JPS62253043A - Tomographic image diagnostic apparatus by nuclear magnetic resonance - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention relates to the nuclear magnetic resonance phenomenon (Nuclear Magnetic Resonance phenomenon).
etic Resonance), the specific atomic nuclear density distribution and spin/lattice relaxation time of each tissue inside the human body are measured non-invasively as nuclear magnetic resonance (NMR) signal data from outside the subject, and the The present invention relates to a tomographic imaging apparatus (hereinafter abbreviated as NMR-CT) that displays information as an image and provides information for medical diagnosis.

[Conventional technology]

Conventionally, diagnostic X-ray devices (including CT scanners), nuclear medicine devices, ultrasound devices, etc. have been widely used as devices for externally diagnosing tomographic planes of organs and tissues inside the human body. Each of these has its own advantages and disadvantages. Although a description of these devices will be omitted, matters necessary for comparison with the NMR-CT device targeted by the present invention will be briefly described below.

That is, the diagnostic XL device performs morphological discrimination of tissues based on differences in X-ray transmission characteristics in human tissues, and diagnoses each disease based on this. Ultrasonic devices also provide diagnosis by determining the morphology of tissues or measuring their movements based on differences in acoustic impedance in human tissues. In addition, nuclear medicine devices can measure the biochemical characteristics of each tissue by measuring the distribution within human tissues of radioactive isotopes or their compounds administered to the human body, but they mainly measure the macroscopic characteristics of each tissue. The purpose is to obtain relevant information. On the other hand, the NMR-CT apparatus that is the subject of the present invention has the following characteristics.

(1) NMR-CT only uses magnetic fields (static magnetic field and high-frequency magnetic field) as a medium to extract information from the human body, and does not use ionizing radiation like X-ray CT.
Therefore, there is no risk of radiation exposure.

(2) In NMR-CT, it is possible to image and display information about specific atomic nuclei, particularly hydrogen nuclei (1H), which is a basic constituent of human tissue.

(3) In NMR-CT, biochemical information regarding biochemical changes in living tissues can be obtained through spin and lattice relaxation times, which are the basic signals measured.

Therefore, it becomes possible to distinguish malignant tumors such as cancer from normal tissue, and it has the potential to perform image diagnosis at the cellular level.

(4) Since NMR-CT provides different signals in response to dynamic changes in substances such as flow and diffusion, it is an effective means for observing blood flow.

(5) In X-ray CT, images of bones appear strongly, which often hinders diagnosis, but in NMR-CT, bones and metals other than ferromagnetic materials hardly affect the images.

(6) In NMR-CT, tomographic images (tilt images) from multiple directions can be freely measured without moving the subject (patient).

As described above, NMR-CT has many features that cannot be obtained with other medical image diagnostic equipment, and research and development is currently progressing in various countries as a next-generation medical imaging diagnostic equipment following X-ray CT.

Next, the operating principle of NMR-CT will be briefly explained below to the extent necessary for understanding the present invention.
Educated Local Interacti
ons: Examples Emp
loyingNuclear Magnetic
Re5onance"by P, C, Lauterbu
r “N a ture vo1.242 Marc
h6. 1973° pp. 190-191).

In general, a hydrogen atom nucleus, which is a basic constituent of the human body, has a nuclear spin (2), and a nuclear magnetic moment μ (μ) associated with this spin. A μ-γhI relationship exists between these. However, γ is the nuclear gyromagnetic ratio, and h is the blank constant/2π
It is.

Here, when a specimen such as a human body is placed in a static magnetic field H0, the nuclear magnetic moment of the hydrogen nucleus changes around the static magnetic field H.
. It precesses at an angular frequency (Larmor frequency) of −rHo. Here, the angular frequency ω is further added to the plane perpendicular to the static magnetic field H0. When a rotating high frequency magnetic field (RF magnetic field) H+ is applied, a nuclear magnetic resonance phenomenon occurs. At this time, it can be seen that the magnetization vector M, which is the sum of the nuclear magnetic moments, tilts in a direction different from the direction of the static magnetic field H0 when viewed from the rotating coordinate system.

