.

In addition, in conventional computed DWI (cDWI), diffusion-weighted imaging is used to generate images with small b values to large b values, thereby improving the CNR (Contrast-to-Noise Ratio) between tumors and normal tissues in the abdomen, etc. However, according to the method described above, it is possible to improve quantitative parameters up to ADC and FA.

Although Examples 1 to 5 have been described above, it is also possible to implement the respective Examples in combination as appropriate.

For example, in the above-mentioned Examples 2 and 3, an example of generating the T2 map T2 _t in which the water component is suppressed is explained, and in Examples 4 and 5, the entire ADC map and the case in which the water component is suppressed are explained. Although an example of generating an ADC map has been described, in any embodiment, the generation function 17c can generate the entire image in the same manner as in the first embodiment by additionally collecting necessary images by the collection function 17a. It is possible to generate T1 maps, T2 maps, and PD maps in which water components and water components are suppressed. Also in Example 1, by additionally collecting necessary images by the collection function 17a, the generation function 17c generates an ADC map in which the overall and water components are suppressed in the same manner as in Examples 4 and 5. It is possible to generate

In any of the embodiments, the generation function 17c generates a necessary quantitative parameter map, so that the synthesis function 17d can create various synthetic contrast enhancement maps as in the first embodiment.

(Example 6)
Next, Example 6 will be described. This example uses an image captured by diffusion weighted imaging to further resolve moving blood flow components in addition to stationary water components and tissue components.

For example, in an image taken with a long TE, both moving and stationary water components cannot be distinguished, so both water components are included, but the blood flow component has a different T2 than the water component (the blood flow component has a different T2). is not included because its T2 is shorter than that of the water component). Therefore, if an image taken with a long TE is subtracted from an image taken with a short TE, blood flow components and tissue components will remain. Furthermore, because the ADC of the blood flow component becomes larger due to movement than that of the water component that is stationary, the water component and the blood flow component can be separated based on the difference in b value in diffusion weighted imaging. Based on these facts, it is assumed that a unit voxel contains three components: a water component, a blood flow component, and a tissue component, and it is possible to separate each component by using the difference in T2 and ADC between each component. It is.

Here, in this embodiment, an MR signal model in which three components, a water component, a blood flow component, and a tissue component are included in a unit voxel, is called a three-comparment model.

For example, for a three-component model, if the volume of the water component in a unit voxel is Vw, the volume of the blood flow component is _Vb , and the volume of the tissue component is _Vt , then the volume V of the unit voxel is calculated using the following formula (47). It is expressed as Further, if PD of the water component is PDw, PD of the bloodstream component is PDb, and PD of the tissue component is _PDt , then the average proton density PD within a unit voxel is expressed by the following (Equation 48).

V=V _w +V _b +V _t =1 (Formula 47)
PD=V _w *PD _w +V _b *PD _b +V _t *PD _t ... (Formula 48)

Then, the MR signal of the water component is S _w , the MR signal of the blood flow component is S _b , and the MR signal of the tissue component is S _t
Then, the MR signal of each component is expressed by the following (Formula 49), (Formula 50), and (Formula 51), respectively.

S _w =V _w *PD _w *exp(-b*ADC _w ) ... (Formula 49)
S _b =V _b *PD _b *exp(-b*ADC _b ) ... (Formula 50)
S _t =V _t *PD _t *exp(-b*ADC _t ) ... (Formula 51)

Here, ADC _w is the ADC of water, ADC _b is the ADC of blood flow, and ADC _t is the ADC of tissue, and the magnitude relationship of each is ADC _b > ADC _w > ADC _t .

For example, as in Example 1, S _t +S _b and S _w can be separated using the difference in T2. Furthermore, if images are collected by applying GMN, the component due to the movement of the CSF is considered to be included in S _w , so S _b is the blood flow where T1 and T2 are approximately between the water component and the tissue component. It can be considered as an ingredient. Therefore, since the image after the water component is suppressed becomes S−S _w =S _t +S _b , it can be considered as a problem of separating two components, St and S _b .

On the other hand, for example, by expanding the separation of two components using the difference in ADC as in Example 2 to three components, it is also possible to separate the three components in the same way. However, when separating three components based only on the difference in ADC, at least four levels of b values (b0, b1, b2, b3) are used, which requires a large b value, so the TE must also be extended. In addition to decreasing the SNR, analytical accuracy is also considered to be insufficient.

FIG. 7 is a diagram showing the relationship between signal intensity and b value in DWI collected by diffusion weighted imaging according to Example 6.

Here, the graph in (a) of FIG. 7 shows the DWI signal strength and b value based on the three-component model for each of the water component signal S _w , blood flow component signal S _b , and tissue component signal S _t . It shows the relationship between In addition, the graph in FIG. 7(b) shows a long TE with b=0 and TE=300 [ms] for the water component signal S _w , the blood flow component signal S _b , and the tissue component signal S _t , respectively. It shows the relationship between the DWI signal strength and the b value after subtracting the images captured in .

Note that the graphs in FIGS. 7A and 7B show the relationship between the DWI signal strength and the b value when each parameter is set as follows.

PD _w =1, PD _b =1, PD _t =1, V _w =0.2, V _b =0.2, V _t =0.6, T1 _w =4000 [ms], T1 _b =1600 [ms] ], T1 _t =1000 [ms], T2 _w =1910 [ms], T2 _b =200 [ms], T2 _t =100 [ms], ADC _w =3.0*10 ^-3 [mm ² /s] , ADC _b = 1.0*10 ^-2 [mm ² /s], ADC _t = 8.0*10 ^-4 [mm ² /s], TR = 10000 [ms], TE = 80 [ms]

For example, as shown in FIG. 6(a), the b values are b0 = 0 [s/mm ² ], b1 = 300 [s/mm ² ], b2 = 1000 [s/mm ² ], b3 = 2000 [ s/mm ² ], S _w ≒ 0 when b > 1000 [s/mm ² ], so the three components can be calculated without subtracting images captured with a long TE. It is possible to separate. However, in this case, a large b value is required in order to use four levels of b values, and the shortest TE becomes longer accordingly, so in addition to the SNR decreasing due to the longer TE, analytically It is also thought that the accuracy will be insufficient.

On the other hand, for example, as shown in FIG. 6(b), when images captured with a long TE are subtracted, S _w ≒0 when b > 300 [s/mm ² ], so b0 It is possible to separate the three components even with three levels of b values such as =0 [s/mm ² ], b1 = 300 [s/mm ² ], and b2 = 1000 [s/mm ² ]. That is, it is possible to separate the three components by omitting b3, which is the largest among the four levels of b values shown in FIG. 6(a). Therefore, in this case, the required maximum b value can be reduced, and as a result, the TE can be shortened and the SNR can be improved.

For this reason, in this example, after separating the water component using the difference in T2, the blood flow component and the tissue component are separated using the difference in ADC. Separate three components: component and tissue component.

That is, in this embodiment, the acquisition function 17a captures images of the subject using three different b values by diffusion weighted imaging that captures DWI.

Specifically, the acquisition function 17a collects a first TE close to T2 of the tissue, a second TE longer than the first echo time such that the water component is dominant, and a water component and a tissue in the signal. A first b value such that all components and blood flow components are mixed, and a second b value that is larger than the first b value such that tissue components and water components are dominant in the signal compared to the first b value. capturing an image of the subject using a b value of 2 and a third b value that is larger than the second b value such that a tissue component becomes dominant in the signal compared to the second b value; , a first image captured using the first TE and the first b value, a second image captured using the second TE and the first b value, and a first image captured using the first TE and the second b value. A third image taken with the second TE and the second b value, and a fifth image taken with the second TE and the third b value are collected.

For example, the acquisition function 17a captures images of the subject by _varying the _TE and b values using diffusion weighted imaging using the SE method. ), S _SE3 (TE3, b0), S _DWI1 (TE1, b1), S _DWI2 (TE2, b1), S _DWI3 (TE3, b1), and S _DWI4 (TE2, b2) correspond to seven collection sample points. Collect 7 images.

Here, the shortest value that can be set in the MRI apparatus 100 is used for TE1. As a result, S _SE1 (TE1, b0) becomes a PD-enhanced image. Further, for TE2, a value longer than TE1 and close to T2 of the tissue is used. As a result, S _SE2 (TE2, b0) becomes a T2-weighted image. Furthermore, a value that is sufficiently longer than TE1 and TE2 is used for TE3. As a result, S _SE3 (TE3, b0) becomes an image in which the water component is dominant. Furthermore, a value is used for b0 such that all of the water component, tissue component, and blood flow component are mixed in the signal. Furthermore, a value larger than b0 is used for b1, so that tissue components and water components become dominant in the signal compared to b0. Furthermore, a value larger than b1 is used for b2 so that the tissue component becomes dominant in the signal compared to b1. Note that TR is constant when collecting each image, and is set to be sufficiently large especially when T1 of the water component is desired to be accurately determined (for example, TR>5*T1 _w ).

As a result, the water component exists only in the images of b=b0 and b=b1, and can be ignored in the image of b=b2. Further, the blood flow component exists only in the image of b=b0, and can be ignored in the images of b=b1 and b=b2. Moreover, the tissue component will exist in all images of b=b0, b=b1, and b=b2.

