JP2018196582A - Magnetic resonance imaging apparatus and correction method of diffusion weighted image - Google Patents

Abstract

To reduce an affection of a contrast difference between images and heat in acquisition time, improve correction accuracy of a diffusion weighted image, for acquiring an accurate calculation image.SOLUTION: The invention is configured to use an image of a b value≠0 as a reference image, make application axes of an MPG pulse different, for calculating a correction amount between plural images which are acquired at the same b value. At this time, a diffusion weighted image having a prescribed b value (b≠0) at a prescribed application axis is used as a reference image, and correction amounts of application axes other than the prescribed application axis are calculated sequentially. In addition, the diffusion weighted images having plural b values at the same application axis are corrected using a correction amount calculated for the application axis.SELECTED DRAWING: Figure 3

Description

  The present invention relates to a magnetic resonance imaging (hereinafter referred to as “MRI”) apparatus having a function of acquiring a diffusion weighted image, and in particular, between a plurality of diffusion weighted images acquired using a plurality of (two or more) b values. The present invention relates to a technique for correcting misalignment and distortion.

  An MRI apparatus measures NMR signals generated by nuclear spins, mainly protons, which constitute a subject's tissue, particularly human tissue, and forms the shape and function of the head, abdomen, limbs, etc. two-dimensionally or three-dimensionally. It is a device that automatically images. One of the images that can be acquired by the MRI apparatus is a diffusion weighted image obtained by imaging the ease of movement of water molecules.

  In imaging for acquiring a diffusion-weighted image, an NMR signal is acquired by applying a diffusion-sensitive gradient magnetic field called MPG (Motion Probing Gradient). By performing a plurality of measurements by varying the axis to which the MPG pulse is applied and the b value which is a parameter representing the intensity of the pulse, an apparent diffusion coefficient (ADC) and diffusion anisotropy (FA) : Fractal Anisotropy) can be used to obtain a calculation image useful for diagnosis.

  When the calculation is performed using a plurality of diffusion weighted images, it is necessary to correct in advance the distortion of the diffusion weighted image, the positional deviation between the images, and the like. As a correction method thereof, Patent Document 1 discloses that a b-value = 0, that is, an image acquired without applying an MPG pulse is used as a reference for a diffusion-weighted image captured with a b-value of 2 or more. A technique for correcting the position and distortion of a diffusion-weighted image is disclosed. By this method, it is possible to calculate various calculation images with less positional displacement and distortion and clearer contrast of the lesion.

JP 2012-157687 A

  However, an image acquired without applying an MPG pulse (an image with a b value = 0) has an image contrast that is significantly different from other diffusion-weighted images. Therefore, when correction of distortion and misalignment of other b-value diffusion weighted images is performed using an image with b value = 0 as a reference, a desired correction result may not be obtained due to the difference in image contrast. In this case, there is a possibility that a phenomenon in which the contrast of the lesioned part becomes unclear in a calculation image such as ADC or FA. Moreover, since diffusion imaging repeats measurement under a plurality of conditions, the measurement time becomes longer, and the temperature of the apparatus (gradient magnetic field coil or the like) increases accordingly. The positional deviation and distortion of the diffusion weighted image are easily affected by the heat generated by the apparatus, but these effects are not considered in the conventional correction method.

  Therefore, an object of the present invention is to reduce the contrast difference between images and the influence of heat at the time of acquisition, and improve the accuracy of correction of the diffusion weighted image.

  In order to solve the above-described problem, the present invention calculates an amount of correction between a plurality of images acquired with the same b value by using an image with b value ≠ 0 as a reference image and changing the application axis of the MPG pulse. . At this time, the correction amount of the application axis other than the predetermined application axis is sequentially calculated using a diffusion-weighted image having a predetermined b value (b ≠ 0) of the predetermined application axis as a reference image. Also, a plurality of b-value diffusion weighted images of the same application axis are corrected using the correction amount calculated for the application axis.

  That is, the MRI apparatus of the present invention includes a magnetic field generating unit that generates a static magnetic field, a gradient magnetic field, and a high-frequency magnetic field, and a receiving unit that receives a nuclear magnetic resonance signal from a subject placed in the static magnetic field. An imaging unit that performs a plurality of measurements under different diffusion-sensitive gradient magnetic field conditions and acquires a plurality of measurement data composed of nuclear magnetic resonance signals for each diffusion-sensitive gradient magnetic field condition; and the measurement data acquired by the imaging unit. And a calculation unit that creates a diffusion weighted image, wherein the calculation unit corrects misalignment and distortion between a plurality of diffusion weighted images, and a correction amount for calculating a correction amount used for the correction. A calculation unit. The correction amount calculation unit uses, as a reference image, a diffusion-weighted image having a predetermined b-value (b ≠ 0) of a predetermined application axis for a plurality of diffusion-weighted images having the same b-value of the diffusion-sensitive gradient magnetic field and different application axes. The correction amount of the diffusion weighted image of the predetermined b value of each application axis is calculated by repeating the procedure of sequentially calculating the correction amount of the application axis other than the predetermined application axis, and the correction unit For each of the axes, the diffusion-weighted image having a b value other than the predetermined b value is corrected using the correction amount of the diffusion-weighted image having the predetermined b value for each applied axis.

  According to the MRI apparatus and the diffusion-weighted image correction method of the present invention, it is possible to accurately perform misalignment and distortion correction on a diffusion-weighted image having a different contrast due to a difference in b value. The accuracy of the calculated image can be calculated.