In the pulsed NMR method, when the angle θ of the magnetization vector M becomes 90@ or 180@, the application of the high-frequency magnetic field is cut off (the resulting high-frequency magnetic field pulse is
A high frequency magnetic field pulse of 0@ or 180° is called an RF pulse for short. ), when the nuclear spin system returns to its original thermal equilibrium state (during relaxation), an exponential decay signal (FREE INDUCTION D) is induced in the nearby coil.
ECAY (abbreviated as FID signal or echo signal)
Measure.

By repeatedly applying such high frequency magnetic field pulses (RF pulses) with θ = 900 or 1806, information (NMR signals) such as the concentration of hydrogen nuclei and relaxation time can be obtained, and information obtained by combining pulses ((pulse sequence)) can be obtained. The time constants of relaxation include the spin-lattice relaxation time T and the spin-spin relaxation time T2, and the hydrogen nuclear density ρ, which indicates the signal intensity, and N
It constitutes a basic parameter of the signal observed in MR-CT.

Many methods have been developed for pulse sequences, and the representative one is the SR method (Saturday).
ation Recovery method), IR method (In
version Recovery method), SE method (Spi
n Echo method), CPMG method (Carr-Purc method)
ell and Meibooa+-Gill method)
and so on. Since this specification is not intended to describe these pulse sequences in detail, by way of example, an overview of the SR method will be described with reference to FIGS. 4 and 4A.

FIG. 4 is an explanatory diagram showing a pulse sequence by the SR method,
FIG. 4A is a perspective view showing coordinate axes in the subject.

In period 1, in order to select a slice plane in the Z-axis direction (body axis direction) of the subject, a 90° RF pulse and a Z-direction gradient control pulse are applied by a selective excitation method. Period 2 is
X
The zero period 4 during which the FID signal is collected as data while applying the gradient control pulse in the Y direction is a period until the excited nuclear spins return to the original equilibrium state. Note that in FIG. 4, T indicates a -projection period.

Since the data obtained by the pulse sequence in this way contains coordinate position information of the part to be measured due to the action of the gradient magnetic field, the filtered back projection method, which has been conventionally used in X-ray CT, is used. By applying such techniques, NMR signals can be displayed as image information. Many algorithms have been developed for imaging techniques other than the above-mentioned filtered back projection method, but since they have no direct relation to the present invention, detailed explanations will be omitted.

FIG. 5 is a sequence explanatory diagram of a single slice using the SR method, and shows a case where a filtered back projection method is used as a reconstruction method.

In this case, as shown in Figure 5A, one slice plane S is selected in the direction perpendicular to the Z-axis, and the data of that plane is changed by increasing the gradient in the X-Y direction by 1 degree from 0 degree. Collect data and perform up to 180°.

In addition, in Fig. 5, the data collection for 1° is
It is expressed as ga.

In addition, in single slice, data is collected only once in selective excitation (application of RF90'' pulse) and once in the FED signal between views, but in multi-slice, data is collected in one view as shown in Figure 6. The first
The frequency of selective excitation of (r+) is different from the frequency of selective excitation (r+).
fz), and data is collected from the FID signal of a slice plane different from the first slice plane. In this way, typically n slices of data can be collected at n frequencies of f, to f,% during IVieII4. Therefore, by repeating this process up to 180Vie%1, data on n slice planes can be collected in approximately the same time as a single slice.

Figure 7 is an explanatory diagram for explaining the IR method, in which slice selection is not performed in one excitation, gradient control pulses are applied from three directions (x, y, and z), three-dimensional data is collected, and calculations are performed. The three-dimensional data is reconstructed using a three-dimensional Fourier transform.

By the way, the image reconstructed by these pulse sequences has the above-mentioned parameters T'+ and Tt.

It is formed in such a way that all ρ are included. However, due to differences in the characteristics of each pulse sequence, images obtained by the SR method are hydrogen nucleus density ρ-weighted images, images obtained by the IR method are T1-weighted images, and images obtained by the SR method and CPMG method are T1-weighted images.
It is characterized as a 2-enhanced image. This situation is shown in the following table, in which the parameters to be emphasized are shown surrounded by a square frame.

