JP6752064B2 - Magnetic resonance imaging device, image processing device, and diffusion-weighted image calculation method - Google Patents

Description

The present invention relates to a magnetic resonance imaging (hereinafter referred to as “MRI”) device that performs imaging using an MPG (Motion Diffusion Gradient), particularly under conditions in which the b value, which is a parameter representing the strength of MPG, is different. The present invention relates to a technique for acquiring a plurality of diffusion-weighted images and obtaining a diffusion-weighted image having an arbitrary b value by using the plurality of diffusion-weighted images.

Diffusion-weighted imaging is one of the imaging methods using an MRI apparatus. Diffusion-weighted imaging uses high-intensity gradient pulsed fields that change the signal intensity from the diffusion spins. This gradient magnetic field pulse is called an MPG (diffusion-sensitive gradient magnetic field) pulse, and by performing a plurality of imaging with variously different intensities and application axes, diffusion along a specific fiber such as diffusion in the brain is performed. Can be visualized (tractography) to display the directionality of fibers (Non-Patent Document 1).

In general, the larger the b value, the more diffusion-weighted images are obtained, and it is possible to calculate the diffusion-weighted coefficient (ADC: Apparent Diffusion Co-efficient) for each pixel from a plurality of diffusion-weighted images having different b values. it can.

In order to obtain various diagnostic images including the diffusion-weighted coefficient and the above-mentioned tractography from the diffusion-weighted image obtained by diffusion-weighted imaging, it is necessary to image with various changes in the intensity and axis of the MPG pulse. It involves a significant increase in imaging time. Further, in the low magnetic field MRI apparatus, the intensity of MPG pulses that can be used is limited, and a diffusion-weighted image having a high b value cannot be obtained. To solve this problem, a method has been proposed in which actual imaging is performed with a small number of b-values and a desired diffusion-weighted image with a b-value is calculated from the result.

For example, in the method disclosed in Patent Document 1, two or more original images (diffusion-weighted images) are obtained by imaging each imaging region using two or more b values, and the apparent image is seen from two or more original images. Calculate the diffusion coefficient (ADC). An arbitrary cutoff diffusion coefficient is set for this ADC, and a diffusion-weighted image having an arbitrary b value is calculated using an ADC having a cutoff diffusion coefficient or more. According to this method, by setting the cutoff diffusion coefficient, it is possible to reduce the error of the diffusion coefficient due to noise, and to calculate a diffusion-weighted image with less noise and clear contrast of the lesion. Hereinafter, the apparent diffusion coefficient (ADC) is simply referred to as a diffusion coefficient (ADC) unless a misunderstanding occurs.

Japanese Unexamined Patent Publication No. 2016-102

Masutani Y, Aoki S, Abe O, et al; MR diffusion tensor imaging recent advance and new techniques for diffusion tensor visualization. Eur J Radiol 46: 53-66, 2003.

However, in the method of Patent Document 1, since a method such as excluding pixels having a cutoff diffusion coefficient or less is excluded from the calculation target, pixel loss occurs in the virtual diffusion-weighted image calculated from the noisy diffusion-weighted image. there is a possibility.

Therefore, an object of the present invention is an MRI apparatus and a diffusion-weighted image calculation method capable of calculating a diffusion-weighted image of an arbitrary b value without eliminating the influence of noise in the calculation of the diffusion coefficient and causing pixel loss. Is to provide.

In order to achieve the above object, the MRI apparatus and the diffusion-weighted image calculation method of the present invention analyze each image acquired with a plurality of b values (including the case where b value = 0), and analyze the diffusion coefficient. Analyze the presence or absence of pixels (abnormal pixels) in which is an abnormal value. If there are abnormal pixels, they are extracted and corrected using the pixel values of each image. The corrected image is used, for example, for the calculation of the diffusion coefficient and the calculation of the diffusion-weighted image (virtual diffusion-weighted image) at an arbitrary b value using the calculated diffusion coefficient.

That is, the MRI apparatus of the present invention uses an imaging unit that performs a plurality of imaging under a plurality of different MPG pulse application conditions and acquires a nuclear magnetic resonance signal for each imaging, and a nuclear magnetic resonance signal acquired by the imaging unit. An image reconstruction unit for reconstructing an image and a calculation unit for calculating a diffusion-weighted image using a plurality of the images having different MPG pulse application conditions are provided. The calculation unit has an analysis unit that analyzes a plurality of images having different MPG pulse application conditions and determines the presence or absence of abnormal pixels, and uses the plurality of images corrected based on the results of the analysis unit to perform diffusion-weighted images. Is calculated. The phrase "using the plurality of corrected images" includes not only the case of using the directly corrected image but also the case of using the result calculated from the corrected image (for example, the diffusion coefficient). The diffusion-weighted image calculated by the calculation unit includes, for example, a diffusion-weighted image (ADC) map or a virtual diffusion-weighted image at an arbitrary b value.

According to the present invention, the apparent diffusion coefficient (ADC) can be calculated accurately without being affected by noise and without causing pixel loss. As a result, the calculation accuracy of the diffusion-weighted image can be improved at an arbitrary b value.

The figure which shows an example of the whole structure of the MRI apparatus which concerns on this invention. An example of a shooting pulse sequence using MPG pulses. The functional block diagram of the arithmetic processing part of 1st Embodiment. The flowchart which shows the processing flow of 1st Embodiment. The figure which shows an example of GUI used for the processing of 1st Embodiment. The figure which shows the flow of the process of 1st Embodiment and the image obtained by each process. The figure which shows the process flow of the comparative example and the image obtained by each process. The flowchart which shows the processing flow of the modification of 1st Embodiment. The flowchart which shows the processing flow of 2nd Embodiment. The figure which shows an example of GUI used for the processing of 2nd Embodiment. The flowchart which shows the processing flow of the modification of the 2nd Embodiment. The functional block diagram of the arithmetic processing part of 3rd Embodiment. The flowchart which shows the processing flow of 3rd Embodiment. The figure explaining the fitting in 3rd Embodiment. The flowchart which shows the processing flow of 4th Embodiment.

Hereinafter, preferred embodiments of the MRI apparatus of the present invention will be described in detail with reference to the accompanying drawings. In all the drawings for explaining the embodiment of the invention, those having the same function are designated by the same reference numerals, and the repeated description thereof will be omitted.

First, an overall outline of an example of the MRI apparatus according to the present invention will be described with reference to FIG. FIG. 1 is a block diagram showing an overall configuration of an embodiment of an MRI apparatus according to the present invention. This MRI apparatus obtains a tomographic image of a subject by utilizing an NMR phenomenon, and includes a static magnetic field generating unit 2, a gradient magnetic field generating unit 3, a transmitting unit 5, a receiving unit 6, and a signal processing unit 7. , The sequencer 4 and the arithmetic processing unit (CPU) 8 are provided.

The static magnetic field generating unit 2 generates a uniform static magnetic field in the space where the subject 1 is placed, and is composed of a permanent magnet type, a normal conducting type, or a superconducting type static magnetic field generating source (static magnetic field generating magnet). The static magnetic field generation source includes a perpendicular magnetic field method that generates a static magnetic field in a direction orthogonal to the body axis of the subject 1 and a horizontal magnetic field method that generates a static magnetic field in the body axis direction of the subject 1. Can also be applied to.