The figure which shows an example of the whole structure of the MRI apparatus with which this invention is applied. An example of the imaging | photography pulse sequence using a diffusion perception gradient magnetic field. FIG. 3 is a functional block diagram of a calculation unit according to the first embodiment. FIG. 3 is a diagram illustrating a processing flow according to the first embodiment. The figure explaining the axis | shaft of a diffusion perception gradient magnetic field, and the measurement order of b value. The figure which shows an example of the condition setting screen of a diffusion weighted imaging. The flowchart which shows the detail of step S205 of FIG. FIG. 3 is a diagram illustrating an image acquired by the method according to the first embodiment. The figure which shows the image acquired with the conventional method. The figure which shows the processing flow of Embodiment 2. FIG. FIG. 10 is a functional block diagram of a calculation unit according to the third embodiment. The figure which shows the processing flow of Embodiment 3. The figure which shows the processing flow of Embodiment 4.

  Hereinafter, embodiments of the MRI apparatus of the present invention will be described in detail with reference to the accompanying drawings. Note that components having the same function are denoted by the same reference symbols throughout the drawings for describing the embodiments of the invention, and the repetitive description thereof is omitted.

  First, an overall outline of an example of an MRI apparatus to which the present invention is applied will be described. As shown in FIG. 1, this MRI apparatus mainly includes a static magnetic field generation unit 2, a gradient magnetic field generation unit 3, a transmission unit 5, a reception unit 6, a signal processing unit 7, a sequencer 4, and a calculation unit. (CPU) 8. In the following description, the static magnetic field generation unit 2, the gradient magnetic field generation unit 3, the transmission unit 5, and the reception unit 6 are collectively referred to as an imaging unit.

  The static magnetic field generation unit 2 includes a permanent magnet type, normal conduction type, or superconductivity type static magnetic field generation source, and generates a uniform static magnetic field in a space where the subject 1 is placed. There are a vertical magnetic field method and a horizontal magnetic field method depending on the direction of the static magnetic field, and the present invention can be applied to either method.

  The gradient magnetic field generating unit 3 drives three sets of gradient magnetic field coils 9 that apply gradient magnetic fields in the three axial directions of X, Y, and Z, which are the coordinate system (stationary coordinate system) of the MRI apparatus, and the respective gradient magnetic field coils. The gradient magnetic field power supply 10 is provided, and the gradient magnetic field power supply 10 of each coil is driven in accordance with a command from the sequencer 4 to be described later, so that the gradient magnetic fields Gx, Gy, Gz in the three axial directions of X, Y, Z Apply. By combining these three-axis gradient magnetic fields, an arbitrary direction can be set on the slice plane (imaging section), and the phase encode gradient can be set in the remaining two directions orthogonal to the slice plane and orthogonal to each other. By applying a magnetic field pulse (Gp) and a frequency encoding gradient magnetic field pulse (Gf), position information in each direction can be encoded in the echo signal. Furthermore, an MPG pulse can be applied to an arbitrary axis by combining a gradient magnetic field. The intensity of the gradient magnetic field pulse is determined by its waveform and is controlled by the sequencer 4.

  The sequencer 4 is a control unit that repeatedly applies a high-frequency magnetic field pulse (hereinafter referred to as “RF pulse”) and a gradient magnetic field pulse in a predetermined pulse sequence. Various commands necessary for data collection are sent to the transmission unit 5, the gradient magnetic field generation unit 3, and the reception unit 6. There are various pulse sequences depending on the imaging method, and they are stored in advance in a storage device or the like. In this embodiment, as a pulse sequence, a diffusion imaging pulse sequence is read from a storage device or the like and executed.

The transmitter 5 irradiates the subject 1 with an RF pulse in order to cause nuclear magnetic resonance to occur in the nuclear spins of the atoms constituting the living tissue of the subject 1, and includes a high frequency oscillator 11, a modulator 12, and a high frequency amplifier. 13 and a high frequency coil (transmission coil) 14a on the transmission side. The RF pulse output from the high-frequency oscillator 11 is amplitude-modulated by the modulator 12 at a timing according to a command from the sequencer 4, and after the amplitude-modulated RF pulse is amplified by the high-frequency amplifier 13, the RF pulse is arranged close to the subject 1. By supplying to the high frequency coil 14a, the subject 1 is irradiated with the RF pulse.
The static magnetic field generation unit 2, the gradient magnetic field generation unit 3, and the transmission unit 5 constitute a magnetic field generation unit in the present invention.

  The receiving unit 6 detects an echo signal (NMR signal) emitted by nuclear magnetic resonance of nuclear spins constituting the biological tissue of the subject 1, and includes a receiving-side high-frequency coil (receiving coil) 14 b and a signal amplifier 15. And a quadrature detector 16 and an A / D converter 17. After the NMR signal of the response of the subject 1 induced by the electromagnetic wave irradiated from the high frequency coil 14a on the transmission side is detected by the high frequency coil 14b arranged close to the subject 1 and amplified by the signal amplifier 15, The signals are divided into two orthogonal signals by the quadrature phase detector 16 at a timing according to a command from the sequencer 4, converted into digital quantities by the A / D converter 17, and sent to the signal processing unit 7.

  The signal processing unit 7 performs various data processing and display and storage of processing results, and includes an external storage device such as an optical disk 19 and a magnetic disk 18 and a display 20 including a CRT. When data from the receiving unit 6 is input to the calculation unit (CPU) 8, the calculation unit 8 executes processing such as signal processing and image reconstruction, and the tomographic image of the subject 1 as a result is displayed on the display 20. The information is displayed and recorded on the magnetic disk 18 of the external storage device.

  The arithmetic unit 8 functions as a control unit that controls the imaging unit via the sequencer 4 in addition to the above-described processing such as image reconstruction. In FIG. 1, the CPU is shown as a specific means for realizing the function of the calculation unit 8, but a part of the function of the calculation unit 8 is ASIC (Application Specific Integrated Circuit), FPGA (Field Programmable Gate Array), and the like. It can also be realized with other hardware. When the CPU is realized, the CPU reads and executes a processing program stored in a memory or the like.