In this way, in NMR-CT, the image obtained by the pulse sequence is generated in a form that includes all 'r' + + Tt * ρ, so the true T, image, true T, image, true In order to obtain 9 images, it is necessary to calculate the intensity of the NMR signal from a plurality of images using the following equation.

For example, in order to obtain a T8 image, calculation is performed using the equation of the signal intensity I for each pixel using the IR method as shown in equation (1) below.

τ, C is a constant, τ is the interval between the 180' pulse and the 90'' pulse (inversion time T,'), and T is the repetition time (repetition time).In other words, the above (1
A T3 image can be created by performing calculations for each pixel using Equation 1, which is T2. The same applies to ρ.

Now, the general system configuration of NMR-CT will be explained with reference to FIG. Incidentally, FIG. 8 is a schematic diagram showing the system configuration.

In the figure, 21 is a static magnetic field coil, 22 is a power source for the static magnetic field coil, 23 is a gradient magnetic field coil, 24 is a power source for the gradient magnetic field coil, 25 is an RF coil, 26 is an RF spectrometer, 27 is a patient bed control, 28 is a cooling system, and these constitute the RF section. In addition, X is a ★ body, and 20 is a computer system. Therefore, the specimen
An RF pulse is applied by the coil 25, and as a result, the FID signal or echo signal output from the specimen
The data is detected by the F spectrometer 26, converted into digital data, and then imported into the computer system 20.

The computer system 20 includes, for example, a main processor (host CPU) 4 and an RF Zainface 5 as shown in FIG.
, a host CPU memory 6, an array processor 7, a magnetic tape device (MT) 8, a magnetic disk device 9, an image display 10, a system bus 11, and the like. Note that 2e, 2f, and 2g are interfaces. That is, the FID signal or echo signal given from the RF section is sent to the host CP via the RF sensor face 5.
The image is taken into U4, where the image is reconstructed in cooperation with the array processor 7, and the result is stored in the magnetic disk device 9. The image data taken into the magnetic disk device 9 is subjected to predetermined processing in the array processor 7, and then displayed as an image on the image display 10 via the interface 2e.

[Problem that the invention seeks to solve]

Now, as mentioned above, in conventional X-ray equipment (X-ray imaging equipment, X-ray CT) and ultrasound diagnostic equipment, there is usually only one type of target image parameter (for X-ray absorption coefficient, acoustic impedance in the case of ultrasound, etc.). Therefore, in actual diagnosis, although a number of images of a patient may be taken at various locations, there is basically only one type of image for each location. For this reason, with conventional X-ray and ultrasound equipment, when it comes to image handling and storage issues, the emphasis is on how to efficiently manage and easily access images of a given patient. It has been placed.

However, in the case of NMR images, as mentioned above, the typical ones are SR images, IR images, and SE images.

One part, such as T1 image, T2 image, density (ρ) image, etc.
It is possible to compare one slice plane using different imaging methods, or even with one imaging method, to change parameters such as τ and T1 in various ways (in the case of equation 11) and compare them. In many cases, it can be said that there are almost no cases in which a diagnosis can be made with only one image.In other words, in NMR image diagnosis, diagnosis is often made by comparing different images of the same region.However, conventional NMR- CT (MRi) apparatuses often have system configurations based on X-ray CT, and therefore have a problem in that they are not conveniently configured to compare multiple images.

In fact, NMR-CT (MRi
) When using NMR-CT (MRf), there is no need to take such circumstances into consideration; rather, the problem tends to be the quality of the image. However, as NMR-CT (MRf) becomes more widespread, screening When used for especially,
NMR-CT (MRi) takes a long time to scan
Compared to line-CT, throughput is often a problem. In this sense, an efficient interpretation system for NMR images is strongly desired.

One of the methods that have been tried so far to solve this problem is
The first is classification of image density levels using a pseudocolor method. Although this colorization makes it easier to interpret NMR images and saves time, subtle information such as at density level boundaries is often lost due to the classification of density information, and There is no international standard for what colors are considered high quality, so although it is being used on a trial basis, it has not yet reached a practical level. In addition, PAC3 (Picture
Archining and Com+
5unication system11), a system for storing and managing medical images by networking within a hospital using optical fibers, etc., has been extensively researched.
This makes the access and storage of images from various modalities more efficient, but this mainly focuses on the efficient storage and operation of images between each modality within a hospital. ) does not directly solve the specific problem.