The gradient magnetic field generating unit 3 includes a gradient magnetic field coil 9 that applies a gradient magnetic field in the three axes of X, Y, and Z, which is a coordinate system (static coordinate system) of the MRI apparatus, and a gradient magnetic field that drives each gradient magnetic field coil. The gradient magnetic field Gx, Gy, and Gz are applied in the three axial directions of X, Y, and Z by driving the gradient magnetic field power source 10 of each coil according to a command from the sequencer 4. At the time of imaging, a slice selection gradient pulse (Gs) is applied in a direction orthogonal to the slice plane (imaging cross section) to set the slice plane with respect to the subject 1, and the remaining 2 are orthogonal to the slice plane and orthogonal to each other. A phase-encoded gradient pulse (Gp) and a frequency-encoded gradient pulse (Gf) are applied in one direction to encode the position information in each direction in the echo signal. Further, by combining the gradient magnetic fields in the three axial directions, an MPG pulse in an arbitrary axial direction can be applied.

The sequencer 4 is a control means 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, operates under the control of the arithmetic processing unit 8, and is a tomographic image of the subject 1. Various commands necessary for data collection are sent to the transmitting unit 5, the gradient magnetic field generating unit 3, and the receiving unit 6. In the MRI apparatus of this embodiment, a pulse sequence of diffusion-weighted imaging described later is used as the pulse sequence.

The transmission unit 5 irradiates the subject 1 with an RF pulse in order to cause nuclear magnetic resonance in the nuclear spins of the atoms constituting the biological tissue of the subject 1, and the high-frequency oscillator 11, the modulator 12, and the high-frequency amplifier It is composed of 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 the timing commanded by the sequencer 4, and the amplitude-modulated RF pulse is amplified by the high-frequency amplifier 13 and then placed close to the subject 1. By supplying the transmission coil 14a, an RF pulse is applied to the subject 1.

The receiving unit 6 detects an echo signal (NMR signal) emitted by nuclear magnetic resonance of the nuclear spins constituting the biological tissue of the subject 1, and is a high-frequency coil (reception coil) 14b and a signal amplifier 15 on the receiving side. It is composed of an orthogonal phase detector 16 and an A / D converter 17. The NMR signal of the response of the subject 1 induced by the electromagnetic wave emitted from the transmitting coil 14a is detected by the receiving coil 14b arranged close to the subject 1, amplified by the signal amplifier 15, and then amplified by the signal amplifier 15. The signals are divided into two systems orthogonal to each other by the orthogonal phase detector 16 at the timing according to the above command, each of which is converted into a digital quantity by the A / D converter 17 and sent to the signal processing unit 7.

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

The operation unit 25 functions as an input unit for inputting various control information of the MRI apparatus and control information of processing performed by the signal processing unit 7 and the arithmetic processing unit 8, and is a trackball or mouse 23, a keyboard 24, and the like. To be equipped. The operation unit 25 is arranged close to the display 20, and the operator interactively controls various processes of the MRI apparatus through the operation unit 25 while looking at the display 20.

Here, the static magnetic field generating unit 2, the gradient magnetic field generating unit 3, the transmitting unit 4, the receiving unit 6, and the sequencer 4 are collectively referred to as an imaging unit.

Currently, the nuclides to be imaged by the MRI apparatus are hydrogen nuclides (protons), which are the main constituents of the subject, which are widely used clinically. By imaging the spatial distribution of the proton density and the spatial distribution of the relaxation time of the excited state, the morphology or function of the human head, abdomen, limbs, etc. is imaged two-dimensionally or three-dimensionally.

In particular, the imaging unit of the present invention performs imaging (diffusion-weighted imaging) in which moving nuclear spins and diffusion spins are emphasized and imaged. A typical pulse sequence used for diffusion imaging is shown in FIG. In this sequence, MPG (diffusion-sensitive gradient magnetic field) pulses 301 and 302 are embedded in a sequence called spin echo type echoplanar imaging (SE-EPI). The SE-EPI consists of two high frequency pulses 201 and 202, slice gradient pulse 203 and 204, phase encoded gradient pulse 205 and 206, and readout gradient pulse 207 and 208, at which time the echo signal 209 has a readout gradient. It is generated in synchronization with the magnetic field pulse 208. The method of acquiring all the image information by executing one pulse sequence is called diffusion imaging by one-shot SE-EPI. Further, a method of acquiring one image information by repeating a series of pulse sequences a plurality of times while changing the shape of the encoded gradient magnetic field pulse 205 as shown by a dotted line is a diffusion imaging by a split type (multi-shot) SE-EPI. be called. In this embodiment, both one-shot and multi-shot can be adopted.

In FIG. 2, the MPG pulses 301 and 302 are applied in the direction of the read gradient magnetic field Gr, but may be applied to other axes (Ge, Gs) or may be applied to a plurality of axes at the same time. The MPG pulses 301 and 302 are generated using the same hardware (gradient magnetic field generator 3) as the gradient magnetic field for photographing.

Further, various pulse sequences including the pulse sequence of diffusion-weighted imaging can be stored in the storage device of the arithmetic processing unit 8 as a program in advance, and are read out and executed according to the imaging method.

In the MRI apparatus of the present embodiment, the arithmetic processing unit 8 creates a diffusion-weighted image using an echo signal (nuclear magnetic resonance signal) acquired by executing such a diffusion imaging pulse sequence. The configuration of the arithmetic processing unit 8 for this purpose is shown in FIG.

The arithmetic processing unit 8 includes an image reconstruction unit 81 that reconstructs an image using an echo signal, and a diffusion image calculation unit (1) that performs various analyzes and calculations on a plurality of diffusion-weighted images created by the image reconstruction unit 81. A calculation unit) 80, and a display control unit 86 for displaying an image created by the image reconstruction unit 81 and the diffusion image calculation unit 80, a GUI, and the like on the display 20. The diffusion image calculation unit 80 analyzes and corrects abnormal pixels of a plurality of images, and calculates a diffusion-weighted image. In the example shown in FIG. 3, in order to achieve these functions, the diffusion image calculation unit 80 includes an analysis unit 82 that analyzes the presence or absence of abnormal pixels in a plurality of images having different MPG pulse application conditions, and an analysis unit. A correction unit 83 that corrects a pixel determined to be an abnormal pixel by 82, a diffusion coefficient calculation unit 84 that calculates a diffusion weighting coefficient using an image whose pixels are corrected by the correction unit 83, and a desired diffusion weighting coefficient. A virtual diffusion image calculation unit 85 that calculates a diffusion-weighted image of the b value of the above is provided. It is also possible to incorporate a machine learning algorithm such as the diffusion image calculation unit 80 neural network as an analysis unit or a correction unit.

The functions of each of these arithmetic processing units 8 are realized by executing a program implemented in the CPU. However, some functions can be realized by hardware such as ASIC (Application Specific Integrated Circuit) and FPGA (Field-Programmable Gate Array).

Hereinafter, embodiments of the MRI apparatus and the diffusion-weighted image calculation method will be described with a focus on the functions of the arithmetic processing unit 8.