For example, the following values are used for the values of each imaging parameter.

TE1 = shortest value, TE2 = 80 [ms], TE3 = 300 [ms], b0 = 0 [s/mm ² ], b1 = 300 [s/mm ² ], b2 = 1000 [s/mm ² ]

In this embodiment, the separation function 17b analyzes the images taken at each of the three different b values collected by the collection function 17a, so that in addition to the stationary water component and tissue component, the moving A separated image is generated in which three components are separated, including the blood flow component.

Specifically, the separation function 17b weights and subtracts the second image from the first image collected by the collection function 17a, so that water components are suppressed and only tissue components and blood flow components remain. The first separated image is generated as the first separated image, and by weighting and subtracting the third to fourth images, an image in which the water component is suppressed and only the tissue component remains is created as the second separated image. Then, using the first separated image and the second separated image, a third separated image in which water components and tissue components are suppressed and only blood flow components remain is generated.

For example, the separation function 17b weights S _SE1 (TE1, b0) to S SE3 (TE3, b0) with a weighting coefficient α and calculates the difference between S SE1 (TE1, b0) and S _SE3 (TE3, b0) using the following (Equation 52). Generate S _wsup1 (TE1, b0). That is, an image S _t (TE1, b0)+S _b (TE1, b0) in which only tissue components and blood flow components remain is obtained.

S _wsup1 (TE1, b0) = S _SE1 (TE1, b0) - α＊S _SE3 (TE3, b0) ... (Formula 52)

Furthermore, the separation function 17b weights and differentiates S _SE2 (TE2, b0) to S _SE3 (TE3, b0) with a weighting coefficient α using the following (Equation 53), thereby producing an image in which the water component is suppressed. Generate S _wsup2 (TE2, b0). That is, an image S _t (TE2, b0)+S _b (TE2, b0) in which only tissue components and blood flow components remain is obtained.

S _wsup2 (TE2, b0) = S _SE2 (TE2, b0) - α＊S _SE3 (TE3, b0) ... (Formula 53)

In addition, the separation function 17b weights and differentiates S _DWI1 (TE1, b1) to S _DWI3 (TE3, b1) with a weighting coefficient α according to the following (Equation 54), thereby producing an image in which the water component is suppressed. Generate S _wsup1 (TE2, b1). That is, since S _b (TE1, b1)≈0, an image S _t (TE1, b1) in which only tissue components remain is obtained.

S _wsup1 (TE1, b1) = S _DWI1 (TE1, b1) - α＊S _DWI3 (TE3, b1) ... (Formula 54)

In addition, the separation function 17b weights S _DWI2 (TE2, b1) to S _DWI3 (TE3, b1) with the weighting coefficient α and calculates the difference between S DWI2 (TE2, b1) and S DWI3 (TE3, b1) using the following (Equation 55), thereby producing an image in which the water component is suppressed. Generate S _wsup2 (TE2, b1). That is, since S _b (TE2, b1)≈0, an image S _t (TE2, b1) in which only tissue components remain is obtained.

S _wsup2 (TE2, b1) = S _DWI2 (TE2, b1) - α＊S _DWI3 (TE3, b1) ... (Formula 55)

In this embodiment, the generation function 17c generates the entire ADC map including water components and tissue components using the plurality of images collected by the collection function 17a. Further, the generation function 17c separates three components, a water component, a blood flow component, and a tissue component, using the first separated image, the second separated image, and the third separated image generated by the separation function 17b. generate an ADC map.

For example, similarly to Examples 4 and 5, the generation function 17c generates the entire ADC map ADC _orig using S _SE2 (TE2, b2) and S _DWI4 (TE2, b2) according to the following (Equation 56). generate.

ADC _orig = ln [S _SE1 (TE2, b0) / S _DWI4 (TE2, b2)] / (b2 - b0) ... (Formula 56)

In addition, since the water component and blood flow component can be ignored in the image of b=b2, the generation function 17c uses the following (Equation 57) to generate the image using S _DWI4 (TE2, b2) and the separation function 17b. An ADC map ADC _t of the tissue component is generated using S _wsup1 (TE2, b1).

ADC _t = ln [S _wsup1 (TE2, b1) / S _DWI4 (TE2, b2)] / (b2 - b1) ... (Formula 57)

Furthermore, the generation function 17c performs approximation using the nonlinear least squares method based on the bi-exponential model using S _SE2 (TE2, b0), S _DWI2 (TE2, b1), and S _DWI4 (TE2, b2). Then, an ADC map ADC _b of the blood flow component is generated.

In addition, since the water component is dominant in the image of TE=TE3, the generation function 17c uses S _SE3 (TE3, b0) S _DWI3 (TE3, b1) according to the following (Equation 58). , generate an ADC map ADC _t of the water component.

ADC _w = ln[S _SE3 (TE3, b0)/S _DWI3 (TE3, b1)]/(b1-b0) ... (Formula 58)

Further, the generation function 17c generates the tissue component image S _t (TE1, b0) using S _wsup1 (TE1, b1) and the ADC _t generated by the generation function 17c according to the following (Equation 59). generate. In addition, the generation function 17c generates the tissue component image S _t (TE2, b0) using S _wsup2 (TE1, b1) and the ADC _t generated by the generation function 17c according to the following (Equation 60). generate.

S _t (TE1, b0) = S _wsup1 (TE1, b1)/exp[-(b1-b0)*ADC _t ] ... (Formula 59)
S _t (TE2, b0) = S _wsup2 (TE2, b1)/exp[-(b1-b0)*ADC _t ] ... (Formula 60)

Furthermore, the generation function 17c generates an image S _b (TE1, b0) of the blood flow component by subtracting S _t (TE1, b0) from S _wsup1 (TE1, b0) using the following (Equation 61). do. In addition, the generation function 17c generates an image S _b (TE2, b0) of the blood flow component by subtracting S _t (TE2, b0) from S _wsup2 (TE2, b0) using the following (Equation 62). do.

S _b (TE1, b0) = S _wsup1 (TE1, b0) - S _t (TE1, b0) ... (Formula 61)
S _b (TE2, b0) = S _wsup2 (TE2, b0) - S _t (TE2, b0) ... (Formula 62)

Furthermore, the generation function 17c generates a water component image S _w by subtracting S _b (TE1, b0) and S _t (TE1, b0) from S _SE1 (TE1, b0) using the following (Equation 63). (TE1,b0) is generated. Furthermore, the generation function 17c generates a water component image S _w by subtracting S _b (TE2, b0) and S _t (TE2, b0) from S _SE1 (TE2, b0) using the following (Equation 64). (TE2, b0) is generated.

S _w (TE1, b0) = S _SE1 (TE1, b0) - S _b (TE1, b0) - S _t (TE1, b0) ... (Formula 63)
S _w (TE2, b0) = S _SE2 (TE2, b0) - S _b (TE2, b0) - S _t (TE2, b0) ... (Formula 64)

Furthermore, similarly to the first embodiment, the generation function 17c generates the entire T2 map T2 _orig using S _SE1 (TE1, b0) and S _SE2 (TE2, b0) according to the following (Equation 65). .

T2 _orig = (TE2-TE1)/ln[S _SE1 (TE1, b0)/S _SE2 (TE2, b0)] ... (Formula 65)

Further, the generation function 17c generates the T2 map T2 _w of the water component using S _w (TE1, b0) and S _w (TE2, b0) according to the following (Equation 66). Furthermore, the generation function 17c generates the T2 map T2 _w of the blood flow component using S _b (TE1, b0) and S _b (TE2, b0) according to the following (Equation 67). Furthermore, the generation function 17c generates a T2 map T2 _w of tissue components using S _t (TE1, b0) and S _t (TE2, b0) according to the following (Equation 68).

T2 _w = (TE2-TE1)/ln[S _w (TE1, b0)/S _w (TE2, b0)] ... (Formula 66)
T2 _b = (TE2-TE1)/ln[S _b (TE1, b0)/S _b (TE2, b0)] ... (Formula 67)
T2 _t = (TE2-TE1)/ln[S _t (TE1, b0)/S _t (TE2, b0)] ... (Formula 68)

Furthermore, similarly to the first embodiment, the generation function 17c generates the entire T2 map T2 _orig using S _SE1 (TE1, b0), T2 _orig , and ADC _orig according to the following (Equation 65).

PD _orig = S _SE (TE1, b0) / DecayT2 (TE1, T2 _orig ) / DecayADC (b0, ADC _orig ) ... (Formula 69)

Furthermore, the generation function 17c generates a PD map PD _w of the water component using S _w (TE1, b0), T2 _w , and ADC _w according to the following (Equation 70). Further, the generation function 17c generates a PD map PD _b of the blood flow component using S _b (TE1, b0), T2 _b and ADC _b according to the following (Equation 71). Furthermore, the generation function 17c generates a PD map PD _t of the tissue component using S _t (TE1, b0), T2 _t and ADC _t according to the following (Equation 72).