As described above, in NMR-CT (MRi), it is necessary to interpret images in a short time and increase diagnostic throughput.

Therefore, in NMR-CT (or MRi), the present invention reduces the burden on the diagnostician by providing a system that can efficiently identify groups of images drawn using each pulse sequence and each parameter, and ultimately The purpose is to enable automatic diagnosis of NMR images using a computer.

[Means for solving problems]

A plurality of auxiliary processors each take out a predetermined set of a plurality of sets of tomographic image data stored in a disk memory based on instructions from the main processor, either as is or by performing a predetermined operation, and a plurality of auxiliary processors each retrieve the data from the auxiliary processors. A plurality of image plane memories are provided to store each image plane memory, and a plurality of desired image data can be retrieved simultaneously by giving a predetermined instruction from the main processor to the auxiliary processor.

[Effect]

The present invention independently stores tomographic images according to each pulse sequence and each parameter and auxiliary pattern images for image analysis in NMR-CT (MRi),
By providing multiple image plane memories (hereinafter simply referred to as image brains) within a host computer or image display that can be accessed independently and at high speed, multiple heterogeneous images essentially required in NMR image diagnosis can be realized. The purpose of the present invention is to provide a system that facilitates simultaneous diagnosis of NMR images and enables automatic diagnosis of NMR images by a computer by computer processing a plurality of different types of images. Here, the "auxiliary pattern image for image analysis" refers to, for example, a frame (area designation) for displaying a ROI (region of interest) or a profile (Profi
le) A general term for linear markers for display, window levels for specifying the gray scale of images, and various other auxiliary patterns for tomographic image analysis. An example of the auxiliary pattern is shown in FIG. 3. FIG. 3(a) shows an example of a region of interest (ROI), and FIG. 3(b) shows an example of a marker.

In other words, by providing multiple image planes like this, it is no longer necessary to access the image file on the disk each time the image parameters or pulse sequences differ, and it is possible to load related image groups into each image plane. By doing so, it becomes possible to compare images using high-speed simultaneous access. In addition, in NMR image diagnosis, the same region is often imaged with different pulse sequences and different sequence parameters, and the images are compared with each other, but in the case of the same region, it is possible to align the images. After alignment, various image processing can be performed for mutual comparison of images.

Various types of image processing include simultaneous extraction of different types of images using a common auxiliary pattern image. For example, it is possible to extract an IR image and an SE image in the same ROI area and display them simultaneously, or to perform mutual comparison by simultaneously displaying T, images, and T2 images at the same window level.

Furthermore, by overlapping the aligned images with appropriate weighting, it is possible to create a composite image that shows features effective for a certain lesion.

As mentioned above, in the uninvention, NMR images have various parameters unlike X-ray images and ultrasound images, so image diagnosis is performed by comparing multiple different types of images. , different types of images can be stored independently in a host computer or an image display, and multiple image planes that can be independently accessed at high speed are provided, allowing simultaneous access of different types of images, extraction of the same area,
It also facilitates NMR image diagnosis by performing superimposition, etc., and can serve as a basic tool for automatic diagnosis using computers.
It indicates the direction, which can also be called diagnosis.

〔Example〕

FIG. 1 is a block diagram showing an embodiment of the present invention.

In the figure, 1 (13 to ld) is an image plane memory, 2 (2a to 2g) is an interface, and 3 (
3a to 3d) are auxiliary processors such as microprocessors, and the rest are the same as in FIG. 9.

Note that the display 10 has two systems (10a, 1
0b) Provided.