<First Embodiment>
In this embodiment, a pattern such as a pixel value or a structure of an image (an original image of a diffusion-weighted image, which is a kind of diffusion-weighted image) acquired by using a plurality of (2 or more) b values is analyzed. Then, the pixels (abnormal pixels) whose diffusion coefficient (ADC) is an abnormal value are extracted, the abnormal pixels in each diffusion-weighted image are corrected, and the diffusion coefficient (ADC) is calculated using each corrected diffusion-weighted image. A diffusion-weighted image of an arbitrary b value is calculated and displayed using the calculated diffusion coefficient (ADC).

Hereinafter, the processing of the arithmetic processing unit 8 of the present embodiment will be described in detail with reference to the processing flow shown in FIG. The processing performed by the arithmetic processing unit 8 is stored in the magnetic disk 18 as a program in advance, and is executed by the arithmetic processing unit 8 reading the program from the magnetic disk 18 and executing the program.

[Step S201]
This step is a process performed by the operator, and settings for imaging are performed. First, the operator sets up the subject in the imaging space of the static magnetic field generating unit 2. Then, the pulse sequence for diffusion-weighted imaging and the imaging conditions are set via the operation unit 25. The imaging conditions include the setting of the MPG pulse axis and the b value in addition to the general imaging parameters such as TE and TR. In this embodiment, a plurality of b values including b = 0 are set. These settings can be made via the setting screen displayed on the display 20. After the setting process is completed, imaging by the imaging unit starts.

[Step S202]
The imaging unit executes a predetermined pulse sequence under the imaging conditions set in step S201, and acquires the number of echo signals required for image reconstruction with a plurality of b values for the same cross section of the subject. .. The plurality of b values may include b value = 0. That is, the imaging is repeated at least as many as the set number of b values. When the application axes of the MPG pulses are different, imaging is performed for b values other than b value = 0 for each application axis. The image creation unit 81 reconstructs the echo signal obtained in each imaging, creates a diffusion-weighted image, and stores each image data in the magnetic disk 18. Hereinafter, for the sake of simplicity, a case where the b value is 0 and a case where imaging is performed for each of a plurality of application axes for a predetermined b value (bmsr) other than 0 will be described as an example. When distinguishing the diffusion weighted images obtained respectively by the imaging of a plurality of application axis, each diffusion weighted image is denoted as S _n. The subscript n represents the type of application shaft.

[Step S203]
The arithmetic processing unit 8 corrects the diffusion-weighted image acquired in step S202 by using a known body motion correction process, if necessary. The body movement correction includes not only the processing performed after the fact on the acquired diffusion-weighted image, but also the correction processing premised on the electrocardiographic synchronized imaging performed at the time of imaging and the imaging using the navigator echo for monitoring the body movement. Further, this step may be omitted when the target imaging region is a region that is hardly affected by body movement.

[Step S204]
The arithmetic processing unit 8 receives an arbitrary b value (b _factor ) input by the operator via the operation unit 25. The arbitrary b value is a b value used by the diffusion image calculation unit 80 to calculate the virtual diffusion-weighted image. The input of the arbitrary b value by the operator may be performed before the diffusion image calculation unit 80 calculates the diffusion-weighted image, and may be input to the imaging condition setting screen displayed in step S201. Then, when calculating the diffusion-weighted image, the display 20 may display a screen prompting the input of an arbitrary b value. FIG. 5 shows an example of the GUI 500 displayed on the display 20 for accepting the input of an arbitrary b value. In the example shown in FIG. 5, an image display unit 510 such as a diffusion-weighted image (DWI) or a diffusion-weighted coefficient map (ADC) and a menu display unit 320 for the operator to input are provided, and the b value can be input. It is configured to be performed by the numerical input box set in the analysis setting menu 521 of the menu display unit 520. Instead of inputting a numerical value, the b value may be input by operating a slider bar or the like.

[Step S205]
The analysis unit 82 of the diffusion image calculation unit 80 uses the diffusion-weighted image S0 with b value = 0 acquired in step S203 and the diffusion-weighted image S with a plurality of b values in each axis (MPG application axis) direction. Pixels with an abnormal diffusion coefficient (ADC) are determined and extracted. The method of determining whether or not an abnormal ADC occurs is that the pixel value decreases monotonically with respect to the b value (therefore, the pixel value when the b value = 0 is greater than or equal to the pixel value when the b value ≠ 0. Yes) is used. That is, the pixel value p0 of the diffusion-weighted image S0 with b value = 0 is compared with the pixel value p of the diffusion-weighted image S of each MPG application axis, and the pixel with p0 <p is determined to be an abnormal pixel. The comparison is performed between pixels having the same position on the image, and all pixels on the image are compared. However, when the ROI is set for the image, only the pixels in the ROI may be compared.

The above-mentioned example is an example of comparing pixel values, but image features may be compared not only by pixel values but also by pattern recognition or the like. For example, in the analysis using the pattern recognition algorithm, a machine learning algorithm such as Neural Network is used. As a feature of machine learning, the pixel value of the diffusion-weighted image decreases monotonically depending on the size of the b value, and the anisotropy of each tissue depends on the direction of the MPG application axis. And learn the pixel value of the abnormal example and create a model to judge the abnormality. Using a model based on this learning, it is determined whether or not there is an abnormality for each pixel of a plurality of diffusion-weighted images having different MPG application conditions, and for each tissue as well as the pixels.
Further, it is conceivable to use machine learning to correct the pixels in the next step. For example, the most suitable method when correcting using the pixels of a diffusion-weighted image and various reference values or correction values is learned for each part using a machine learning algorithm, and the optimum correction method is determined from the pixels. Create a model. Using this model, the optimum correction method is determined for the pixel determined to be abnormal by the early determination unit.

[Step S206]
The correction unit 83 corrects the pixels that are determined to be abnormal pixels by the analysis unit 82 and extracted. The correction may be to correct the pixels of the diffusion-weighted image S0 with b value = 0 or to correct the pixels of the diffusion-weighted image S with b value ≠ 0. Here, the pixel value of the diffusion-weighted image S0 is used. to correct. Specifically, the pixel value of the diffusion-weighted image S0 in the pixels extracted in step S205 is replaced with the pixel value of the diffusion-weighted image S. Maximum pixel value of the diffusion weighted image S, if there is a diffusion weighted image S of a plurality of application shaft _{(S 1, S 2 ··· S} n) , of the pixel values of corresponding pixels of each of the diffusion weighted image to replace Can be used. Further, not only one pixel but also the average value of a plurality of pixel values including the pixel values of the surrounding pixels may be used. When there are a plurality of diffusion-weighted images having a b value ≠ 0, a value calculated from pixels of another diffusion-weighted image or pixels of a plurality of diffusion-weighted images may be used.

Further, in the above example, the pixel value of b value = 0 is replaced, but the image of another b value (≠ 0) may be replaced. For example, in the abnormal pixel, the b value is compared with the average value of the peripheral pixels for each axis, and the pixel having a large degree of abnormality is corrected. When the degree of abnormality of the pixel with b value = 0 is large, the pixel with b value = 0 is corrected, and when the degree of abnormality of the pixel with b value ≠ 0 is large, the pixel with b value ≠ 0 is corrected. To do.
In the following description, a case where the pixel value of the diffusion-weighted image S0 with b value = 0 is corrected will be described as an example. The diffusion-weighted image after correcting the abnormal pixels is referred to as S0crt.