  The operation unit 25 inputs various control information of the MRI apparatus and control information of processing performed by the signal processing unit 7 and the calculation unit 8 and includes input devices such as a trackball or a mouse 23 and a keyboard 24. The operation unit 25 is disposed in the vicinity of the display 20, and an operator controls various processes of the MRI apparatus interactively through the operation unit 25 while looking at the display 20.

  Next, a diffusion imaging pulse sequence targeted by the present invention will be described. The pulse sequence of diffusion imaging is obtained by adding the above-described MPG to a normal imaging sequence. The imaging sequence is not particularly limited, but an echo planar imaging (EPI) pulse sequence in which the influence of MPG is likely to appear clearly is usually used. A typical imaging sequence is shown in FIG. In this sequence, MPG pulses 35 and 36 are embedded in a sequence called spin echo type echo planar imaging (SE-EPI). In the figure, RF, Gs, Ge, and Gr represent an RF pulse, a slice gradient magnetic field, an encode gradient magnetic field, and a read gradient magnetic field, respectively. The EPI sequence is a well-known sequence and will not be described in detail, but is a one-shot SE-EPI that encodes all image information necessary for image reconstruction in one execution of a pulse sequence. There is a split-type (multi-shot) SE-EPI that acquires a piece of image information by repeating a series of pulse sequences a plurality of times while changing the shape of a gradient magnetic field pulse as shown by a dotted line. Any of these may be adopted in the present invention. In multi-slice imaging, image information for a predetermined number of slices is acquired.

  In FIG. 2, the MPG pulses 35 and 36 are applied in the direction of the readout gradient magnetic field Gr, but may be applied to other axes (Ge, Gs) or may be applied to a plurality of axes simultaneously. When applying simultaneously to a plurality of axes, the magnitude of the sum vector of the vectors of each axis is the strength of the gradient magnetic field, and the angle is the angle of the application axis. In diffusion imaging, diffusion of water molecules along the application axis of the MPG pulse is imaged. Therefore, by measuring not only the Gs, Ge, and Gr axes but also the MPG application axis appropriately synthesized, Tissues that suppress the movement of molecules, such as nerve fibers, can be easily depicted.

  The MPG pulses 35 and 36 are generated using the same hardware (gradient magnetic field generator 3) as the gradient magnetic field for imaging, and the pulse intensity (b value) is changed for each measurement by controlling the waveform. Can do. In the present embodiment, a plurality of b values are measured for a plurality of application axes and for each application axis, and a plurality of diffusion weighted images (for example, the number of application axes × the number that makes the b value different) are acquired. The plurality of images may include b value = 0, that is, measurement without applying the MPG pulse. A plurality of diffusion-weighted images obtained by a plurality of measurements with different diffusion-sensitive gradient magnetic field conditions are further used to calculate calculation images such as ADC, FA, and mean kurtosis (MK). In order to improve the accuracy of the calculated image, processing for correcting misalignment and distortion between the images is performed on a plurality of diffusion weighted images.

  These functions are realized by the calculation unit 8. Hereinafter, specific embodiments of functions related to image reconstruction realized by the calculation unit 8 will be described.

<< Embodiment 1 >>
The calculation unit of the present embodiment includes a correction unit that corrects a shift and distortion between a plurality of diffusion-weighted images, and a correction calculation unit that calculates a correction amount used for correction. Furthermore, you may provide the extraction part which extracts the correction amount calculation area | region for calculating a correction amount. The correction amount calculation unit is configured to perform diffusion enhancement of a predetermined b value (b ≠ 0) of a predetermined application axis for a plurality of diffusion weighted images having the same b value of the diffusion-sensitive gradient magnetic field and different application axes. The correction amount of the diffusion weighted image of the predetermined b value of each application axis is calculated by repeating the procedure of sequentially calculating the correction amount of the application axis other than the predetermined application axis using the image as a reference image, and the correction The unit corrects a diffusion-weighted image having a b value other than the predetermined b value by using the correction amount of the diffusion-weighted image having the predetermined b value on each application axis.

  FIG. 3 shows a functional block diagram showing functions of the calculation unit 8 of the present embodiment, particularly functions related to image reconstruction. As shown in the figure, the calculation unit 8 includes a control unit 80, an image reconstruction unit 81, an extraction unit 82, a correction amount calculation unit 83, a correction unit 85, a calculated image calculation unit 87, and a display control unit 89. The control unit 80 controls the imaging unit and each functional unit in the calculation unit. The image reconstruction unit 81 performs an image reconstruction operation such as Fourier transform on measurement data collected for each measurement with different diffusion-sensitive gradient magnetic field conditions (application axis, b value), and creates a diffusion-weighted image. The extraction unit 82 extracts a correction amount calculation area, that is, a correction amount calculation area, from each diffusion weighted image. The correction amount calculation unit 83 calculates the correction amount of another diffusion weighted image using one of the plurality of diffusion weighted images as a reference. An image having a b value other than zero is used as a reference. Preferably, an image having a minimum b value among a plurality of b values is used. The correction unit 85 corrects misalignment and distortion of a plurality of diffusion weighted images using the correction amount calculated by the correction amount calculation unit 83. The calculated image calculation unit 87 calculates a calculated image using the corrected diffusion weighted image. The display control unit 89 creates an image created by the calculation unit 8 or a display image such as a GUI for prompting the operator to input, and causes the display 20 to display the image.

  Based on the above configuration, the flow of processing by the MRI apparatus of this embodiment will be described with reference to the processing flow shown in FIG.

<Step S201>
After setting up the subject 1 in the examination space of the static magnetic field generation unit 2, imaging conditions for imaging each diffusion weighted image of a plurality of b values including b value = 0 are received via the operation unit 25. When the calculation unit 8 receives the imaging condition, the calculation unit 8 calculates a gradient magnetic field waveform used in each measurement in the diffusion weighted imaging, and passes the result to the sequencer 4.