PD _w = S _w (TE1, b0) / DecayT2 (TE1, T2 _w ) / DecayADC (b0, ADC _w ) ... (Formula 70)
PD _b = S _b (TE1, b0) / DecayT2 (TE1, T2 _b ) / DecayADC (b0, ADC _b ) ... (Formula 71)
PD _t = S _t (TE1, b0) / DecayT2 (TE1, T2 _t ) / DecayADC (b0, ADC _t ) ... (Formula 72)

Here, DecayT2 (TE1, T2 _orig ), DecayT2 (TE, T2 _w ), DecayT2 (TE, T2 _b ), and DecayT2 (TE, T2 _t ) are each determined by the above-mentioned (Equation 5). Furthermore, DecayADC (b0, ADC _orig ), Decay ADC (b0, ADC _w ), Decay ADC (b0, ADC _b ), and Decay ADC (b0, ADC _t ) are each determined by the above-mentioned (Equation 6).

Note that when TE1<20 [ms], DecayT2≈1, so correction of T2 attenuation is not necessary. Furthermore, when b0≈0, DecayADC≈1, so there is no need to correct ADC attenuation.

In addition, although the generation of the ADC map, T2 map, and PD map has been explained here, by using S _IR1 (TR1, TI1, TE1, b0) as in Example 1, the entire T1 map T1 _orig , water It is also possible to generate a component T1 map T1 _orig , a blood flow component T1 map T1 _b and a tissue component T1 map T1 _t .

In this embodiment, the synthesis function 17d, as in the first embodiment, combines the entire T1 map T1 _orig , T2 map T2 _orig , and PD map PD _orig generated by the generation function 17c with the water component being suppressed. The T1 map T1 _t , T2 map T2 _t , and PD map PD _t are used in combination to generate a composite contrast-enhanced image of the SE method and a composite contrast-enhanced image of the FLAIR method corresponding to arbitrary imaging parameters. .

Although an example of collecting S _DWI3 (TE3, b1) has been described here, for example, by using ADC _W assumed or measured as in Examples 1 and 3, the following (Equation 73 ) may generate S _DWI3 (TE3, b1) from S _SE3 (TE3, b0). In that case, there is no need to collect S _DWI3 (TE3, b1).

S _DWI3 (TE3, b1) = S _SE3 (TE3, b0) * exp [-(b1-b0) * ADC _W ] ... (Formula 73)

In the sixth embodiment described above, since the difference in ADC between the tissue component and the blood flow component is large, the error in analysis due to noise is reduced compared to the case of separating water and tissue, which have a small difference in ADC. Further, for simplicity, if TR>5*T1 _w , T1 relaxation can be ignored.

The above is a method of generating quantitative parameter maps, or a method of suppressing the influence of PVE in voxels where different tissues are mixed in synthetic MRI in a broad sense, which generates synthetic contrast-enhanced images based on the quantitative parameter maps. However, the types of tissues to be separated, the generation method (original image acquisition sequence type and calculation algorithm) and types of quantitative parameter maps, and the types of synthetic contrast-enhanced images are not limited to those described above, but can be extended. Can be used. In this example, the original data collection was explained using a general SE sequence, but in recent years, as typified by the MRF mentioned above, the speed of original image collection and quantitative parameter map generation has increased (how much data can be generated from a small amount of data). For example, SSFP-based sequences, spiral incoherent acquisition, etc. are used as the acquisition unit, and the separation unit that generates separated images by data analysis is used in the acquisition sequence. Machine learning techniques such as dictionary matching and convolutional neural networks (CNN) based on compatible signal models can also be applied to quantitative parameter maps of water and tissue components obtained by separation. It is.

(Example 7)
Next, Example 7 will be described. The present embodiment is intended to suppress a decrease in SNR and a decrease in separation accuracy due to the difference between the images described above.

For example, in a difference using the difference in T2, in an image taken with a long TE in which the water component is dominant, if the TE is not long enough, the T2 contained in the tissue such as edema, which is a lesion, is It is also conceivable that the long portion may be included as a water component. In that case, instead of using a simple weighted difference or a difference in signal strength obtained from T2 of water, a threshold value is set for the signal strength or T2 of an image in which the water component to be differentiated is dominant, and the water component and tissue component are determined. A method of separating it is also possible. Similarly, in the separation using the difference in ADC, a method may be adopted in which a threshold value is set for the ADC to separate the water component and the tissue component.

Regarding this, when PVE occurs due to a mixture of water and tissue components, T2 and ADC continuously change between the water value and the tissue value depending on the volume ratio. Although it is difficult to completely separate water and tissue components using thresholds, if the thresholds for T2 and ADC are set midway between the water and tissue values with a margin, it is possible to , the risk that tissues with relatively large ADCs disappear due to the difference is reduced. Furthermore, it is also possible to combine and separate the threshold for T2 and the threshold for ADC.

In addition, with a 1:1 difference between images with the same SNR, the noise SD is multiplied by √2, so the SNR decreases to 1/√2. In the resulting image, there is almost no PVE due to the water component, and there is only tissue component noise, and the SNR is only degraded by the difference in the tissue component portion. Therefore, it is desirable not to differentiate the tissue component portion.

For this reason, in this embodiment, the separation function 17b subtracts the second image from the first image so that areas with high signal strength are subtracted, and areas with low signal intensity are not subtracted. An image in which the water component is suppressed is generated as a separated image.

That is, in this embodiment, the separation function 17b sets the weighting to 1 for a signal whose signal strength is greater than a predetermined threshold, sets the weighting to 0 for a signal whose signal strength is less than or equal to the threshold, and further sets the weighting to 0 for a signal whose signal strength is less than or equal to the threshold. After weighting the second image using a mask function that sets the weighting so that the signal near the threshold changes smoothly between 0 and 1, the second image is subtracted from the first image. By doing so, a separated image is generated.

FIG. 8 is a diagram illustrating an example of generation of separated images performed by the separation function 17b according to the seventh embodiment.

For example, as shown in the upper part of FIG. 8A, the separation function 17b sets a threshold value T2th for the signal strength T2(x) of the T2 map in which the water component is suppressed. For example, for T2th, the T2 of tissue in a standard T2 map in which water components are not suppressed is used. For example, when the object to be imaged is the brain, a value of about 100 [msec] is set.

As shown in the middle part of FIG. 8(a), the separation function 17b sets the weighting to 1 for signals whose T2(x) is greater than T2th, and sets the weighting to 0 for signals whose T2(x) is less than or equal to T2th. A mask function Mask(x) to be set is generated.

Thereafter, as shown in the lower part of FIG. 8(a), the separation function 17b performs smoothing processing on Mask(x) using a low-pass filter, so that for the signal where T2(x) is around T2th, the separation function 17b A mask function Mask _fil (x) is generated that sets the weighting so that it changes smoothly between .

In addition, as another method, for example, as shown in the upper part of FIG. Set Sth. Here, Sth is determined based on the amount of noise in S(x). For example, Sth is set to a value three to four times the noiseSD of S(x).

As shown in the middle part of FIG. 8(b), the separation function 17b sets the weighting to 1 for signals where S(x) is greater than Sth, and sets the weighting to 0 for signals where S(x) is less than or equal to Sth. A mask function Mask(x) to be set is generated.

Thereafter, as shown in the lower part of FIG. 8(b), the separation function 17b performs smoothing processing on Mask(x) using a low-pass filter, so that for the signal where T2(x) is around T2th, the separation function 17b A mask function Mask _fil (x) is generated that sets the weighting so that it changes smoothly between .

After generating Mask _fil (x) in this way, as shown in FIG _. By calculating the product with (x), an image S _cor (x) in which noise in portions where the signal value is zero or small is reduced is generated.

The separation function 17b then generates an image in which the water component is suppressed by subtracting the generated S _cor (x) from the PD-weighted image or T2-weighted image before the water component is suppressed. As a result, a separated image in which only the water component is suppressed and the SNR of the tissue component is maintained is obtained.

Note that here, the image taken with a long TE in which the water component is dominant is expressed as S(x), but if the image is a two-dimensional image S(x, y), The mask function and the image after conversion are also Mask (x, y), Mask _fil (x, y), and S _cor (x, y), respectively, and if it is a three-dimensional image S (x, y, z), The mask function and the converted image are also Mask (x, y, z), Mask _fil (x, y, z), and S _cor (x, y, z), respectively.

(Example 8)
Next, Example 8 will be explained. This example, like Example 7, suppresses the reduction in SNR and separation accuracy due to differences, but uses a nonlinear function instead of a mask function.

That is, in this embodiment, after the separation function 17b transforms the second image using a nonlinear function that nonlinearly transforms the signal strength in the range of 0 to 1 times for signals whose signal strength is below a predetermined threshold, A separated image is generated by subtracting the second image from the first image.

FIG. 9 is a diagram illustrating an example of generation of a separated image performed by the separation function 17b according to the eighth embodiment.

For example, as shown in the lower right of FIG. 9, the separation function 17b sets a threshold Sth for an image S(x) captured with a long TE in which the water component is dominant. Here, Sth is determined based on the amount of noise in S(x). For example, Sth is set to a value three to four times the noise of S(x).

Then, as shown in the upper right corner of FIG. 9, the separation function 17b generates a nonlinear function F(S) that nonlinearly transforms the signal strength in the range of 0 to 1 times for a signal where S(x) is less than or equal to Sth. Here, F(S) is a function that transforms S(x) so that the signal strength after transformation smoothly decreases to 0 in the range of S(x)<Sth with Sth as the boundary.