By analyzing and correcting the pixels in each axial direction in this way, the final diffusion coefficient becomes 0 in one direction, but has a value close to normal in the other direction. As a result, a more appropriate diffusion coefficient (ADC) can be calculated as compared with the method of excluding the calculation target by using the cutoff diffusion coefficient for the ADC map (image in which the pixel value is ADC) as in Patent Document 1. ..

[Step S207]
Diffusion coefficient calculating unit 84, using the corrected diffusion weighted image S0crt the diffusion weighted image S _n in step S206, for example, to calculate the diffusion coefficient ADC by the following equation.

ここで、
ＡＤＣ：計測された見かけの拡散係数（計測拡散係数）
Ｓ_ｎ：ＭＰＧ軸＝ｎの拡散強調画像
Ｓ０crt：ｂ値＝０の補正後の拡散強調画像

here,
ADC: Measured apparent diffusion coefficient (measured diffusion coefficient)
S _n : Diffusion-weighted image with MPG axis = n S0crt: Diffusion-weighted image after correction of b value = 0

Ｓ_ｂ２：ｂ２の拡散強調画像
Ｓ_{ｂ１ｃｒｔ}：ｂ値＝ｂ１の補正後の拡散強調画像 When b1 and b2 (b2> b1) whose b value is other than 0 are used and one of them is corrected by S206, the equation (1) becomes the following equation (2).

S _b2 : Diffusion-weighted image of b2 S _b1crt : Diffusion-weighted image after correction of b value = b1

The diffusion coefficient ADC is calculated for all pixels to obtain an ADC map. That is, the ADC map is an image in which the pixel value is ADC.

[Step S208]
Virtual diffusion weighted image calculating unit 85, and the diffusion coefficient ADC calculated in step S207, using any b values input in step S204 the (b _factor), diffusion weighted image of arbitrary b values _(S bfactor) Calculate using the following formula (3) or formula (3').

ここで、
Ｓbfactor: 任意のｂ値の拡散強調画像（仮想拡散強調画像：ｃＤＷＩ）
ｂ_factor：任意のｂ値

here,
Sbfactor: Diffusion-weighted image of arbitrary b value (virtual diffusion-weighted image: cDWI)
b _factor : Arbitrary b value

When an image with a b value = 0 corrected for all pixels is used, the formula (3') is a case where the original image before correction is used, and any of them may be used. If the pixel corrected for each pixel has a b value = 0, the image with the corrected b value = 0 is used, and if the corrected pixel has a b value ≠ 0, the image with the b value = 0 of the original image is used. It may be used.

In addition, although equation (2) uses "Mono Exponential Fitting" in which the change in signal strength due to the b value is fitted to a single exponential function, the signal strength is attenuated according to, for example, a binomial exponential function other than equation (2). Other mathematical models (such as Bi-Exponential Fighting) or mathematical models that do not make assumptions about the biophysical model as the distribution of diffusion coefficients (such as Diffusion Kurtosis Imaging) may be used.

Further, in step S204, a case where one b value (b _factor ) is input is shown, but it is also possible to set a plurality of b values as arbitrary b values, and in that case, a plurality of set b values are set. By performing the calculation in step S208 for the b value of, each virtual diffusion-weighted image can be calculated.

[Step S209]
The display control unit 86 displays a diffusion-weighted image of an arbitrary b value (b _factor ) calculated in step S208 on the display 20. On the screen 500 shown in FIG. 5, a virtual diffusion-weighted image cDWI is displayed on the image display unit 510 along with the diffusion-weighted image DWI and the ADC map (an image having an ADC value as a pixel value) obtained from the original image. These images are stored in a predetermined storage device.

The above is the description of the processing flow of the arithmetic processing unit 8. Needless to say, it is possible to perform known tensor analysis using the diffusion-weighted image, the diffusion-weighted coefficient, and the virtual diffusion-weighted coefficient in the processing of the arithmetic processing unit 8 to obtain further developed diagnostic information. For example, the anisotropy of diffusion is analyzed to image fibers (white matter tracts of the brain, etc.), the difference in fiber direction is imaged by the tractography technique disclosed in Non-Patent Document 1, and the diffusion anisotropy is imaged. The degree of can be quantified.

6 and 7 show a comparison of the method and effect of the virtual diffusion-weighted image calculation according to the present embodiment with the case of calculating using the cutoff diffusion coefficient (comparative example). FIG. 6 shows a procedure and an image according to the present embodiment, and FIG. 7 shows a procedure and an image according to a comparative example. In the present embodiment, the two DWI images that are the original images for the ADC calculation are corrected in advance, then the ADC is calculated and the virtual diffusion-weighted image cDWI is calculated. Therefore, the obtained cDWI does not have image deterioration due to pixel loss. On the other hand, in the comparative example, the ADC is calculated from the original image and the cutoff ADC is set to improve the accuracy of the ADC. However, since the pixels below the cutoff ADC are not used, the cDWI is used. The image quality of the cDWI obtained by calculating is deteriorated.

The MRI apparatus of the present embodiment uses an imaging unit that performs a plurality of imaging under a plurality of different MPG pulse application conditions and acquires a nuclear magnetic resonance signal for each imaging, and a nuclear magnetic resonance signal acquired by the imaging unit. The image reconstruction unit for reconstructing an image and a calculation unit for calculating a diffusion-weighted image using the plurality of images having different MPG pulse application conditions are provided, and the calculation unit includes a plurality of calculation units having different MPG pulse application conditions. It has an analysis unit that analyzes the image of the above and determines the presence or absence of abnormal pixels, and calculates a diffusion-weighted image using the plurality of images corrected based on the results of the analysis unit.

Further, the calculation unit includes a correction unit that corrects pixels determined to be abnormal by analysis, and calculates a diffusion-weighted image using the plurality of images in which the pixels have been corrected. The calculation unit applies an arbitrary MPG pulse using the diffusion weight calculation unit that calculates the diffusion weighting coefficient using the plurality of images whose pixels have been corrected by the correction unit, and the diffusion weighting coefficient calculated by the diffusion coefficient calculation unit. It further includes a diffusion-weighted image calculation unit that calculates a virtual diffusion-weighted image under the conditions.

According to the MRI apparatus of the present embodiment, it is possible to obtain a virtual diffusion-weighted image having a high b-value even when it is difficult to capture a diffusion-weighted image having a high b-value in the low-field MRI apparatus. At that time, since the cutoff is not set in the ADC, it is possible to calculate an accurate diffusion-weighted image of an arbitrary b value with few pixel defects.

<Modified example of the first embodiment>
In the first embodiment, the analysis unit 82 extracts the abnormal pixels by comparing the pixel values, but further, the abnormal pixels may be determined and extracted by using the information of the imaging conditions.

The processing flow of this modification is shown in FIG. Similar to the first embodiment, this processing flow is stored in the magnetic disk 18 as a program in advance, and is executed by the arithmetic processing unit 8 reading the program from the magnetic disk 18 and executing the program. In FIG. 8, the same processing as in FIG. 4 is indicated by the same reference numerals, and detailed description thereof will be omitted.

[Steps S201 to S204]
The operator sets the imaging conditions, the imaging unit performs imaging based on diffusion-weighted imaging, body movement correction is performed as necessary, and the b value is set for calculating the virtual diffusion-weighted image.