<Step S202>
Under the control of the sequencer 4, the imaging unit (the transmission unit 5, the gradient magnetic field generation unit 3, and the reception unit 6) operates to perform imaging according to the diffusion weighted pulse sequence under the imaging conditions set in step S 201. Diffusion-weighted image data (measurement data) is acquired on the same cross section of the subject. The calculation unit 8 receives the measurement data, and the image reconstruction unit 81 creates a plurality of b-value diffusion weighted images including b value = 0.

  The order of measurement of a plurality of diffusion-sensitive gradient magnetic field conditions is not particularly limited. For example, a plurality of measurements are performed while changing the b value on the first MPG application axis, and then the b value is changed on another MPG application axis. Make multiple measurements. In the following description, the first application axis in the measurement order is called the first axis, the second application axis is called the second axis, and so on. The order of changing the b value may be changed, for example, from a small b value (Low-b) to a large b value (High-b), or vice versa. FIG. 5 schematically shows an example of the measurement order of the application axis and b value. In FIG. 5, the radial arrows indicate the MPG axes, and the numbers attached to the arrows indicate the measurement order. The axis of the MPG in FIG. 5 is not related to the arrangement of the real space and may be any arrangement in space. The radial axis of each radiation is the b value axis, and shows that the b value increases from the center toward the outside. Here, the numbers representing the measurement orders 1 to 6 are assigned to the first and second axes. As shown by the enclosing figure, the case of measuring from High-b to Low-b on each axis is shown.

<Step S203>
An operation for executing correction of the diffusion weighted image is received via the operation unit 25. In this step, it is accepted whether or not correction is to be executed. When correction is to be performed, an input of a b value as a reference is accepted. FIG. 6 shows an example of a screen (GUI) for accepting the operation of the operator. In the example shown in the figure, an image display block 61 such as a diffusion weighted image (DWI) created in step S202, a block 62 for displaying a menu for selecting an image to be displayed on the image display block 61, and a block for performing settings related to correction. 63, a block 64 for displaying a menu for selecting various calculation images to be created. In the block 63, a check box 631 for selecting distortion or misalignment correction is displayed.

  The b value may be preset as a default, or a box for inputting the b value may be displayed. In the latter case, when correction is selected by the check box 631, a box for inputting the b value may be displayed. The b value used as a reference is set to a value other than zero in order to prevent a correction error due to a contrast difference with other b-value images. Furthermore, in order to correct distortion due to eddy current with high accuracy, it is preferable to set the minimum value where the influence of eddy current is the smallest. Further, when the imaging time is long, the b value of the axis that is imaged earliest is suitable for the positional deviation of the image due to the heat of the apparatus because the influence of the heat is the smallest. Further, in an MRI apparatus using a permanent magnet, image distortion and misalignment due to a residual magnetic field occur, and the influence of the residual magnetic field increases as the b value increases, so the lowest value is preferable.

  For example, the operator can check the diffusion weighted image displayed on the image display block 61 and determine whether correction is necessary. However, the distortion and misalignment correction may be automatically executed instead of being selected via the operation unit 25. The correction execution setting may be performed in step S201.

<Step S204>
The calculation unit 8 creates a reference image by using the b-value specified in step S203 or the b-value set by default among the plurality of diffusion-weighted images created in step S202. Here, by using a diffusion weighted image S ₁ of the smallest b value of the first axis in the measurement order (where b value ≠ 0), an example of making the reference image R.

Reference image R is is also possible to use as a diffusion weighted image S _1, for the present embodiment to increase the accuracy of the calculation of the correction amount will be described later, the extraction unit 82 is the reference image was extracted regions . The region to be extracted is preferably a region that is not affected by the diffusion of water molecules. For example, in the case of an image of the head, it is a brain parenchymal region. The extraction method is not particularly limited, and known threshold processing such as discriminant analysis, region extraction processing, and machine learning processing can be used.

  Here, the reference image used the diffusion-weighted image of the first axis with the minimum b value excluding the b value = 0 as the preferred b value. However, the reference image is an image with a smaller degree of positional deviation and distortion from the b value = 0. If it exists, it may be another axis or another image of another b value.

<Step S205>
Correction amount calculation unit 83, the second axis and subsequent diffusion weighted images, by using the diffusion-weighted images of the same b values and diffusion weighted image S ₁ that created the reference image R, the correction amount for performing the position shift distortion correction Is calculated and corrected. Details of this step are shown in FIG.

[Step S2051]
First, the diffusion weighted image S ₂ of the second axis, to extract the same region as the reference image A, to create a region extraction image B.

[Step S2052]
A positional deviation amount and a distortion amount of the region extraction image B with respect to the reference image A are calculated. For example, as shown in the following equation (1), the displacement amount and the distortion amount can be calculated as a shift amount or a conversion matrix for matching the extracted image B with the reference image A. The shift amount and the transformation matrix are calculated so that the index values such as (Dice's similarity coefficiency) and mutual information (MI) are maximized.

[Equation 1]
A = c × B + d (1)
In the formula, c represents a transformation matrix (distortion), and d represents a shift amount. A is a reference image, and B is an extracted image obtained by extracting a correction amount calculation region from a correction target image.

  For example, in the correction using the dice coefficient, the shift amount (d in the expression (1)) is changed so as to maximize the dice coefficient s (simularity) represented by the following expression (2), and the calculation is repeatedly performed.

式中、ｓはダイス係数、Ｂは拡散強調画像Ｓ_２の領域抽出画像、Ａはリファレンス画像である（以下、同じ）。

Wherein, s is a die coefficient, B is the diffusion weighted image S ₂ of the region extraction image, A is the reference image (hereinafter, the same).