Now, NMR signals (FID signals or echo signals) collected by exciting hydrogen nuclei in a designated cross section of a specimen such as a human body are sent to the host CPU 4 via the RF interface 5. The NMR signal is image-reconstructed by the host CPU 4, array processor 7, etc.
The NMR image is created as a digitized grayscale image of bit or 512x512x8 bit. In the conventional method, these images are stored on disk 9 along with information such as patient name, age, gender, date, and pulse sequence parameters.
It was stored in an image file within. Therefore, when interpreting images, it is necessary to make hard copies of several images of the same region due to differences in pulse sequences and parameters, etc. using a multi-format camera, etc., and then compare the images with each other using a scanner such as Scherkasten. I was making a diagnosis while doing so. In contrast, in the present invention, the first
As shown in the figure, there are multiple image planes 1 (1a) within a computer system or within an image display.
~) d) By preparing image planes 1 and making it possible to access each image plane 1 independently and simultaneously, it is possible to improve the efficiency of image comparison, and also to combine and extract images under the same conditions, which is generally complicated. This facilitates NMR image diagnosis, which is a common problem.

That is, as shown in the figure, in this method, a plurality of memory planes 1 for images are prepared. Each of the image planes 1a to 1d is not dedicated to a certain pulse sequence, but is designed so that various images can be loaded as required.

The number of image planes 1 needs to be at least three, considering the flexibility of various processes described below.

In this method, the larger the number of memory planes, the easier the processing, but the number is decided based on cost considerations (in Figure 1, it is assumed that there are four memory planes). A microprocessor 3 dedicated to image handling of the image plane is connected via an interface 2 (2a to 2d) for image handling.

Microprocessor 3 is the computer's system bus 1
1, and is configured to exchange commands with the host CPU 4 and control the loading of image data from the disk 9 onto the image plane 1.

First, the images reconstructed by the host CPU 4 and the array processor 7 include the date, patient name, and age of 1 year.

The information is temporarily stored in an image file on the disk 9 together with information such as gender and pulse sequence parameters. This is called image creation mode.Next, in the stage of interpreting the various created images (image interpretation mode), the host CPU 4 does not directly handle the various images, but only gives commands to the microprocessor 3. Each microprocessor 3
handles the image in the image brain 1. By doing so, it becomes possible to simultaneously handle a large number of different types of images, and high-speed access can be realized. The microprocessor 3 receives image processing commands from the host CPU 4, loads image data onto the image plane 1 from the disk 9, and synthesizes images.

Performs the extraction and waits for the next command when it is finished. Furthermore, in this embodiment, two image display monitors (10a, 10b) are arranged side by side in order to make it easier to compare images of different types, but this is not an essential condition.

Now, with the above configuration, the following image processing can be performed in this method.

■ Simultaneous display of different images of the same region For example, for IR images, the repetition time T is 1400 m5° and the inversion time Ta'400 (71 images (hereinafter referred to as IR (1400/4
00)) and the repetition time T in the SE image.

= 1160ms, Eco 1 hour Tt -160ms image (hereinafter referred to as SE (1160/160)), etc., and interpreting while comparing the differences in shading is as follows.
This is often done in NMR image diagnosis.

In such a case, in this method, the operator first enters the date and
If you specify IR (1400/400) and 5E (1160/160) along with patient name etc., host C
For example, the PU 4 transmits an image data load command to the microprocessor 3a using parameters such as date, patient name, and JR (1400/400), and transmits the date, patient name, and 5E (1160/16) to the microprocessor 3b.
A command to load image data using parameters 0) is transmitted. The microprocessors 3a and 3b load images with specified parameters from the disk 9, and store the images in image planes la and lb, respectively.
The respective images are displayed on the image displays 10a and 10b. FIG. 2A shows an example of the SE image and the IR image. When the parameter Tt is long (A), the area A7□ becomes white, and when the parameter T is long (B), the area Ay
It can be seen that , becomes black.

In addition, if the arrow "↓" in the table mentioned above is black, it will be "↑
” indicates the case where the color becomes white.

■ Superposition of images The IR (1400/400) image and S
The E (1160/160) image has a completely different meaning. Therefore, by superimposing these images (with weighting if necessary), it is possible to create an image with new meaning. In this case, by simply issuing an image superimposition command from the host CPU 4, IR (1400/400) and SE (1160/16
0) images can be superimposed. At this time, if it is necessary to leave the original IR image and SE image as they are on the image plane, the resulting images are transferred to the image plane IC via the microprocessor 3c, for example.