[Step S301]
When determining the abnormal pixel, the analysis unit 82 determines the pixel value of the pixel to be compared with reference to the imaging conditions. For example, when the reference imaging condition is the number of integrations, the median pc of each pixel of the diffusion-weighted image of each b value is calculated. The median p _c, each accumulation number, each pixel of the diffusion weighted image in the same pixel, sorted by the magnitude of the pixel value is a value which becomes the center. A median p _c in each pixel, which is calculated by comparing the pixels p ₀ of the diffusion weighted image S0 in b value = 0, p ₀ is extracted becomes smaller pixels than the median p _c. The value used for determining the pixel value may be an average value or the like instead of the median value.

[Step S302]
The correction unit 83 corrects the abnormal pixel using a value corresponding to the imaging condition. For example, in the pixels extracted in step S301, the pixel value of each pixel of the diffusion-weighted image S0 with b value = 0 is replaced with the median value of each pixel calculated in step S301. Alternatively, it is replaced with the average value of each pixel.

The imaging condition to be referred to is not limited to the number of integrations, and may be, for example, the number of MPG axes or the value of the b value. For example, when there are a plurality of MPG axes, the median value of the pixel value in the same pixel may be calculated, and the abnormal pixel may be replaced with the median value in each axis. The median value in this case is also a value that is the center value of each pixel of the diffusion-weighted image in the same pixel sorted by the size of the pixel value for each MPG axis. By making such a correction, the accuracy of the ADC can be improved when the ADC is calculated by the equation (1) in the subsequent step S207. Similarly, for the b value, a value calculated from the corresponding pixels of a plurality of b value images can be used as the correction value. For example, when an image is taken with a plurality of b values, the degree of the abnormal value is calculated by comparing with the average value of the peripheral pixels based on the value of each b value in the same pixel. A threshold value of the degree of the abnormal value is set for each magnitude of the b value, and correction is performed from the one whose degree of the abnormal value is larger than the threshold value. An image with a lower b value is considered to have a large degree of abnormality, so the threshold value is set low.

In addition, as an imaging condition, it is also possible to change the value to be replaced in consideration of the characteristics of the image in k-space, that is, whether or not the pixel is a pixel of a fine structure portion (containing many high frequency components).

According to this modification, in addition to the same effect as the effect obtained by the MRI apparatus of the first embodiment, the pixel extraction and correction are performed using the imaging conditions, so that the virtual diffusion-weighted painting monk has higher accuracy. Can be calculated.

<Second embodiment>
In this embodiment as well, the processing performed by the arithmetic processing unit 8 is basically the same as in the first embodiment, and the pixel values of each diffusion-weighted image are analyzed from the diffusion-weighted images acquired using two or more b values. By doing so, pixels having an abnormal diffusion coefficient are extracted, the extracted pixels in each diffusion-weighted image are corrected, and each corrected diffusion-weighted image is used to calculate a diffusion coefficient ADC, and the calculated ADC is used. A diffusion-weighted image of an arbitrary b value is calculated and displayed. However, the present embodiment is characterized in that the strength of correction by the correction unit 83 can be adjusted (the correction unit has an adjustment unit). The strength of the correction may be adjusted by the setting of the operator or automatically. Hereinafter, the processing of the arithmetic processing unit of the present embodiment will be described focusing on the differences from the first embodiment.

The processing flow of this embodiment is shown in FIG. Similar to the first embodiment, this processing flow is stored in the magnetic disk 18 as a program in advance, and is executed by the arithmetic processing unit 8 reading the program from the magnetic disk 18 and executing the program. Note that, in FIG. 9, the same processing as in FIG. 4 is indicated by the same reference numerals, and detailed description thereof will be omitted.

[Steps S201 to S204]
Similar to the first embodiment, the operator sets the imaging conditions, the imaging unit performs imaging based on diffusion-weighted imaging, body movement correction as necessary, and b-value setting for calculating the virtual diffusion-weighted image.

[Steps S205 to S209]
The diffusion-weighted image calculation unit 80 first determines and extracts pixels in which the ADC becomes abnormal with respect to the plurality of diffusion-weighted images acquired by imaging, and corrects the pixel values for the pixels (S205, S206). As the correction method, the same method as that described in the first embodiment or its modified example can be adopted. After that, the ADC is calculated using the corrected diffusion-weighted image (S207), and the calculated ADC and the diffusion-weighted image S0 (original image before correction or diffusion-weighted image S0cor after correction) with a b value = 0 are used. The b-value virtual diffusion-weighted image set in S204 is calculated (S208) and displayed (S209).

[Steps S901, S902]
If the correction made in step S206 is not made appropriately (S901), the strength of the correction is changed (S902). Whether or not the correction is properly performed may be performed by, for example, the operator confirming the ADC map or cDWI displayed on the display 20 in step S209, or the diffusion-weighted image based on the ADC value or the like. The calculation unit 80 (for example, the correction unit) may automatically determine.

The strength of the correction is, for example, when replacing the pixel value of the diffusion-weighted image S0 with a b value = 0 for a pixel recognized as abnormal with the pixel value of the diffusion-weighted image S with a b value ≠ 0, the pixel value of S is changed. Instead of using it as it is, it is adjusted by multiplying it by a predetermined weighting coefficient. Alternatively, the pixel value to be replaced may be changed. The weighting coefficient may be calculated based on the calculated ADC, or a predetermined coefficient may be used according to the b value. When calculating based on the ADC, for example, the weighting coefficient is determined so that the ADC value of the abnormal pixel approaches the average value of the ADC values of the peripheral pixels. The peripheral pixels may be pixels above and below and / or left and right of the target pixel, or may be a pixel to which pixels located in the diagonal direction are added. When calculating based on the b value, for example, when replacing the pixel value of the diffusion-weighted image with b value = b1 with the pixel value of the diffusion-weighted image with b value = b2, the difference between b1 and b2 ( Alternatively, the weighting coefficient may be determined so that the weight changes according to the ratio).

FIG. 10 shows an example of the GUI when the operator inputs the strength used for the correction via the operation unit 25. In this example, the image correction box 522 for inputting the strength of the image correction is displayed on the menu display unit 520 of the display screen 500 of the display 20. The operator sets the degree of correction by inputting a numerical value such as "1.1" or "0.7" or selecting a character indicating the degree of correction such as "strong" or "weak". To do.

When the arithmetic processing unit 8 automatically calculates the weighting coefficient, the image correction box 522 may be a display for selecting the necessity of correction. When the correction requirement is selected through such a GUI, or when the degree of correction is set (S902), the diffusion-weighted image calculation unit 80 returns to step S205 based on the input information, and again. The diffusion-weighted image is corrected, the diffusion coefficient is calculated, the virtual diffusion-weighted image is calculated, and the calculated image is displayed (S206 to S209). The calculated image is stored in a predetermined storage device as needed.

As described above, in the MRI apparatus of the present embodiment, in addition to the configuration of the MRI apparatus of the first embodiment, the correction unit further includes an adjustment unit for adjusting the strength of the correction. When the pixel values are simply replaced according to the method of the first embodiment, the pixel values of the calculated correction pixels of the diffusion-weighted image may be too high or too low, making it difficult to observe. According to the form, it is possible to improve the diagnostic accuracy using the obtained virtual diffusion-weighted image by optimizing the correction.