Thereafter, as shown in the upper left of FIG. 9, the separation function 17b converts S(x) using F(S), thereby eliminating noise in the portion where the signal value is zero or small, as in the seventh embodiment. Generate a reduced image S _cor (x).

The separation function 17b then generates an image in which the water component is suppressed by subtracting the generated S _cor (x) from the PD-weighted image or T2-weighted image before the water component is suppressed. As a result, a separated image in which only the water component is suppressed and the SNR of the tissue component is maintained is obtained.

FIG. 10 is a diagram illustrating an example of a nonlinear function used by the separation function 17b according to the eighth embodiment.

For example, as shown in Figure 10, F(x) has a threshold value for x of 1, and in the range of x < 1, x smoothly decreases from a value of 0.5 times to a value of 0 times. This is a function that converts it so that For example, F(x) is a modified sigmoid function, and is expressed by the following (Equation 74).

F(x)=x/{1+exp[-a*(|x|-1)]} ...(Formula 74)

Here, the smoothness of the change in signal strength after conversion near the threshold value can be controlled by the parameter a. Specifically, the smaller a is, the more smoothly F(x), which is the signal strength after conversion, changes near the threshold value.

For example, the separation function 17b obtains F(x) as x=S/Sth, and generates the converted image S _cor from S _cor =Sth*F(x). As a result, an image in which noise is reduced and the boundary between the water component and the tissue component is smoothly connected can be obtained.

Note that here, the image taken with a long TE in which the water component is dominant is expressed as S(x), but if the image is a two-dimensional image S(x, y), The nonlinear function for conversion and the image after conversion are also F (x, y) and S _cor (x, y), respectively, and if the image is a three-dimensional image S (x, y, z), the nonlinear function for conversion is The converted images are also F(x, y, z) and S _cor (x, y, z), respectively.

The eighth embodiment described above can also be used as a general-purpose image processing method and as a method for improving a portion of an image where the SNR is low.

Several examples have been described above as a method for suppressing deterioration in areas with low SNR when suppressing the water component by subtracting long TEs in which the water component is dominant. Although the method based on the T2 map described in Example 7 requires the T2 map, its greatest advantage is that the absolute value of T2 can be used as the threshold. In addition, the method based on the signal strength of images with long TE described in Example 7 requires noise measurement to determine the threshold value, but once the imaging conditions are determined, empirical values can be used without having to measure each time. It seems possible.

Although ADC is not explained here, similar to T2, a threshold value such as ADC=2*10 ^-3 [mm ² /s], which is an intermediate value between water ADC and tissue ADC, is used. By using this, it is possible to similarly apply the method based on the T2 map described in Example 7.

(Example 9)
Next, Example 9 will be explained. This example, like Example 7 and Example 8, suppresses a decrease in SNR and a decrease in separation accuracy due to differences, but uses a weighted difference using a simple index value.

For example, in a difference using a difference in T2, in an image taken with a long TE in which water components are dominant, the TE may not be sufficiently long compared to the T2 of the tissue, or the tissue may When T2 itself is long, the signal intensity of the tissue component becomes a value that cannot be ignored compared to the signal intensity of the water component, and even with the methods of Examples 7 and 8 described above, it is difficult to distinguish between the water component and the tissue component. It could happen.

On the other hand, as a method for more accurately separating the water component and tissue component even when the tissue component has a value that cannot be ignored in an image taken with a long TE in which the water component is dominant, for example, 4 If there are images that are collected at more than one collection sample point and include many images taken with a particularly long TE, a method of separating the images using the least squares method may be considered. However, such methods have large errors when the number of sample points collected in the image is small.

For this reason, in this embodiment, a simple index is used to enable more accurate separation of water components and tissue components.

That is, in this embodiment, the separation function 17b calculates an index value that reflects the ratio of the water component to the tissue component based on at least three images with different TEs among the plurality of images collected by the collection function 17a. A separated image is generated by calculating the difference from the first image after weighting the second image according to the index value.

Specifically, the separation function 17b uses three images with different TEs to calculate T2 obtained from the image with a short TE and an intermediate image as T2fast, and calculates T2 obtained from an image with a long TE and an intermediate image as T2fast. Calculate T2 as T2slow.

For example, if S(TE1), S(TE2), and S(TE3) are used, and TE3 is a long TE in which the water component is predominant, the separation function 17b is as follows (Equation 75) and T2fast and T2slow are each calculated using (Equation 76).

T2fast=(TE2-TE1)/ln(S1/S2)...(Formula 75)
T2slow=(TE3-TE2)/ln(S2/S3)...(Formula 76)

Furthermore, the separation function 17b calculates the ratio T2ratio between T2fast and T2slow using (Equation 77) below.

T2ratio=T2fast/T2slow...(Formula 77)

Here, when the component within a voxel is single, that is, when either the water component or the tissue component is dominant within the voxel, T2ratio approaches 1. On the other hand, when the water component and the tissue component are about half and half in a voxel, T2ratio approaches zero.

Then, the separation function 17b weights S(TE3) with a weighting coefficient αcor determined according to T2ratio and subtracts it from S(TE1) using the following (Equation 77), so that the water component is suppressed. An image S (TE1) is generated. In addition, the separation function 17b weights S(TE3) with a weighting coefficient αcor determined according to T2ratio and subtracts it from S(TE2) according to the following (Equation 78), so that the water component is suppressed. Generate image S _t (TE2).

S _t (TE1) = S (TE1) - αcor * S (TE3) ... (Formula 77)
S _t (TE2) = S (TE2) - αcor * S (TE3) ... (Formula 78)

At this time, the separation function 17b determines αcor depending on whether the tissue component is dominant, the water component and the tissue component are mixed, or the water component is dominant.

For example, the separation function 17b determines αcor using two thresholds T2Th1 and T2Th2 for T2 (T2Th1<T2Th2) and a threshold T2ratioTh for T2ratio (0<T2ratioTh<1). For example, when the imaging target is the brain, T2Th1=200 [ms], T2Th2=500 [ms], and T2ratioTh=0.7.

Specifically, the separation function 17b determines that tissue is dominant when T2<T2Th1 and T2ratio>T2ratioTh, and determines αcor=0. Furthermore, when 0<T2ratio≦T2ratioTh, the separation function 17b determines that the water component and tissue component are mixed, and determines that αcor=T2ratio ^β . Furthermore, when T2≧T2Th2 and T2ratio>T2ratioTh, the separation function 17b determines that water is dominant and determines αcor=1. Here, β is a coefficient of 0 or more for controlling the weight of T2ratio.

Note that when this example is combined with Example 7, in Mask _fil , the weighting for the water component is set to 1, the weighting for the tissue component is set to 0, and the weighting for the boundary part between the water component and the tissue component is set to 0 to 1. Therefore, αcor may be determined as αcor=Mask _fil *T2ratio ^β .

Furthermore, when this embodiment is combined with embodiment 8, the signal strength is obtained when TE=TE3, so αcor may be determined as αcor=F(S(TE3))*T2ratio ^β .

As a result, for example, even if the T2 or signal intensity is the same, if a voxel contains many tissue components, the T2ratio will become smaller, which will also reduce the weighting given to the tissue components, and as a result, after the difference Also, tissue components tend to remain.

Note that, in order to simplify the calculation, a simple signal strength difference, slope ratio, or the like that indicates the correlation with T2 may be used without calculating up to T2.

(Example 10)
Next, Example 10 will be explained. In this embodiment, the water component and tissue component are not separated by subtracting images, but by approximation using an MR signal model.

That is, in this embodiment, the separation function 17b uses at least three images with different TEs based on a model that defines the relationship between the overall signal strength of the signal and the signal strength of each of the water component and tissue component within the signal. A separated image is generated by performing least squares approximation.

Specifically, the separation function 17b uses three images with different TE according to the two-component model described above, and the water T2 _w assumed using the theoretical value or measured from the images, using the least squares method. By performing the approximation, four parameters of the two-component model, namely PD _t *V _t , T2 _t , PD _w *V _w and T2 _w , are obtained.

Here, it is theoretically possible to obtain the four parameters by using four images with different TE, but the number of necessary images can be calculated by using the assumed or measured water T2 _w . By setting the number to three, the approximate calculation can be stabilized.

Generally, the least squares method approximation algorithm is a nonlinear method, and it is known that the solution becomes unstable when the SNR of the data is low or when the number of data is small. Therefore, in this example, in order to be able to separate water components and tissue components with high precision, it is desirable to use images collected with a sufficiently long TE such that the tissue component is approximately zero. Even if the image contains some real components, it is approximated by a 2-compartment model formula using the entire data of 3 or more points, so it is much slower than approximating by a single-compartment model after performing weighted differences. , it is possible to avoid differences in real components.

In addition, in the nonlinear least squares approximation algorithm that calculates the parameters of a function expressed as the sum of exponential functions of two or more components, data is internally divided into temporal ranges, and a time constant is calculated from data with a long time. There is also a method of finding a component with a large (T2) and finding a short component while sequentially subtracting the data. In this embodiment, by using such a method, noise can be made more robust than when all parameters are approximated using all data at the same time.