ここで、ａ、ｂは画像Ａ、Ｂの画素値、ｍ、ｎは画像Ａ、Ｂの階調数を示す。p(a_i, b_j)は、a_i, b_jが同時におこる確率、p(a_i), p(b_j)は、a_i , b_jがそれぞれおこる確率を示す。 In the correction using the MI, a correction condition that maximizes the MI represented by the following equation (3) is calculated.

Here, a and b are the pixel values of the images A and B, and m and n are the number of gradations of the images A and B. p (a _i , b _j ) indicates the probability that a _i and b _j occur simultaneously, and p (a _i ) and p (b _j ) indicate the probability that a _i and b _j occur, respectively.

  Specifically, the correction condition is calculated by the following procedure. First, a predetermined correction process is performed on the distortion / position correction target image B to calculate MI. Further, the correction condition is changed to calculate MI. The correction condition (correction amount) that maximizes the MI is obtained by repeating the change of the correction condition and the calculation of the MI.

[Steps S2053 and S2054]
The correction amount calculation unit 83 calculates the correction amount by repeating steps S2051 and S2052 for the diffusion-weighted images on and after the third axis.

[Step S2055]
Correcting unit 85, the calculated shift amount and the transformation matrix in step S2052, is applied to the diffusion weighted image of the second axis of the diffusion weighted image S ₂ and b values other than the b value, the position correction, performs distortion correction . Similarly, with respect to the diffusion-weighted images on and after the third axis, the shift amount and the conversion matrix calculated in S2053 are applied to the diffusion-weighted image of each b value on that axis to perform position correction and distortion correction.

  When steps S2051 to S2055 are completed for all measured MPG axes, the correction process (FIG. 4: S205) is completed. Note that steps S2052 to S2054 and step S2055 may be performed in parallel. That is, after the correction amount for the second axis is calculated, the correction for the second axis may be performed, and then the correction amount calculation / correction for the third axis may be performed.

In this way the step S205, the correction amount calculation unit 83 and correction unit 85, the extracted shift amount of diffusion weighted image S ₂ calculated in step S2053, based on the transformation matrix, the imaging order of the same value of b 3 Axis to diffusion weighted images S ₃ since, misalignment, performs distortion correction. That is, the positional deviation and distortion of the diffusion weighted images of all axes are corrected using the correction amount calculated for the diffusion weighted image of the previous axis as an initial value.

  For the coaxial diffusion-weighted image, the diffusion-weighted image having a high b value (High-b) is corrected using a correction amount calculated with a low b value (Low-b). Thereby, correction can be performed with high accuracy. For example, a diffusion-weighted image such as High-b is different from a diffusion-weighted image having a contrast of Low-b, and thus there is a case where the correction amount cannot be calculated correctly by the same processing as Low-b. Further, when the measurement data of Low-b and High-b are continuously measured with the same application axis, the positional deviation and distortion correction are small between the coaxial Low-b and High-b from the measurement order. Accordingly, the correction accuracy can be maintained without calculating the correction amount for all the b-value images.

  In the above description, in step S205, the correction amount calculated for the diffusion weighted image with different b values on the same axis is used as the correction amount for the diffusion weighted image with other b values. However, the present embodiment is not limited to this. Instead, the correction amount of the image on the other axis may be used as long as the image has the same degree of displacement and distortion.

  In step S205, the shift amount and the distortion amount are calculated using the index after performing the region extraction. However, the calculation process may be performed as the image without performing the region extraction. In this case, the extraction unit 82 shown in FIG. 3 can be omitted. Further, the correction amount calculation method may be analysis using publicly known machine learning or analysis using other features, instead of maximizing the index.

<Step S206>
After performing steps S204 and S205 described above for the required number of slices, the calculated image calculation unit 97 uses the corrected diffusion weighted image and the diffusion weighted image with b value = 0 to generate various calculated images. calculate. The calculated image is an image in which calculated values such as apparent diffusion coefficient (ADC), diffusion anisotropy (FA), or average kurtosis (MK) are calculated for each pixel, and exponential function fitting (for example, Mono Exponential Fitting, Bi Exponential). It can be calculated based on a known diffusion model such as Fitting and diffusion kurtosis imaging.

<Step S207>
The display control unit 89 displays the calculation image calculated in step S206 on the display 20. At this time, for example, as shown in FIG. 6, the diffusion-weighted image created in step S202 or the corrected diffusion-weighted image may be displayed together.

  As described above, in Embodiment 1 of the present invention, the calculation unit 8 performs diffusion of a predetermined b value (b ≠ 0) from a plurality of b-value diffusion weighted images in each axial direction including b value = 0. Using the enhanced image as a reference, the position and distortion of the diffusion enhanced image on the other axis are corrected. At this time, the correction amount of the previous axis in the imaging order is set as an initial value, and the correction amount of the next axis in the imaging order is calculated. Further, the correction amount calculated for each axis is applied to the diffusion-weighted image of the b value in addition to the same axis to correct the position and distortion. After that, various calculation images are calculated. By such a method, an image with b value = 0 is used as a reference to correct a diffusion weighted image with each axis and each b value, and an image with clearer tissue contrast is calculated as compared with the calculated image. Can do.

  The brain diffusion weighted image obtained by the method of the present embodiment and the image of the same part obtained by the conventional method (comparative example) are shown in FIGS. 8 and 9, respectively. Here, as an example, a DWI image with b value = 0, a DWI image with b value = 1000, and an MK image are shown. In the DWI of the comparative example (b value = 1000), blurring of the image due to misalignment is seen around the brain parenchyma, and when the calculated image calculated based on it is compared with the calculated image obtained by this embodiment, It can be seen that the image of this embodiment can more accurately depict diffusion information around the brain parenchyma.

  Thus, according to this embodiment, it is possible to obtain a diffusion-weighted image with high contrast by correcting the diffusion-weighted image using a b value other than zero. In particular, the accuracy of the calculated image can be improved even when the degree of displacement and distortion is large due to the influence of heat or the like in the high magnetic field MRI apparatus.