<Modification of Second Embodiment>
The feature of this modification is that, in addition to the diffusion-weighted image which is the original image for calculating the virtual diffusion-weighted image, a corrected image indicating that the pixels have been corrected is created and displayed on the display.

FIG. 11 shows a processing flow of the arithmetic processing unit 8 of this modification. In FIG. 11, the same processing as in FIGS. 4 and 9 is indicated by the same reference numerals, and detailed description thereof will be omitted.

In the present embodiment, an abnormal image is corrected for a diffusion-weighted image, virtual diffusion-weighted images are created (S201 to S209), and then a corrected image is created and displayed (S903). As the corrected image, an image showing the position of a pixel determined to be an abnormal pixel on an image of the same size as the original image (corrected position image), and a corrected value image created from the corrected pixel and the corrected value (corrected pixel). There is an image consisting of only images) or an image showing the position of the corrected pixel in an identifiable display on the diffusion-enhanced image (corrected diffusion-enhanced image) after the pixel value is corrected, but the correction position is corrected. The image is not limited to these as long as it is an image that can be seen in relation to the front and rear diffusion-enhanced images. The corrected image is displayed together with the virtual diffusion-weighted image created in step S209.

By displaying such a corrected image together with the original image and the virtual diffusion-weighted image, the operator can easily determine whether or not the correction performed in step S206 is appropriate. After that, it is determined whether or not the correction is necessary (S901), and if the correction is necessary, the strength of the correction is adjusted or the pixel value of the pixel used for the correction is changed (S902), and the correction is performed again. Do (S206). The processing of steps S901 and S902 is the same as that of the second embodiment. When it is determined that the correction is properly performed, the ADC is calculated using the corrected extended-weighted image, and the virtual diffusion-weighted image is created and displayed using the ADC (S207 to S209). When repeating the correction, the pixel position may be specified so that only the pixel for which the correction is desired to be corrected is corrected.

According to this modification, when correcting a diffusion-weighted image, it becomes easy to judge the appropriateness of the correction, unnecessary repetition of calculation can be eliminated, the accuracy of the correction is improved, and the virtual diffusion finally obtained is obtained. The accuracy of the emphasized image can be improved. In particular, by designating and enabling a specific pixel among the corrected pixels, the amount of calculation at the time of repetition can be reduced.

<Third embodiment>
This embodiment is different from the first embodiment in the method of determining abnormal pixels in a diffusion-weighted image. That is, in the MRI apparatus of the present embodiment, the imaging unit performs imaging with a b value of 3 or more, for example, a diffusion-weighted image having a b value = 0 and a diffusion-weighted image of 2 or more to obtain a diffusion-weighted image of 3 or more. The feature is that the diffusion-weighted image calculation unit (calculation unit) includes a fitting unit for fitting pixel values, and the correction unit for correcting abnormal pixels corrects pixels according to the fitting result.

Also in this embodiment, the overall configuration of the MRI apparatus is the same as that in the first embodiment, but the functions of the arithmetic processing unit are different. FIG. 12 shows a functional block diagram of the arithmetic processing unit of the present embodiment. In FIG. 12, parts having the same functions as those shown in FIG. 3 are indicated by the same reference numerals, and detailed description thereof will be omitted.

As shown in FIG. 12, the arithmetic processing unit 8A of the present embodiment includes an image reconstruction unit 81, a diffuse image arithmetic unit 80A, and a display control unit 86. The diffusion image calculation unit 80A includes a fitting unit 87 (determination unit), a diffusion coefficient calculation unit 84, and a virtual diffusion-weighted image calculation unit 85. The fitting unit 87 fits the signal values (pixel values) of a plurality of diffusion-weighted images obtained by the imaging unit performing imaging with three or more different b values by executing a pulse sequence of diffusion-weighted imaging, and makes an abnormality. While determining the pixel, the pixel value of the abnormal pixel is corrected to the pixel value on the fitting curve.

The functions of each unit included in the arithmetic processing unit 8A are realized by executing the program implemented in the CPU, as in the first and second embodiments. Further, some functions may be realized by hardware such as ASIC (Application Specific Integrated Circuit) and FPGA (Field-Programmable Gate Array).

Hereinafter, the processing of the arithmetic processing unit 8A in the present embodiment will be described with reference to the processing flow shown in FIG. In FIG. 13, the same processing as in FIG. 4 is indicated by the same reference numerals, and detailed description thereof will be omitted.

[Steps S201 to S204]
Also in this embodiment, first, the operator sets the imaging conditions (S201), and then the imaging unit performs imaging under the set conditions (S202). However, in the condition setting, in addition to the b value = 0, a b value of 2 or more is set as the b value. The imaging unit performs imaging by diffusion-weighted imaging on a predetermined MPG application axis with a plurality of set b-values, and acquires a plurality of diffusion-weighted images corresponding to the plurality of b-values. The image reconstruction unit 81 reconstructs the diffusion-weighted image and corrects the body movement as necessary (S203). Further, in step S204, an arbitrary b value for calculating the virtual diffusion-weighted image is set.

[Step S701]
The diffusion-weighted image calculation unit 80 (fitting unit 87) uses the plurality of diffusion-weighted images obtained in the above steps to fit each pixel of those images. In the fitting, an approximate curve that minimizes the distance between the pixel value and the approximate curve is obtained for the corresponding pixels of the images having different b values. At this time, some constraint conditions (for example, "if b1> b2, p1>p2","if b0 = 0, p0 ≧ p _max ", etc .: where b1 and b2 are arbitrary b values, p0, p1, and p2 are , Each of which represents a pixel value when the b value is 0, b1 and b2. P _max represents the maximum pixel value among the target pixel values for fitting).

Further, the fitting unit 87 determines that a pixel in which the distance (distance along the vertical axis) between the approximate curve thus obtained and the actual pixel value is equal to or greater than a predetermined threshold value is determined as an abnormal pixel. The threshold value is set for each magnitude of the b value, and the correction is performed from the one in which the degree of the abnormal value (here, the distance) is larger than the threshold value. An image with a lower b value is considered to have a large degree of abnormality, so the threshold value is set low.

FIG. 14 illustrates the relationship between the approximate curve and the pixel value. As the threshold value for determining an abnormality, for example, a ratio to the pixel value (distance is 10% or more of the pixel value, etc.) or a predetermined value may be set in advance, or the operator sets or changes the threshold value. You may be able to do it. In the example shown in FIG. 14, the pixel 801 of the image with b value = b1 is far from the approximate curve 800. Therefore, the pixel value (outlier value) of the pixel 801 of b1 is replaced with the pixel value at the time of b1 in the approximate curve. The fitting unit 87 performs the above-mentioned fitting and correction for all the pixels of each diffusion-weighted image, and corrects each diffusion-weighted image. In addition, instead of correcting all the pixels, if the ROI is set in advance, the pixels in the ROI may be corrected.

[Steps S207 to S209]
The ADC is calculated using the diffusion-weighted image that has been fitted and corrected by the fitting unit 87, and the diffusion-weighted image (cDWI) at the set b value is calculated using the b value set in S204 using this ADC. Doing, displaying and storing these images is the same as in the first embodiment.