In addition, here we have explained the separation of water components and tissue components using images with different TE, but by considering T2 as 1/ADC, the separation of water components and tissue components using images with different b values is possible. Separation can be performed in the same way.

In addition, here we have explained an example in which each image is treated as one-dimensional data, such as images with different TEs or images with different b values, and approximation is performed. By treating each image as two-dimensional data as described above, it is also possible to perform approximation using the nonlinear least squares method. Theoretically, if there are images corresponding to the same number of collected sample points as the number of unknown parameters to be sought, the combination of (TE, b) for each point may be randomly different, but the water component In order to separate tissue components with high precision, images must be taken with a sufficiently long TE so that the water component is dominant, or with a sufficiently large b value that the water component can be ignored. It is desirable to include the image in the images used for analysis. Even when each image is treated as two-dimensional data in this way, the number of necessary images can be reduced by using the assumed or measured water T2 _w and ADC, as in the case of one-dimensional data described above. Therefore, it is possible to improve the accuracy of approximation even with a small number of images.

In addition, when combining this example with Example 7 or 8, in addition to short TE images, analysis can be performed using long TE images after weighting by a mask function and conversion by a nonlinear function. become.

Examples 7 to 10 have been described above, and the calculation costs of these examples are lower in the order of Example 10, Example 9, Example 7, and Example 8. The methods of these embodiments can be selected as appropriate depending on conditions such as the SNR of the original image, the number of sampling points for collecting TE or b values, and required accuracy. Furthermore, this method can be commonly applied to other embodiments as a method for suppressing deterioration in SNR and reduction in separation accuracy due to differences.

As described above, according to the first embodiment and each example, a quantitative parameter map is generated using a separated image in which water components and tissue components in a signal are separated. PVE occurring in quantitative parameter maps can be suppressed by mixing with tissue components. Furthermore, by generating a composite contrast-enhanced image using a quantitative parameter map in which PVE is suppressed, it is possible to suppress high-signal artifacts that occur in the composite contrast-enhanced image due to PVE.

For example, according to the first embodiment and examples described above, by simply adding collection of images in which the water component is dominant, PVE due to the water component is suppressed by simple processing of the PD, T1T2, and ADC. Quantitative parameter maps such as the following can be easily generated. Moreover, it can be easily applied to conventional methods such as the FAST SE method.

Further, for example, according to the first embodiment and examples described above, it is possible to generate a Synthetic FLAIR image in which PVE due to water components, which cannot be obtained by actual acquisition, is suppressed.

Further, for example, according to the first embodiment and examples described above, the water component itself and PVE due to the water component, which cannot be obtained by actual collection, are suppressed, and the SNR and different tissues are lower than that of the FLAIR image. Synthetic SE images with high inter-CNR can be generated.

Further, for example, according to the first embodiment and examples described above, by combining the T2 difference and the ADC difference between the water component and the tissue component during actual collection, water can be collected in a shorter data collection time than in the conventional method. The effect of suppressing components or the effect of separating water components and tissue components can be improved.

Further, for example, according to the first embodiment and examples described above, by combining T2-based water component suppression and DWI-based water component suppression, the water component is suppressed more than when each is performed alone. The effect can be improved.

Furthermore, for example, according to the first embodiment and examples described above, it is possible to separate the three components of water, blood, and tissue, which is difficult to do with only diffusion weighted imaging (requiring images of four or more points with different b values). , by combining T2-based water component suppression, it can be performed with a relatively small b value (b≦1000 [s/mm ² ]) and a small number of images (at least 3 points).

Further, for example, according to the first embodiment and examples described above, the accuracy of quantitative parameter maps such as ADC and FA can be improved by performing simpler and more accurate water component suppression than conventional methods using diffusion weighted imaging. can be improved.

Further, for example, according to the first embodiment and examples described above, in addition to the synthetic contrast-enhanced image and the quantitative parameter map in which the water component is suppressed, if necessary, a conventional similar image in which the water component is not suppressed You can also get images of

Further, for example, according to the first embodiment and examples described above, by adding nonlinear differential processing, it is possible to reduce only the CSF signal that causes unnecessary PVE without reducing the SNR of the parenchymal part. becomes.

For example, according to the first embodiment and examples described above, Computed DWI can be extended to improve quantitative parameter maps such as ADC and FA.

From the above, according to the first embodiment and examples described above, in synthetic MRI, even when the mixing ratio of water component (CSF) and tissue component for each voxel is different, T2-FLAIR images in particular can be It is possible to provide a method for easily suppressing high signal artifacts estimated to be caused by PVE during generation.

Further, for example, according to the first embodiment and examples described above, the quantitative parameter map is created by suppressing the signal of the water component of the original image itself used for calculating the quantitative parameter according to the volume of water in the voxel. , and the water signal itself in the synthetic MRI image can be suppressed according to the water volume.

Further, for example, according to the first embodiment and examples described above, quantitative parameter maps (PD, T1, T2, ADC, FA, etc.) in which water components are suppressed, or their respective water components and structures. A quantitative parameter map with separated components can be generated.

Further, for example, according to the first embodiment and examples described above, a Synthetic FLAIR image, which is a water-suppressed synthetic image, is a contrast-enhanced image calculated using a quantitative parameter map equivalent to an actually acquired FLAIR image. , and can provide synthetic SE images with excellent water-suppressed tissue contrast resolution.

Further, for example, according to the first embodiment and examples described above, by diffusion weighted imaging, simple and highly accurate water suppression is performed on an SE basis without using FLAIR, and the ADC, FA, etc. of the tissue part is The accuracy of quantitative parameter maps can be improved.

Further, for example, according to the first embodiment and examples described above, by combining diffusion weighted imaging with a water component suppression method, water components, blood flow components, and tissue components can be separated more easily and with high precision. method can be provided.

Further, for example, according to the first embodiment and examples described above, by separating a certain tissue from a plurality of general tissues, PVE due to components excluded from a voxel in which a plurality of types of tissues are mixed can be reduced. A suppressed quantitative parameter map and a composite contrast-enhanced image can be obtained.

Further, for example, according to the first embodiment and example described above, by optimally controlling the weight of the difference for a simple difference where the weight of the difference is constant, the normal tissue and T2 are shorter than the CSF. However, it is possible to improve the SNR of relatively long tissues or lesions compared to normal tissues.

(Second embodiment)
As mentioned above, in the first embodiment, an example in which the technology disclosed in the present application is applied to an MRI apparatus has been described, but the embodiment is not limited to this. For example, the technology disclosed in this application can also be applied to an image processing device connected to the MRI apparatus 100 via a network. Therefore, an embodiment of an image processing apparatus will be described below as a second embodiment.

FIG. 11 is a diagram illustrating a configuration example of an image processing apparatus according to the second embodiment.

For example, as shown in FIG. 11, in this embodiment, an MRI apparatus 100 and an image processing apparatus 200 are communicably connected via a network 300.

The MRI apparatus 100 collects image data of a subject using magnetic resonance phenomena. Specifically, the MRI apparatus 100 collects magnetic resonance data from a subject by executing various imaging sequences based on imaging conditions set by an operator. Then, the MRI apparatus 100 generates two-dimensional or three-dimensional image data (MR image) by performing image processing such as Fourier transform processing on the collected magnetic resonance data.

The image processing device 200 processes image data collected by the MRI device 100. Specifically, the image processing device 200 acquires image data from the MRI device 100 via the network 300, and stores it in a storage circuit provided within the device or outside the device. The image processing device 200 also performs various image processing on the acquired image data, and displays the image data before or after the image processing on a display or the like. For example, the image processing device 200 is realized by computer equipment such as a workstation.

Specifically, the image processing device 200 includes a network (NW) interface 210, a storage circuit 220, an input interface 230, a display 240, and a processing circuit 250.

The NW interface 210 controls transmission and communication of various data sent and received between the image processing device 200 and other devices connected via the network 300. Specifically, the NW interface 210 is connected to the processing circuit 250, converts image data output from the processing circuit 250 into a format compliant with a predetermined communication protocol, and transmits the converted image data to the MRI apparatus 100. Further, the NW interface 210 outputs image data received from the MRI apparatus 100 to the processing circuit 250. For example, the NW interface 210 is realized by a network card, network adapter, NIC (Network Interface Controller), or the like.

The storage circuit 220 stores various data. Specifically, the storage circuit 220 is connected to the processing circuit 250 and stores input image data or transfers stored image data to the processing circuit 250 in accordance with a command sent from the processing circuit 250. Output. For example, the storage circuit 220 is realized by a semiconductor memory element such as a RAM (Random Access Memory) or a flash memory, a hard disk, an optical disk, or the like.

The input interface 230 receives various instructions and input operations for various information from the operator. Specifically, the input interface 230 is connected to the processing circuit 250, converts an input operation received from the operator into an electrical signal, and outputs the electrical signal to the control circuit. For example, the input interface 230 includes a trackball for setting a region of interest (ROI), a switch button, a mouse, a keyboard, a touchpad for performing input operations by touching the operation surface, a display screen, and a touchpad. This is realized by a touch screen integrated with a pad, a non-contact input interface using an optical sensor, a voice input interface, etc. Note that in this specification, the input interface 230 is not limited to one that includes physical operation parts such as a mouse and a keyboard. For example, an example of the input interface 230 includes an electrical signal processing circuit that receives an electrical signal corresponding to an input operation from an external input device provided separately from the device and outputs this electrical signal to a control circuit.