  In the above description, the correction method using a two-dimensional image has been described, but a three-dimensional image may be used.

<< Embodiment 2 >>
In the first embodiment, the case where the correction amount is calculated for the diffusion-weighted images of all the measured MPG application axes has been described. However, in the present embodiment, only a part of the measured MPG application axes is corrected. The feature is that the amount is calculated and the other axes are corrected using the correction amounts calculated for some of the axes.

  Also in the present embodiment, the configuration of the apparatus is the same as that of the first embodiment, but the function of the calculation unit 8 is different. Hereinafter, this embodiment will be described with a focus on processes different from those of the first embodiment, with appropriate reference to the drawings used in the first embodiment. The processing flow of this embodiment is shown in FIG.

<Step S201 to Step S204>
The processing of these steps is the same as the processing of S201 to S204 of the first embodiment, and the process from setting of imaging conditions to imaging, acquisition of a diffusion weighted image under each diffusion-sensitive gradient magnetic field condition, and creation of a reference image is performed.

<Step S301>
The correction amount calculation unit 83 selects (extracts) one or more images for calculating the correction amount from the plurality of diffusion weighted images created in step S202, and calculates the correction amounts. As an image selection method, for example, an MPG application axis that is an odd number in the measurement order or a 3n + 1th MPG application axis (n is an integer of 1 or more) (its image) is selected, and an image of the application axis is selected. Then, an image having the same b value as that of the diffusion weighted image in which the b value is created as the reference image is selected. Alternatively, a method may be used in which images having similar distortion and positional deviation are clustered using a machine learning process or the like and a part thereof is extracted.

<Step S302>
The calculation of the correction amount is the same as that in the first embodiment (step S205). For example, the first correction amount is calculated in the order of measurement, starting from the diffusion weighted image closest to the measurement order of the reference image, and then the initial correction amount is calculated. The correction amount is sequentially calculated by repeating the procedure of calculating the correction amount of the diffusion-enhanced image in the next order as the value (precisely, the diffusion-weighted image corrected with the correction amount is the initial value). At this time, for example, a correction amount for maximizing the index is obtained using an index such as a dice coefficient or a mutual information amount. When the dice coefficient is used, it is preferable to perform the region extraction described in step S <b> 204 of the first embodiment on the reference image and the diffusion weighted image for which the correction amount is obtained. Thereby, the accuracy of the correction amount calculation can be increased.

<Step S303>
The correction unit 85 further applies the correction amount calculated in step S to the image of the other application axis for which the correction amount is not calculated. For example, when the correction amount is calculated for the odd-numbered application axis, the correction amount is applied to the image of the even-numbered application axis. The even-numbered application axis may be preceded in order or may be subsequent. When every other application axis is selected, the correction amount with the closest measurement order is applied. When clustering, the same correction amount is applied to images clustered in the same classification.

<Step S304>
The correction unit 85 corrects all b-value images of the corresponding application axis using the correction amount of the application axis calculated in steps S302 and S303.

<Steps S206 and S207>
Similar to steps S206 and S207 of the first embodiment, a calculated image is calculated using the corrected diffusion weighted image and displayed on the display 20.

  As described above, in the present embodiment, the correction amount is calculated only for some diffusion-weighted images, and correction is performed on the other diffusion-weighted images using the calculated correction amounts. According to the present embodiment, as in the first embodiment, the diffusion weighted image can be corrected without depending on the diffusion weighted image with b value = 0, and the contrast of the calculated image can be improved. Also, images with close measurement order and clustered images have the same distortion and displacement caused by heat, so the correction amount calculated for one of these images can be applied to other images. The calculation cost for calculating the correction amount can be reduced.

<< Embodiment 3 >>
In the present embodiment, the correction amount of each applied axis is calculated sequentially using a reference image created from an image having a predetermined b value on a predetermined applied axis, as in the first or second embodiment. In the present embodiment, whether or not the correction amount is appropriate is determined, and the correction amount determined to be inappropriate is substituted with the correction amount calculated for the image of the other axis.

  The function and processing of the calculation unit 8 of this embodiment will be described with reference to the block diagram of FIG. 11 and the flow of FIG. As shown in FIG. 11, the calculation unit 8 of the present embodiment has a determination unit 84 that determines whether the correction amount calculated by the correction amount calculation unit 83 is appropriate. Other than that, the correction unit 85 is the same as the calculation unit 8 of the first embodiment illustrated in FIG. 3, but the correction unit 85 determines the correction amount applied to each image according to the result of the determination unit 84. Either the amount or the correction amount calculated for another image is used.

Hereinafter, the processing flow of this embodiment will be described with reference to FIG.
<Step S201 to Step S204>
The processing of these steps is the same as the processing of S201 to S204 of the first embodiment, and the process from setting of imaging conditions to imaging, acquisition of a diffusion weighted image under each diffusion-sensitive gradient magnetic field condition, and creation of a reference image is performed.

<Step S401>
The correction amount calculation unit 83 calculates the correction amount by using the method of the first embodiment or the second embodiment with respect to an image having an application axis different from the application axis of the diffusion weighted image that has created the reference image and the same b value.

<Step S402>
The determination unit 84 determines whether the calculated correction amount is appropriate. It is determined whether the correction amount is appropriate within a preset reference value. For example, in the case of the amount of misalignment (B in formula (1)), the number of pixels and the ratio to the entire pixel are used as a criterion. When the pixel exceeds 10 pixels or the ratio to the entire pixel exceeds a predetermined ratio (%) Can be determined as abnormal (unsuitable). Further, when the correction amount calculation unit calculates using an index such as a dice coefficient or MI, it is determined as abnormal if the correction amount that maximizes the index cannot be calculated even if the number of repetitions of the calculation exceeds a predetermined number. Also good.