Although the operation of the arithmetic processing unit 8 of the present embodiment has been mainly described, also in the present embodiment, the corrected image may be displayed on the display 20 as in the modified example of the second embodiment, or the first embodiment may be displayed. As described in the above, it is also possible to create a diagnostic image such as tractography using the DWI and cDWI created for each MPG application axis.

As described above, as a premise of the MRI apparatus of the present embodiment, a plurality of different MPG pulse application conditions, which are imaging conditions, include three or more different b-value conditions, and the analysis unit and correction of the first embodiment. As an analysis unit instead of the unit, a fitting unit is provided which fits the pixel values of a plurality of diffusion-weighted images obtained under a plurality of MPG pulse application conditions and replaces the pixel values of the abnormal pixels with values on the fitting curve for correction. ..

According to the MRI apparatus of the present embodiment, the accuracy of the ADC calculation and the virtual diffusion-weighted image calculation can be improved by correcting the diffusion-weighted image which is the original image of the virtual diffusion-weighted image calculation by fitting.

<Fourth Embodiment>
The present embodiment is characterized in that the MRI apparatus of the first to third embodiments is provided with a function of re-imaging according to the image quality of the diffusion-weighted image to be corrected.

The apparatus configuration and the functions of the arithmetic processing unit of this embodiment are almost the same as those of the first to third embodiments, but the analysis unit determines the degree of abnormality in addition to the determination of abnormal pixels, and the abnormality It is different that the imaging unit performs reimaging when the degree is large and there is a high possibility that the correction cannot deal with it. Hereinafter, the operation of the MRI apparatus of the present embodiment, particularly the processing of the arithmetic processing unit, will be described with reference to FIGS. 4 and 12 as appropriate. FIG. 15 shows the processing flow of the arithmetic processing unit of the present embodiment. In FIG. 15, the same processing as in FIG. 4 is indicated by the same reference numerals, and detailed description thereof will be omitted.

[Steps S201 to S203]
Similar to the first embodiment, the process from imaging to acquisition of diffusion-weighted image is performed.

[Step S910]
The analysis unit 82 first analyzes the diffusion-weighted image to determine abnormal pixels. This analysis method may be any of the methods adopted in the first to third embodiments. After determining the abnormal pixel, the degree of abnormality and the occurrence rate of abnormality are further determined. The degree of abnormality is determined by, for example, whether or not the pixel p0 of the diffusion-weighted image with b value = 0 and the pixel p1 of the diffusion-weighted image with b value = b1 are significantly separated from each other. To judge. For example, when a predetermined threshold value is set for the ratio to the pixel value [(p1-p0) / p1] and the difference between the pixel values [p1-p0] and these values exceed the predetermined threshold value (for example, the pixel of an abnormal pixel). When the value exceeds twice the pixel value of the diffusion-weighted image with b value = 0), the degree of abnormality is determined to be “large”. Regarding the occurrence rate of abnormality, for example, when the number of pixels determined to be abnormal (number of abnormal pixels / total number of pixels) exceeds a predetermined ratio (for example, 50%) with respect to the total number of pixels, the degree of abnormality is ". Judged as "Large". By combining both, a threshold value for determining pixel abnormality (first threshold value) and a threshold value for determining the ratio (second threshold value) are set, respectively, and the ratio of pixels having a pixel value equal to or higher than the first threshold value becomes equal to or higher than the second threshold value. When it becomes, the degree of abnormality may be determined to be "large". Alternatively, a known pattern recognition technique may be used to extract a diffusion-weighted image having a particularly large degree of abnormality from among the plurality of diffusion-weighted images.

[Steps S911 to S913]
If the degree of abnormality is determined to be "large" (S911), if it is the first determination (step S912), the process proceeds to step S913, and diffusion weighting is determined to be "large". Reimaging only the image is performed. For example, when an image having a predetermined b value is determined to be "abnormally large", only the image of the b value is taken. Further, when a plurality of slices are imaged and the image of a predetermined slice is determined to be "abnormally large", only that slice is re-imaged.

With respect to the diffusion-weighted image obtained after the re-imaging, another diffusion-weighted image is used to determine the abnormal pixels and the degree of abnormality (step S910). If any of the images is determined to be "abnormally large" again in this second determination (S911, S912), the process returns to step S202, and all imaging is repeated. Alternatively, a known pattern recognition technique may be used to extract a diffusion-weighted image having a particularly large degree of abnormality from among the plurality of diffusion-weighted images, and the extracted diffusion-weighted image may be re-imaged.

[Steps S206 to S209]
If the degree of abnormality is determined to be "small" in step S911, the process proceeds to the steps after step S206, correction of abnormal pixels (S206), calculation of ADC (S207), calculation and display of virtual diffusion-weighted image (S207). S208, S209). The contents of these processes are as described in the first embodiment.
In the embodiment shown in FIG. 15, the reimaging process is different depending on whether the determination is the first time or later, but the reimaging may be performed by returning to step 202 from the beginning, and the determination result shows "abnormality". When the result of "" reaches a predetermined number of times, the process may return to step 202.

The example of applying this embodiment to the first embodiment has been described above. However, the present embodiment is characterized in that the degree of abnormality is determined and automatic reimaging is performed using the determination result, and the first embodiment is characterized. It can be similarly applied not only to the embodiment but also to the modified examples thereof and the second and third embodiments.

Further, the determination criteria and thresholds exemplified in the above description are merely examples, and various changes can be made. The target for reimaging can also be changed as appropriate. It is also possible to incorporate the operator's selection (selection including the possibility of reimaging) at the time of reimaging.

In the MRI apparatus of the present embodiment, the analysis unit determines the presence or absence of abnormal pixels and the degree of abnormality. Then, depending on the degree of abnormality, the imaging unit is made to perform reimaging.
According to the present embodiment, when there are many abnormal values, reimaging is performed and a virtual diffusion-weighted image is created using the image after reimaging. This is expected to improve inspection efficiency and shorten inspection time.

Although each embodiment of the present invention has been described above, the present invention is not limited to these, and it is also included in the present invention that elements and steps that are not essential in the present invention are appropriately deleted or added. .. Further, in each of the above embodiments, the diffusion-weighted image calculation method executed by the MRI apparatus and its arithmetic processing unit has been described, but the diffusion-weighted image calculation method of the present invention is realized by an image processing apparatus independent of the MRI apparatus. Is also possible, and such an image processing apparatus is also included in the present invention.