Display 240 displays various information and various images. Specifically, the display 240 is connected to the processing circuit 250 and displays images in various formats based on image data output from the processing circuit 250. For example, the display 240 is realized by a liquid crystal monitor, a CRT (cathode ray tube) monitor, a touch panel, or the like.

The processing circuit 250 controls each component included in the image processing apparatus 200 according to input operations received from the operator via the input interface 230. Specifically, the processing circuit 250 causes the storage circuit 220 to store the image data output from the NW interface 210. Furthermore, the processing circuit 250 displays the image data read from the storage circuit 220 on the display 240. For example, processing circuit 250 is implemented by a processor.

With such a configuration, the image processing apparatus 200 according to the present embodiment extracts quantitative parameters such as PD, T1, and T2 from a plurality of images collected by the MRI apparatus 100 while changing the values of a plurality of types of imaging parameters. It has the ability to generate maps. Furthermore, the image processing device 200 according to the present embodiment calculates and generates a composite contrast-enhanced image corresponding to various contrast-enhanced images, using the signal value of the generated quantitative parameter map and the value of an arbitrary imaging parameter. It has a function.

The image processing device 200 according to the present embodiment is configured to be able to suppress PVE that occurs in the quantitative parameter map due to the mixing of the water component and tissue component in the signal.

Specifically, in this embodiment, the processing circuit 250 shown in FIG. 11 has an acquisition function 251, a separation function 252, a generation function 253, and a composition function 254.

The acquisition function 251 acquires images of the same subject in which multiple types of tissues coexist by changing the values of multiple types of imaging parameters, and acquires images captured at a TE close to T2 of the tissue and signals. A plurality of images are acquired from the MRI apparatus 100, including images taken with a long TE in which the water component is dominant. Note that the acquisition function 251 is an example of an acquisition unit.

Similar to the separation function 17b described in the first embodiment, the separation function 252 analyzes at least three images with different TE among the plurality of images collected by the collection function 17a, thereby separating water components and tissues. A separated image is generated by separating the components. Note that the separation function 252 is an example of a separation unit.

Similar to the generation function 17c described in the first embodiment, the generation function 253 uses the separated image generated by the separation function 17b to generate quantitative parameters of at least one of T1, T2, PD, and ADC. Generate a map. Note that the generation function 253 is an example of a generation unit.

Similar to the synthesis function 17d described in the first embodiment, the synthesis function 254 uses at least one type of quantitative parameter map generated by the generation function 17c to generate a synthesized contrast-enhanced image corresponding to an arbitrary imaging parameter. Generate by calculation. Note that the composition function 254 is an example of a composition section.

Note that the processing circuit 250 described above is realized by, for example, a processor. In this case, each processing function that the processing circuit 250 has is stored in the storage circuit 220 in the form of a computer-executable program, for example. The processing circuit 250 reads each program from the storage circuit 220 and executes it, thereby realizing a processing function corresponding to each program. In other words, the processing circuit 250 in a state where each program is read has each function shown in each processing circuit in FIG. Note that although the processing circuit 250 has been described here as being realized by a single processor, the processing circuit 250 is configured by combining a plurality of independent processors, and each processor executes a program to perform each processing function. It does not matter if it is something that realizes. Further, the processing functions of the processing circuit 250 may be realized by being appropriately distributed or integrated into a single or multiple processing circuits. Furthermore, in the example shown in FIG. 14, the single storage circuit 220 stores programs corresponding to each processing function. A configuration may also be used in which a corresponding program is read out from the circuit.

According to the above-described configuration, in the second embodiment as well, as in the first embodiment and examples described above, it is possible to suppress PVE that occurs in the quantitative parameter map due to the mixing of the water component and tissue component in the signal. I can do it.

Note that in the embodiments described above, the collection unit, separation unit, generation unit, synthesis unit, and acquisition unit in this specification are realized by the collection function, separation function, generation function, synthesis function, and acquisition function of the processing circuit, respectively. Although an example of the case has been described, the embodiment is not limited to this. For example, the collection unit, separation unit, generation unit, synthesis unit, and acquisition unit in this specification may be realized by the collection function, separation function, generation function, synthesis function, and acquisition unit described in the embodiment, as well as by hardware. The same function may be achieved by only one or a combination of hardware and software.

Furthermore, the term "processor" used in the above explanation refers to, for example, a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), an Application Specific Integrated Circuit (ASIC), or a programmable logic device. (For example, it refers to circuits such as Simple Programmable Logic Device (SPLD), Complex Programmable Logic Device (CPLD), and Field Programmable Gate Array (FPGA)). A processor achieves its functions by reading and executing a program stored in a memory circuit. Note that instead of storing the program in the memory circuit, the program may be directly incorporated into the circuit of the processor. In this case, the processor implements its functions by reading and executing a program built into the circuit. Further, the processor of this embodiment is not limited to being configured as a single circuit, but may be configured as a single processor by combining a plurality of independent circuits to realize its functions.

Here, the program executed by the processor is provided by being pre-installed in a ROM (Read Only Memory), a storage circuit, or the like. This program is a file in a format that can be installed or executable on these devices, such as CD (Compact Disk)-ROM, FD (Flexible Disk), CD-R (Recordable), DVD (Digital Versatile Disk), etc. It may be provided recorded on a computer readable storage medium. Further, this program may be stored on a computer connected to a network such as the Internet, and may be provided or distributed by being downloaded via the network. For example, this program is composed of modules including each of the above-mentioned functional units. In actual hardware, a CPU reads a program from a storage medium such as a ROM and executes it, so that each module is loaded onto the main storage device and generated on the main storage device.

According to at least one embodiment described above, it is possible to suppress PVE that occurs in the quantitative parameter map due to the mixing of water components and tissue components in the signal.

Although several embodiments of the invention have been described, these embodiments are presented by way of example 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 changes can be made without departing from the gist of the invention. These embodiments and their modifications are included within the scope and gist of the invention as well as within the scope of the invention described in the claims and its equivalents.

100 MRI apparatus 17 processing circuit 17a acquisition function 17b separation function 17c generation function 17d synthesis function

Claims (26)