  The processing by the determination unit 84 may be performed after the correction amount calculation unit 83 has calculated all the correction amounts, but may be sequentially performed in parallel with the processing of the correction amount calculation unit 83.

<Steps S403 and S404>
The correction unit 85 corrects the image using the correction amount determined to be within the reference value (appropriate) among the correction amounts of the images of the application axes calculated by the correction amount calculation unit 83 (S403). . In addition, the image of the application axis in which the correction amount is determined to be inappropriate or abnormal is corrected using an appropriate correction amount calculated for other images (S404). The other image is not limited, but an image of the application axis that is close to the measurement order is preferable. Instead of using the correction amounts of the other axes, the average value or the median value of the correction amounts of the respective axes may be used.
<Step S405>
For all b-value images on each axis, correction is performed using the correction amount calculated for each applied axis.

<Steps S206 and S207>
Similar to steps S206 and S207 of the first embodiment, a calculated image is calculated using the corrected diffusion weighted image and displayed on the display 20.

  As described above, in the present embodiment, whether or not the correction amount calculated by the correction amount calculation unit is appropriate is determined, and the correction amount of the axis whose correction amount is determined to be appropriate is determined for the axis that is determined to be inappropriate. Is used to correct distortion and misalignment. According to the present embodiment, in addition to the same effects as those of the first embodiment or the second embodiment, the accuracy of correction can be improved by replacing an inappropriate correction amount with another correction amount, and more accurate calculation can be performed. An image can be calculated.

<< Embodiment 4 >>
In the present embodiment, similarly to the third embodiment, the calculation unit 8 includes a unit (determination unit) that determines whether the correction amount is appropriate. However, in the third embodiment, the correction amount determined to be inappropriate is substituted with the correction amount calculated for the image of the other axis, whereas in the present embodiment, a display for prompting remeasurement is performed. .

The process of this embodiment is demonstrated with reference to the flow of FIG.
<Step S201 to Step S204>
The processing of these steps is the same as the processing of S201 to S204 of the first embodiment, and the process from setting of imaging conditions to imaging, acquisition of a diffusion weighted image under each diffusion-sensitive gradient magnetic field condition, and creation of a reference image is performed.

<Steps S401 and S402>
The processing in this step is the same as the processing in the third embodiment and the processing in steps S401 and S402, and the correction amount calculation unit 83 calculates the correction amount and the determination unit 84 determines whether the correction amount is appropriate. The reference value for determining the correction amount may be the same as or different from the reference value in the third embodiment.

<Step S501>
The determination unit 84 passes an application axis with an inappropriate correction amount as a determination result to the control unit 80 that controls the imaging unit. Or you may display on the display 20 via the display control part 89. FIG. The display control unit 89 displays a GUI for instructing re-imaging on the display 20 together with the determination result. Thereby, the operator can determine whether or not to re-image by looking at the determination result.

<Step S502>
When the control unit 80 receives a determination result from the determination unit 84 or receives an instruction from the operator via the GUI, the control unit 80 executes re-imaging. The target of re-imaging may be, for example, only a b-value MPG axis diffusion-weighted image with a large correction amount, that is, a large deviation amount, or may be a plurality of images including images on the previous axis in the imaging order. In addition, when the ratio of the axes for which the correction amount is inappropriate is large or every determination, all diffusion weighted images may be re-imaged. At this time, based on the calculated correction amount value, the imaging order may be changed so that the correction amount does not become abnormal.

<Steps S403 and S405>
When re-imaging is performed for only a part of the MPG axes, the above-described steps S401 and S402 are performed on the diffusion-weighted image obtained thereby, and when all the diffusion-weighted images are re-imaged. Returning to step S202, the processing up to step S502 is repeated, and in all cases, if all of the calculated correction amounts are appropriate, the same correction as in steps S403 and S405 of the third embodiment is performed, and the process proceeds to step S206. move on.

<Steps S206 and S207>
Similar to steps S206 and S207 of the first embodiment, a calculated image is calculated using the corrected diffusion weighted image and displayed on the display 20.

  Although the processing of the fourth embodiment has been described above, the fourth embodiment can be combined with the third embodiment. In this case, the determination unit 84 sets, for example, two reference values with different deviation amounts, a first reference value, and a second reference value as reference values for determining the correction amount. And about the axis | shaft in which the correction amount calculated by the correction amount calculation part 83 exceeded 1st reference value, the process of step S403 of Embodiment 3 is performed, and it correct | amends using the correction amount of another axis | shaft. When the correction amount calculated by the correction amount calculation unit 83 exceeds the first reference value, the processing of steps S501 and S502 of the fourth embodiment is performed, and the axis determined to be inappropriate or a plurality of axes including the same Take an image.

  As described above, in the present embodiment, it is determined whether the correction amount calculated by the correction amount calculation unit is appropriate, and when the correction amount is determined to be inappropriate, the axis for which the correction amount is determined to be inappropriate, or Re-imaging is performed for a plurality of axes including the whole axis or all axes. According to the present embodiment, in addition to the same effects as those of the first embodiment or the second embodiment, the correction accuracy can be improved by eliminating the correction with an inappropriate correction amount, so that a calculation image with higher accuracy can be obtained. Can be calculated.

  As mentioned above, although embodiment of this invention was described, this invention is not limited to embodiment shown by the said embodiment and drawing used for the description. For example, as long as there is no technical contradiction, a plurality of embodiments can be combined, and elements that are not essential to the embodiments can be deleted or added.