■ Image subtraction Similar to section (■), it is also possible to easily create images with new meaning by performing image subtraction processing. In particular, the effectiveness of image subtraction before and after contrast agent injection is important, such as angiographic subtraction (Digit
al 5ubtraction Angiogra
Gd-DTPA (Gadolinium DTPA) has been demonstrated in NMR images.
As the effectiveness of contrast agents such as these is being studied, multiple image branes using this method are ideal for such subtraction processing.

■ Extracting and enlarging the same ROI (region of interest) When comparing different images of the same region, find out how the ROI (region of interest) of one image compares to the ROI of the other image located at the same position. There are many things. In such a case, if the example in section (2) is used as is, a frame (area) of the region of interest ROI is set in the desired part of the IR (1400/400) image, and the ROI of the IR image is extracted. Next, the coordinates of the frame are extracted by the microprocessor 3a, and the coordinates of the frame are extracted by the microprocessor 3a.
Transfer to b. Then, by superimposing the ROI frame and the image plane 1b using the microprocessor 3b, it is possible to extract an area at the same position as that extracted from the IR image. This situation is shown in Fig. 2B, in which A and I indicate the same region of interest, and the extracted images can be stored in image planes lc and ld, for example, and processed such as enlargement. can.

■ Profile display on the same marker This process also ■
This can be done using the same procedure as in Section 1. FIG. 2C shows an example of a profile marker for a heterogeneous image;
are the same profile markers.

■ Display at the same window level Window level (density range frame) from main CPU4
IR (1400/400) and SE (1160/1) are transmitted to each microprocessor 3a, 3b.
60) can be displayed at the same window level.

■ Pseudo-colorization of the same density level This process is also performed in the same way as in item 0, by instructing the density level from the main CPU 4 to each microprocessor 3a, 3b, and also instructing the corresponding display color. Portions with the same density level can be displayed in the same color on the displays 10a and 10b, respectively.

(2) Display of the same density contour line This process is also performed by instructing the density level from the main CPU 4 to each microprocessor, as in the case of the 0th item. This situation is shown in FIG. 2D, where L9 is the concentration contour line.

(2) Comparison of distances between two identical points This is necessary, for example, when specifying two points in an IR image and wanting to compare the distance between them in an SE image. First, two fixed points are specified in the IR image by an operator using a light pen or the like, and the distances are measured. Next, the coordinates of the two fixed points are extracted by, for example, the microprocessor 3a and transferred to the microprocessor 3b, and the distance between the two fixed points in the SE image is measured by the microprocessor 3b. In this way, the same two
Distances between fixed points can be compared. The same 2 images of different types shown in Fig. 2E.
An example of distance between points is shown. L is the distance between two identical points.

[Phase] Comparison of angles between the same three points This can also be performed using the same procedure as the 0th term.

The situation is shown in Fig. 2F, θ1. θ2 is the angle.

■ Comparison of area values of the same region This can also be performed using the same procedure as the extraction of the same ROI 911 area in section (■).

As described above, it has been shown that image processing for various diagnoses can be performed easily and at high speed using the method according to the present invention.

Note that in the above embodiments, no particular mention is made of the alignment of images when superimposing images or extracting by designated area. This is due to the following reason.

Generally, in NMR-CT, imaging of the same part of the same patient using various pulse sequences is almost always done at the same time, and it can be said that all slice planes are the same. , even if no alignment is performed, there is no risk of misalignment as long as the reconstruction is performed using the same algorithm, so there is no problem in overlapping them as is. However, comparing two images of the same patient taken on different days. In this case, the slice plane may be slightly shifted, so it is necessary to perform each of the above image processing after confirming that the position is not shifted. In addition, in the present invention, a microprocessor dedicated to each image plane is used as a means of realizing a system having multiple image planes that can be accessed independently and simultaneously, but it is not necessarily a necessary condition that the microprocessor is a microprocessor. Needless to say, something equivalent to this can be used.

〔Effect of the invention〕

According to the present invention, by providing a plurality of image plane memories that can quickly and independently access various images of the same region reconstructed in NMR-CT, the following effects can be obtained.