By capturing images of the same subject in which multiple types of tissues coexist while changing the values of multiple types of imaging parameters, the first image captured at the first echo time and the time difference between the first image captured at the first echo time and the first image captured at the first echo time are A second image taken at a second echo time that is long and close to the transverse relaxation time of the tissue, and a third image taken at a third echo time that is long and close to the transverse relaxation time of the tissue. a collection unit that collects a plurality of images including;
a separation unit that generates a separated image in which a water component and a tissue component are separated by analyzing at least the first image, the second image, and the third image of the plurality of images; ,
a generation unit that uses the separated images to generate a quantitative parameter map of at least one of longitudinal relaxation time, transverse relaxation time, proton density, and diffusion coefficient;
The collecting unit collects the plurality of images in which water components are suppressed by capturing images of the subject using a non-zero b value by diffusion weighted imaging that captures diffusion weighted images;
The separating unit generates a first separated image in which a water component is suppressed by weighting and subtracting the third image from the first image , and separates the third image from the second image. By weighting and subtracting the images, a second separated image in which the water component is suppressed is generated,
The generation unit generates the at least one type of quantitative parameter map using the first separated image and the second separated image.
Magnetic resonance imaging device. By capturing images of the same subject in which multiple types of tissues coexist while changing the values of multiple types of imaging parameters, the first image captured at the first echo time and the time difference between the first image captured at the first echo time and the first image captured at the first echo time are A second image taken at a second echo time that is long and close to the transverse relaxation time of the tissue, and a third image taken at a third echo time that is long and close to the transverse relaxation time of the tissue. a collection unit that collects a plurality of images including;
a separation unit that generates a separated image in which a water component and a tissue component are separated by analyzing at least the first image, the second image, and the third image of the plurality of images;
a generation unit that uses the separated images to generate a quantitative parameter map of at least one of longitudinal relaxation time, transverse relaxation time, proton density, and diffusion coefficient;
Equipped with
The collecting unit collects the plurality of images in which water components are suppressed by capturing images of the subject using a non-zero b value by diffusion weighted imaging that captures diffusion weighted images;
The separation unit uses the first image and the second image to determine the minimum Generating a first separated image in which water components and tissue components are separated by performing approximation using the square method, and performing approximation using the least square method using the second image and the third image. to generate a second separated image in which water components and tissue components are separated ,
The generation unit generates the at least one type of quantitative parameter map using the first separated image and the second separated image.
Magnetic resonance imaging device. further comprising a synthesis unit that uses the at least one type of quantitative parameter map to generate a synthetic contrast-enhanced image corresponding to an arbitrary imaging parameter by calculation;
The magnetic resonance imaging apparatus according to claim 1 or 2. The separation unit is configured to extract an image of a first water component corresponding to the first echo time from the third image and an image of the second water component corresponding to the second echo time using an assumed or measured transverse relaxation time of water. generate an image of a second water component , and subtract the image of the first water component from the first image to generate the first separated image ; generating the second separated image by subtracting an image from the second image ;
The magnetic resonance imaging apparatus according to claim 1. The generation unit uses the first separated image and the second separated image to generate at least a quantitative parameter map of the transverse relaxation time with the water component suppressed and a quantitative parameter of the proton density with the water component suppressed. generate a map and
The synthesis unit uses the quantitative parameter map of the proton density with the water component suppressed and the quantitative parameter map of the transverse relaxation time with the water component suppressed, to perform the spin echo method with the water component suppressed. generate a synthetic contrast-enhanced image,
The magnetic resonance imaging apparatus according to claim 3. The generating unit generates a quantitative parameter map of the entire longitudinal relaxation time including water components and tissue components using the plurality of images, and using the first separated image and the second separated image, generating a quantitative parameter map of transverse relaxation time in which the water component is suppressed, and a quantitative parameter map of proton density in which the water component is suppressed;
The synthesis unit uses the quantitative parameter map of the overall longitudinal relaxation time, the quantitative parameter map of the transverse relaxation time with the water component suppressed, and the quantitative parameter map of the proton density with the water component suppressed. , generating a synthetic contrast-enhanced image of the spin echo method in which the water component is suppressed;
The magnetic resonance imaging apparatus according to claim 5 . The generation unit uses the first separated image and the second separated image to generate at least a quantitative parameter map of the transverse relaxation time in which a water component is suppressed,
The synthesis unit generates a composite contrast-enhanced image of the FLAIR method in which the water component is suppressed, using the quantitative parameter map of the transverse relaxation time in which the water component is suppressed.
The magnetic resonance imaging apparatus according to claim 3. The acquisition unit applies a gradient magnetic field for Gradient Moment Nulling to capture an image of the subject.
The magnetic resonance imaging apparatus according to any one of claims 1 to 7 . The acquisition unit captures an image of the subject using a multi-echo method.
The magnetic resonance imaging apparatus according to any one of claims 1 to 8 . By capturing images of the same subject containing multiple types of tissue while changing the values of multiple types of imaging parameters, it is possible to differentiate between images captured with an echo time close to the transverse relaxation time of the tissue and water components in the signal. an acquisition unit that collects a plurality of images including images captured with a dominantly long echo time;
a separation unit that generates a separated image in which a water component and a tissue component are separated by analyzing images with different echo times among the plurality of images;
a generation unit that uses the separated images to generate a quantitative parameter map of at least one of longitudinal relaxation time, transverse relaxation time, proton density, and diffusion coefficient;
The acquisition unit uses diffusion weighted imaging to capture a diffusion weighted image to obtain a first echo time close to the transverse relaxation time of the tissue, a long second echo time such that the water component is dominant, and a second echo time that is long enough to be dominated by the water component. By capturing an image of the subject using a b value of 1 and a second b value that is larger than the first b value such that water components are suppressed compared to the first b value. , a first image taken at the first echo time and the first b value, a second image taken at the second echo time and the first b value, and the first echo. and a third image taken at the time and the second b value;
The separation unit generates, as the separated image, an image in which a water component corresponding to the first b value is suppressed by weighting and subtracting the second image from the first image,
The generation unit generates a quantitative parameter map of a diffusion coefficient in which a water component is suppressed, using the third image and the separated image.
Magnetic resonance imaging device. By capturing images of the same subject containing multiple types of tissue while changing the values of multiple types of imaging parameters, it is possible to differentiate between images captured with an echo time close to the transverse relaxation time of the tissue and water components in the signal. an acquisition unit that collects a plurality of images including images captured with a dominantly long echo time;
a separation unit that generates a separated image in which a water component and a tissue component are separated by analyzing images with different echo times among the plurality of images;
a generation unit that uses the separated images to generate a quantitative parameter map of at least one of longitudinal relaxation time, transverse relaxation time, proton density, and diffusion coefficient;
The acquisition unit uses diffusion weighted imaging to capture a diffusion weighted image to obtain a first echo time close to the transverse relaxation time of the tissue, a long second echo time such that the water component is dominant, and a second echo time that is long enough to be dominated by the water component. By capturing an image of the subject using a b value of 1 and a second b value that is larger than the first b value such that water components are suppressed compared to the first b value. , a first image taken at the first echo time and the first b value, a second image taken at the second echo time and the first b value, and the first echo. and a third image taken at the time and the second b value;
The separation unit generates, as the separated image, an image in which a water component corresponding to the first b value is suppressed by weighting and subtracting the second image from the first image,
The generation unit generates an image of a water component corresponding to the second b value from the second image using an assumed or measured water diffusion coefficient, and generates an image of a water component corresponding to the second b value from the third image. By performing the difference, an image in which the water component is suppressed corresponding to the second b value is generated, and using this image and the separated image, a quantitative parameter map of the diffusion coefficient in which the water component is suppressed is generated. generate,
Magnetic resonance imaging device. By capturing images of the same subject containing multiple types of tissue while changing the values of multiple types of imaging parameters, it is possible to differentiate between images captured with an echo time close to the transverse relaxation time of the tissue and water components in the signal. an acquisition unit that collects a plurality of images including images captured with a dominantly long echo time;
a separation unit that generates a separated image in which a water component and a tissue component are separated by analyzing images with different echo times among the plurality of images;
a generation unit that uses the separated images to generate a quantitative parameter map of at least one of longitudinal relaxation time, transverse relaxation time, proton density, and diffusion coefficient;
The acquisition unit captures images of the subject using three different b values by diffusion weighted imaging that captures diffusion weighted images,
The separation unit separates three components including a moving blood flow component in addition to the stationary water component and tissue component by analyzing the images taken at each of the three different b values. generate a separated image,
Magnetic resonance imaging device. The collecting section has a first echo time that is close to the transverse relaxation time of the tissue, a second echo time that is longer than the first echo time such that the water component is dominant, and a water component in the signal. , a first b value such that all tissue components and blood flow components are mixed, and the first b value such that tissue components and water components are dominant in the signal compared to the first b value. of the subject using a second b value that is larger and a third b value that is larger than the second b value such that tissue components are dominant in the signal compared to the second b value. By capturing images, a first image captured at the first echo time and the first b value, and a second image captured at the second echo time and the first b value. , a third image taken at the first echo time and the second b value, a fourth image taken at the second echo time and the second b value, and the second echo a fifth image taken at the time and the third b value;
The separation unit weights and subtracts the second image from the first image to generate an image in which water components are suppressed and only tissue components and blood flow components remain as a first separated image. Then, by weighting and subtracting the fourth image from the third image, an image in which the water component is suppressed and only the tissue component remains is generated as a second separated image, and the first separated image is using the image and the second separated image to generate a third separated image in which water components and tissue components are suppressed and only blood flow components remain;
The generation unit generates a quantitative parameter map of diffusion coefficients in which the three components are separated, using the first separated image, the second separated image, and the third separated image.
The magnetic resonance imaging apparatus according to claim 12 . The separation unit subtracts the third image from the first image and the second image so that areas with high signal strength are subtracted and areas with low signal intensity are not subtracted. , generating the first separated image and the second separated image;
The magnetic resonance imaging apparatus according to claim 1 or 4 . The separation unit sets weighting to 1 for signals whose signal strength is greater than a predetermined threshold, sets weighting to 0 for signals whose signal strength is less than or equal to the threshold, and further sets weighting to 0 for signals whose signal strength is near the threshold. After weighting the third image using a mask function that sets the weighting so that it changes smoothly between 0 and 1, the third image is divided from the first image and the second image. After converting the third image by subtracting or using a nonlinear function that nonlinearly converts the signal intensity in the range of 0 to 1 times for signals whose signal intensity is less than or equal to a predetermined threshold, generating the first separated image and the second separated image by subtracting an image from the first image and the second image;