1: subject, 2: static magnetic field generation unit, 3: gradient magnetic field generation unit, 4: sequencer, 5: transmission unit, 6: reception unit, 7: signal processing unit, 8: calculation unit, 9: gradient magnetic field coil, 10: Gradient magnetic field power supply, 11: High frequency oscillator, 12: Modulator, 13: High frequency amplifier, 14a: High frequency coil (transmitting coil), 14b: High frequency coil (receiving coil), 15: Signal amplifier, 16: Quadrature phase detector 17: A / D converter, 18: magnetic disk, 19: optical disk, 20: display, 21: ROM, 22: RAM, 23: mouse, 24: keyboard, 25: operation unit, 80: control unit, 81: Image reconstruction unit, 82: extraction unit, 83: correction amount calculation unit, 84: determination unit, 85: correction unit, 87: calculation image calculation unit, 89: display control unit

Claims (12)

A magnetic field generation unit that generates a static magnetic field, a gradient magnetic field, and a high-frequency magnetic field, and a reception unit that receives a nuclear magnetic resonance signal from a subject placed in the static magnetic field, under different diffusion-sensitive gradient magnetic field conditions An imaging unit that performs a plurality of measurements and acquires a plurality of measurement data composed of nuclear magnetic resonance signals for each diffusion-sensitive gradient magnetic field condition;
A calculation unit that creates a diffusion-weighted image using the measurement data acquired by the imaging unit;
The calculation unit includes a correction unit that corrects misalignment and distortion between a plurality of diffusion-weighted images, and a correction amount calculation unit that calculates a correction amount used for the correction,
The correction amount calculation unit uses, as a reference image, a diffusion-weighted image having a predetermined b-value (b ≠ 0) of a predetermined application axis for a plurality of diffusion-weighted images having the same b-value of the diffusion-sensitive gradient magnetic field and different application axes. The correction amount of the diffusion weighted image of the predetermined b value of each application axis is calculated by repeating the procedure of sequentially calculating the correction amount of the application axis other than the predetermined application axis,
The correction unit corrects a diffusion-weighted image having a b value other than a predetermined b value by using a correction amount of the diffusion-weighted image having the predetermined b value on each application axis for each applied axis. Magnetic resonance imaging apparatus. The magnetic resonance imaging apparatus according to claim 1,
2. The magnetic resonance imaging apparatus according to claim 1, wherein the predetermined b value is a minimum value other than 0 among a plurality of b values set on the same application axis. The magnetic resonance imaging apparatus according to claim 1,
The magnetic resonance imaging apparatus, wherein the correction amount calculation unit repeats a procedure for calculating the correction amount in accordance with a measurement order of each application axis. The magnetic resonance imaging apparatus according to claim 1,
The calculation unit further includes an extraction unit that extracts a correction amount calculation region from the diffusion weighted image,
The magnetic resonance imaging apparatus, wherein the correction amount calculation unit calculates a correction amount that maximizes the similarity of the correction amount calculation region extracted by the extraction unit. The magnetic resonance imaging apparatus according to claim 1,
The correction amount calculation unit calculates a correction amount for a part of the plurality of application axes,
The magnetic resonance imaging apparatus, wherein the correction unit corrects a diffusion weighted image of an application axis other than the partial application axis using a correction amount calculated for the partial application axis. The magnetic resonance imaging apparatus according to claim 5,
The part of the application axes is an application axis extracted for each of a plurality of application axes in the measurement order of the application axes. The magnetic resonance imaging apparatus according to claim 1,
The magnetic resonance imaging apparatus, wherein the calculation unit further includes a determination unit that determines whether the correction amount calculated by the correction amount calculation unit is appropriate. The magnetic resonance imaging apparatus according to claim 7,
The correction unit is a target of correction using a correction amount of another application axis determined to be appropriate when the correction amount of the application axis that is a correction target is determined to be inappropriate by the determination unit. A magnetic resonance imaging apparatus for correcting a diffusion weighted image of an application axis. The magnetic resonance imaging apparatus according to claim 7,
The magnetic resonance imaging apparatus, wherein the calculation unit performs control to re-measure the imaging unit when the correction amount of the application axis to be corrected is determined to be inappropriate by the determination unit. A plurality of measurements with different diffusion-sensitive gradient magnetic field application axes and b values are made, and a plurality of diffusion-weighted images having different diffusion-sensitive gradient magnetic field conditions are created using the obtained measurement data. A method for correcting misalignment and distortion between
For a plurality of diffusion-weighted images having the same b-value of the diffusion-sensitive gradient magnetic field and different application axes, a diffusion-weighted image having a predetermined b-value (b ≠ 0) of the predetermined application axis is used as a reference image, and other than the predetermined application axis The correction amount of the diffusion weighted image of the predetermined b value of each application axis is calculated by repeating the procedure of sequentially calculating the correction amount of the application axis of
A diffusion-weighted image in which a diffusion-weighted image having a b value other than a predetermined b value is corrected using the correction amount of the diffusion-weighted image having the predetermined b value on each application axis. Correction method. The method of correcting a diffusion weighted image according to claim 10,
A method for correcting a diffusion-weighted image, wherein the correction amount is calculated in parallel with the measurement of each application axis in accordance with the measurement order of the application axis. A method of correcting misalignment and distortion in a plurality of diffusion-weighted images having different diffusion-sensitive gradient magnetic field conditions obtained by a plurality of measurements with different application axes and b values of the diffusion-sensitive gradient magnetic field,
For a plurality of diffusion-weighted images having the same b-value of the diffusion-sensitive gradient magnetic field and different application axes, a diffusion-weighted image having a predetermined b-value (b ≠ 0) of the predetermined application axis is used as a reference image, and other than the predetermined application axis The correction amount of the diffusion weighted image of the predetermined b value of each application axis is calculated by repeating the procedure of sequentially calculating the correction amount of the application axis of
A diffusion-weighted image in which a diffusion-weighted image having a b value other than a predetermined b value is corrected using the correction amount of the diffusion-weighted image having the predetermined b value on each application axis. Correction method.

Images