1: Subject 2: Static magnetic field generator 3: Diagonal magnetic field generator 4: Sequencer, 5: Transmitter, 6: Receiver, 7: Signal processing unit, 8: Arithmetic processing unit (CPU), 8A: Arithmetic processing unit (CPU), 9: gradient magnetic field coil, 10: gradient magnetic field power supply, 11: high frequency transmitter, 12: modulator, 13: high frequency amplifier, 14a: high frequency coil (transmission coil), 14b: high frequency coil (reception) Coil), 15: Signal amplifier, 16: Orthogonal phase detector, 17: A / D converter, 18: Magnetic disk, 19: Optical disk, 20: Display, 21: ROM, 22: RAM, 23: Trackball or mouse , 24: Keyboard, 80: Diffusion-enhanced image calculation unit (calculation unit), 80A: Diffusion-enhanced image calculation unit (calculation unit), 81: Image reconstruction unit, 82: Analysis unit, 83: Correction unit, 84: Diffusion coefficient Calculation unit, 85: Virtual diffusion enhancement image calculation unit, 86: Display control unit, 87: Fitting unit

Claims (19)

An imaging unit that performs a plurality of imaging under a plurality of different MPG pulse application conditions and acquires a nuclear magnetic resonance signal for each imaging, and an image reconstruction that reconstructs an image using the nuclear magnetic resonance signal acquired by the imaging unit. A unit and a calculation unit that calculates a diffusion-weighted image using a plurality of the images having different MPG pulse application conditions are provided.
The calculation unit has an analysis unit that analyzes a plurality of images having different MPG pulse application conditions and determines the presence or absence of abnormal pixels based on the relationship between the pixel values of the plurality of images, and is based on the results of the analysis unit. A magnetic resonance imaging apparatus characterized in that a diffusion-weighted image is calculated using the plurality of images including the corrected image . The magnetic resonance imaging apparatus according to claim 1.
The calculation unit includes a correction unit that corrects pixels determined to be abnormal by analysis, and a diffusion coefficient calculation unit that calculates a diffusion weighting coefficient using the plurality of images including an image whose pixels have been corrected by the correction unit. A magnetic resonance imaging apparatus comprising the above, wherein a diffusion-weighted image is calculated using a diffusion-weighted image calculated by the diffusion-weighted calculation unit. The magnetic resonance imaging apparatus according to claim 2.
The calculation unit uses the plurality of images whose pixels have been corrected by the correction unit or the plurality of images before correction, and the diffusion weighting coefficient calculated by the diffusion coefficient calculation unit under arbitrary MPG pulse application conditions. A magnetic resonance imaging apparatus further comprising a diffusion-weighted image calculation unit that calculates a virtual diffusion-weighted image. The magnetic resonance imaging apparatus according to claim 1.
The analysis unit is a magnetic resonance imaging apparatus characterized in that the degree of abnormality is determined by using a predetermined threshold value according to the magnitude of the b value. The magnetic resonance imaging apparatus according to claim 1.
The analysis unit is a magnetic resonance imaging apparatus characterized in that the conditions for determining abnormal pixels differ depending on any of the imaging conditions, pixel values, and tissues of the image to be analyzed. The magnetic resonance imaging apparatus according to claim 1.
The arithmetic unit, the analysis of the image by the analysis unit, or the analysis of the presence or absence of abnormal pixels, or magnetic resonance imaging apparatus and performing based on machine learning algorithms to correct the pixel. The magnetic resonance imaging apparatus according to claim 1.
The plurality of different MPG pulse application conditions include a plurality of different b values.
The analysis unit is a magnetic resonance imaging apparatus characterized in that abnormal pixels are determined by comparing the pixel values of a plurality of different b-value diffusion-weighted images. The magnetic resonance imaging apparatus according to claim 7.
The plurality of different MPG pulse application conditions include a plurality of different MPG application axes or a plurality of different b values.
The calculation unit compares the pixel values by the analysis unit, determines abnormal pixels based on the pixels in the diffusion-weighted image under the MPG application conditions and a machine learning algorithm, or corrects the diffusion-weighted image. A magnetic resonance imaging apparatus characterized in that. The magnetic resonance imaging apparatus according to claim 1.
The plurality of different MPG pulse application conditions include a plurality of different MPG application axes.
The analysis unit is a magnetic resonance imaging apparatus characterized in that a reference value is calculated from the same pixel of each MPG application axis, and an abnormal pixel is determined by comparing the pixel of each MPG axis with the reference value. The magnetic resonance imaging apparatus according to claim 2.
The correction unit corrects the pixel value of the diffusion-weighted image when the b value = b1 by using the pixel value of the diffusion-weighted image when the b value = b2 (however, b2 ≠ b1). Resonance imaging device. The magnetic resonance imaging apparatus according to claim 10.
The plurality of different MPG pulse application conditions include a plurality of conditions in which the application axes of MPG pulses are different.
The analysis unit is characterized in that the pixel value of the diffusion-weighted image when the b value = b2 is calculated from a plurality of diffusion-weighted images obtained under a plurality of conditions in which the MPG pulse application axes are different. Magnetic resonance imaging device. The magnetic resonance imaging apparatus according to claim 2.
The correction unit is a magnetic resonance imaging apparatus further comprising an adjustment unit for adjusting the strength of correction. The magnetic resonance imaging apparatus according to claim 12.
The adjusting unit is a magnetic resonance imaging apparatus characterized in that the pixel value used for correcting abnormal pixels is multiplied by a predetermined weighting coefficient to adjust the strength of the correction. The magnetic resonance imaging apparatus according to claim 2.
It also has a display device that displays diffusion-weighted images.
The correction unit is a magnetic resonance imaging device characterized in that a corrected image representing the corrected abnormal pixel is created and displayed on the display device. The magnetic resonance imaging apparatus according to claim 2.
The plurality of different MPG pulse application conditions include 3 or more different b values.
The correction unit fits the pixel values of a plurality of diffusion-weighted images obtained under a plurality of MPG pulse application conditions, and replaces the pixel values of the abnormal pixels with values on the fitting curve for correction. Resonance imaging device. The magnetic resonance imaging apparatus according to claim 1.
The analysis unit is a magnetic resonance imaging device that determines the presence or absence of abnormal pixels, determines the degree of abnormality, and causes the imaging unit to perform reimaging according to the degree of abnormality. The magnetic resonance imaging apparatus according to claim 3 .
The arithmetic unit, the diffusion weighted image and using said diffusion weighted image of a virtual, magnetic resonance imaging apparatus characterized by having a diffusion tensor image creating section that creates a diffusion tensor imaging. An image processing device having a calculation unit that calculates a virtual diffusion-weighted image under an arbitrary MPG pulse application condition by using a diffusion-weighted image acquired under a plurality of MPG pulse application conditions using an MRI apparatus.
The calculation unit is a correction unit that determines abnormal pixels based on the relationship between the pixel values of the plurality of diffusion-weighted images for a plurality of diffusion-weighted images having different MPG pulse application conditions, and corrects the determined abnormal pixels. A diffusion weight calculation unit that calculates a diffusion weight coefficient using the plurality of diffusion weight images including a diffusion weighted image whose pixels have been corrected by the correction unit, and a diffusion weighting coefficient calculated by the diffusion coefficient calculation unit. An image processing apparatus characterized in that the virtual diffusion-weighted image is calculated using the above. Diffusion-weighted image calculation to calculate a virtual diffusion-weighted image under desired MPG pulse application conditions using the nuclear magnetic resonance signals obtained from imaging without applying MPG pulses and imaging with one or more MPG pulses applied. It ’s a method,
The steps of reconstructing the nuclear magnetic resonance signal obtained by each imaging to create multiple diffusion-weighted images,
A step of determining abnormal pixels based on the relationship between the pixel values of the plurality of diffusion-weighted images for a plurality of diffusion-weighted images having different MPG pulse application conditions.
The step of correcting the pixels determined to be abnormal pixels and creating the corrected diffusion-weighted image,
A step of calculating the diffusion-weighted coefficient using the plurality of diffusion-weighted images including the corrected diffusion-weighted image , and
A diffusion-weighted image calculation method including a step of calculating a virtual diffusion-weighted image using the calculated diffusion-weighted coefficient.

Images