The magnetic resonance imaging apparatus according to claim 14 . The separation unit calculates an index value that reflects a ratio of a water component to a tissue component based on at least the first image, the second image, and the third image of the plurality of images. The first separated image and the second separated image are obtained by calculating and subtracting the third image from the first image and the second image after weighting the third image according to the index value. generate,
The magnetic resonance imaging apparatus according to claim 14 . By capturing images of the same subject in which multiple types of tissues coexist while changing the values of multiple types of imaging parameters, the first image captured at the first echo time and the time difference between the first image captured at the first echo time and the first image captured at the first echo time are A second image taken at a second echo time that is long and close to the transverse relaxation time of the tissue, and a third image taken at a third echo time that is long and close to the transverse relaxation time of the tissue. a collecting step of collecting a plurality of images including;
a separation step of generating a separated image in which a water component and a tissue component are separated by analyzing at least the first image, the second image, and the third image of the plurality of images; ,
a generation step of generating a quantitative parameter map of at least one of longitudinal relaxation time, transverse relaxation time, proton density, and diffusion coefficient using the separated images,
The collecting step collects the plurality of images in which water components are suppressed by capturing images of the subject using a non-zero b value by diffusion weighted imaging that captures diffusion weighted images;
The separation step includes weighting and subtracting the third image from the first image to generate a first separated image in which the water component is suppressed , and separating the third image from the second image. By weighting and subtracting the images, a second separated image in which the water component is suppressed is generated,
The generating step generates the at least one type of quantitative parameter map using the first separated image and the second separated image.
Magnetic resonance imaging methods. By capturing images of the same subject containing multiple types of tissue while changing the values of multiple types of imaging parameters, it is possible to differentiate between images captured with an echo time close to the transverse relaxation time of the tissue and water components in the signal. an acquisition step of collecting a plurality of images including an image taken with a predominant long echo time;
a separation step of generating a separated image in which a water component and a tissue component are separated by analyzing images with different echo times among the plurality of images;
a generation step of generating a quantitative parameter map of at least one of longitudinal relaxation time, transverse relaxation time, proton density, and diffusion coefficient using the separated images,
In the acquisition step, a first echo time close to the transverse relaxation time of the tissue, a second long echo time such that the water component is dominant, and a second echo time are obtained by diffusion weighted imaging to take a diffusion weighted image. By capturing an image of the subject using a b value of 1 and a second b value that is larger than the first b value such that water components are suppressed compared to the first b value. , a first image taken at the first echo time and the first b value, a second image taken at the second echo time and the first b value, and the first echo. and a third image taken at the time and the second b value;
In the separation step, an image in which a water component corresponding to the first b value is suppressed is generated as the separated image by weighting and subtracting the second image from the first image,
The generation step uses the third image and the separated image to generate a quantitative parameter map of a diffusion coefficient in which a water component is suppressed.
Magnetic resonance imaging methods. By capturing images of the same subject containing multiple types of tissue while changing the values of multiple types of imaging parameters, it is possible to differentiate between images captured with an echo time close to the transverse relaxation time of the tissue and water components in the signal. an acquisition step of collecting a plurality of images including an image taken with a predominant long echo time;
a separation step of generating a separated image in which a water component and a tissue component are separated by analyzing images with different echo times among the plurality of images;
a generation step of generating a quantitative parameter map of at least one of longitudinal relaxation time, transverse relaxation time, proton density, and diffusion coefficient using the separated images,
In the acquisition step, a first echo time close to the transverse relaxation time of the tissue, a second long echo time such that the water component is dominant, and a second echo time are obtained by diffusion weighted imaging to take a diffusion weighted image. By capturing an image of the subject using a b value of 1 and a second b value that is larger than the first b value such that water components are suppressed compared to the first b value. , a first image taken at the first echo time and the first b value, a second image taken at the second echo time and the first b value, and the first echo. and a third image taken at the time and the second b value;
In the separation step, an image in which a water component corresponding to the first b value is suppressed is generated as the separated image by weighting and subtracting the second image from the first image,
The generating step generates an image of the water component corresponding to the second b value from the second image using the assumed or measured water diffusion coefficient, and generates an image of the water component corresponding to the second b value from the third image. By performing the difference, an image in which the water component is suppressed corresponding to the second b value is generated, and using this image and the separated image, a quantitative parameter map of the diffusion coefficient in which the water component is suppressed is generated. generate,
Magnetic resonance imaging methods. By capturing images of the same subject containing multiple types of tissue while changing the values of multiple types of imaging parameters, it is possible to differentiate between images captured with an echo time close to the transverse relaxation time of the tissue and water components in the signal. an acquisition step of collecting a plurality of images including an image taken with a predominant long echo time;
a separation step of generating a separated image in which a water component and a tissue component are separated by analyzing images with different echo times among the plurality of images;
a generation step of generating a quantitative parameter map of at least one of longitudinal relaxation time, transverse relaxation time, proton density, and diffusion coefficient using the separated images,
In the acquisition step, images of the subject are captured using three different b values by diffusion weighted imaging that captures diffusion weighted images;
The separation step separates three components, including a still water component and tissue component, as well as a moving blood flow component, by analyzing images taken at each of the three different b values. generate a separated image,
Magnetic resonance imaging methods. The separating step is performed by subtracting the third image from the first image and the second image such that areas with high signal strength are subtracted and areas with low signal strength are not subtracted. , generating the first separated image and the second separated image;
The magnetic resonance imaging method according to claim 17 . A first image captured at a first echo time, which is collected by capturing images of the same subject in which multiple types of tissues are mixed, while changing the values of multiple types of imaging parameters, and a first image captured at a first echo time; A second image taken with a second echo time that is longer than the echo time and close to the transverse relaxation time of the tissue, and a third image taken with a third echo time that is longer such that the water component becomes dominant in the signal. an acquisition unit that acquires a plurality of images including the image;
a separation unit that generates a separated image in which a water component and a tissue component are separated by analyzing at least the first image, the second image, and the third image of the plurality of images; ,
a generation unit that uses the separated images to generate a quantitative parameter map of at least one of longitudinal relaxation time, transverse relaxation time, proton density, and diffusion coefficient;
The acquisition unit acquires the plurality of images in which water components are suppressed, which are collected by capturing images of the subject using a non-zero b value by diffusion weighted imaging that captures diffusion weighted images;
The separating unit generates a first separated image in which a water component is suppressed by weighting and subtracting the third image from the first image , and separates the third image from the second image. By weighting and subtracting the images, a second separated image in which the water component is suppressed is generated,
The generation unit generates the at least one type of quantitative parameter map using the first separated image and the second separated image.
Image processing device. Images of the same subject in which multiple types of tissues coexist are acquired by changing the values of multiple types of imaging parameters, and images captured with an echo time close to the transverse relaxation time of the tissue are compared with images captured in the signal. an acquisition unit that acquires a plurality of images including an image captured with a long echo time in which a water component is dominant;
a separation unit that generates a separated image in which a water component and a tissue component are separated by analyzing images with different echo times among the plurality of images;
a generation unit that uses the separated images to generate a quantitative parameter map of at least one of longitudinal relaxation time, transverse relaxation time, proton density, and diffusion coefficient;
The acquisition unit obtains a first echo time close to the transverse relaxation time of the tissue, a second echo time so long that the water component is dominant, and a second echo time, which is long enough to be dominated by the water component, by diffusion weighted imaging that captures a diffusion weighted image. By capturing an image of the subject using a b value of 1 and a second b value that is larger than the first b value such that water components are suppressed compared to the first b value. A first image captured at the first echo time and the first b value, a second image captured at the second echo time and the first b value, and a second image captured at the second echo time and the first b value, which are collected. a third image taken at an echo time of 1 and the second b value;
The separation unit generates, as the separated image, an image in which a water component corresponding to the first b value is suppressed by weighting and subtracting the second image from the first image,
The generation unit generates a quantitative parameter map of a diffusion coefficient in which a water component is suppressed, using the third image and the separated image.
Image processing device. Images of the same subject in which multiple types of tissues coexist are acquired by changing the values of multiple types of imaging parameters, and images captured with an echo time close to the transverse relaxation time of the tissue are compared with images captured in the signal. an acquisition unit that acquires a plurality of images including an image captured with a long echo time in which a water component is dominant;
a separation unit that generates a separated image in which a water component and a tissue component are separated by analyzing images with different echo times among the plurality of images;
a generation unit that uses the separated images to generate a quantitative parameter map of at least one of longitudinal relaxation time, transverse relaxation time, proton density, and diffusion coefficient;
The acquisition unit obtains a first echo time close to the transverse relaxation time of the tissue, a second echo time so long that the water component is dominant, and a second echo time, which is long enough to be dominated by the water component, by diffusion weighted imaging that captures a diffusion weighted image. By capturing an image of the subject using a b value of 1 and a second b value that is larger than the first b value such that water components are suppressed compared to the first b value. A first image captured at the first echo time and the first b value, a second image captured at the second echo time and the first b value, and a second image captured at the second echo time and the first b value, which are collected. a third image taken at an echo time of 1 and the second b value;
The separation unit generates, as the separated image, an image in which a water component corresponding to the first b value is suppressed by weighting and subtracting the second image from the first image,
The generation unit generates an image of a water component corresponding to the second b value from the second image using an assumed or measured water diffusion coefficient, and generates an image of a water component corresponding to the second b value from the third image. By performing the difference, an image in which the water component is suppressed corresponding to the second b value is generated, and using this image and the separated image, a quantitative parameter map of the diffusion coefficient in which the water component is suppressed is generated. generate,
Image processing device. Images of the same subject in which multiple types of tissues coexist are acquired by changing the values of multiple types of imaging parameters, and images captured with an echo time close to the transverse relaxation time of the tissue are compared with images captured in the signal. an acquisition unit that acquires a plurality of images including an image captured with a long echo time in which a water component is dominant;
a separation unit that generates a separated image in which a water component and a tissue component are separated by analyzing images with different echo times among the plurality of images;
a generation unit that uses the separated images to generate a quantitative parameter map of at least one of longitudinal relaxation time, transverse relaxation time, proton density, and diffusion coefficient;
The acquisition unit acquires images of the subject collected using three different b values by diffusion weighted imaging that captures diffusion weighted images,
The separation unit separates three components including a moving blood flow component in addition to the stationary water component and tissue component by analyzing the images taken at each of the three different b values. generate a separated image,
Image processing device. The separation unit subtracts the third image from the first image and the second image so that areas with high signal strength are subtracted and areas with low signal intensity are not subtracted. , generating the first separated image and the second separated image;
The image processing device according to claim 22 .

Images