(2) High-speed image handling is possible compared to the method used in conventional X-ray CT, etc., in which a specified image is accessed each time from an image file on a disk. Although this is not so necessary in X-ray CT, this invention is particularly effective because images are compared with each other in NMR-CT.

◎ Since it has multiple image brains that can be accessed independently, it can perform tasks that were considered difficult with conventional methods, such as overlapping images, subtracting images, extracting the same area from different images, or extracting the same amount of features. Processing becomes easier.

Since different types of images can be compared simultaneously and easily, the burden on those who interpret images (such as doctors) can be significantly reduced.

- If a diagnostic method for NMR images is established in the future, by converting that method into an algorithm and incorporating it into an image processing system based on this method, it will become possible to automatically diagnose NMR images using a computer.

[Brief explanation of drawings]

FIG. 1 is a configuration diagram showing an embodiment of the present invention, and FIG. 2A is a SE
FIG. 2B is an explanatory diagram for explaining the same region of interest in different images extracted at the same time. FIG. 2C is an explanatory diagram for explaining the same profile marker in different images. Figure 2D is an explanatory diagram for explaining the same density contour line of different types of images, Figure 2E is an explanatory diagram for explaining the distance between the same two points of different types of images, and Figure 2F is an explanatory diagram for explaining the same density isocapsular line of different types of images. An explanatory diagram for explaining the angle between three points,
Fig. 3 is an explanatory diagram for explaining the auxiliary pattern, Fig. 4 is an explanatory diagram showing the pulse sequence by the SR method, and Fig. 4A is an explanatory diagram for explaining the auxiliary pattern.
The figure is an explanatory diagram for explaining the coordinate axes of the specimen, and Figure 5 is S
An explanatory diagram showing a single slice pulse sequence according to the R method, FIG. 5A is an explanatory diagram for explaining the slice plane, FIG. 6 is an explanatory diagram showing a multi-slice pulse sequence according to the SR method, and FIG. 7 is an explanatory diagram showing the IR method. FIG. 8 is a schematic diagram showing a general system configuration of NMR-CT, and FIG. 9 is a configuration diagram showing a specific example of the computer system of FIG. 8. Code explanation 1 (1a to ld)...Image plane memory, 2
(2a~2g)...Ink face, 3 (3a~
3d)...auxiliary processor (microprocessor),
4...Host CPU (main processor), 5...
- RF interface, 6... Memory for host CPU, 7... Array processor, 8... Magnetic tape device (MT), 9... Magnetic tape device, 10 (10a
, 10b)...Image display, 11...System bus, 20...Computer system, 21.
...Static magnetic field coil, 22...Static magnetic field coil power supply, 2
3... Gradient magnetic field coil, 24... Power source for gradient magnetic field coil, 25... RF coil, 26... RF spectrometer, 27... Patient bed control. Agent Patent Attorney Akio Namiki Agent Patent Attorney Kiyoshi Matsuzaki Fig. 1 Fig. 2A Fig. 2B Fig. IR image SE Wei Fig. 2C Fig. 2D Fig. IR image SE image Fig. 2E IR Fig. 2F Figure 1 Rat SE image display 1
00 Display 10b Figure 3 (A) (B) Figure 4 Figure 5 Figure 9o Pulse Figure 5A Figure 8 Figure 9

Claims (1)

[Claims] 1) Two-dimensional detection of each part of the specimen from nuclear magnetic resonance signals obtained by applying a uniform static magnetic field, a static magnetic field with a linear gradient, and a high-frequency magnetic field pulse of a constant period to the specimen. A method using nuclear magnetic resonance comprising: a main processor that reconstructs a tomographic image; a memory that stores the reconstructed tomographic image data; and a display unit that displays a tomographic image of a specimen based on the data from the memory. In the tomographic image diagnostic apparatus, a plurality of auxiliary processors each retrieves a predetermined set of a plurality of sets of tomographic image data stored in a memory based on instructions from the main processor, either as is or after processing as appropriate; and the auxiliary processor. A plurality of image plane memories each storing data given through the image plane memory are provided, and by giving a predetermined instruction from the main processor to the auxiliary processor, a plurality of types of desired screens can be retrieved individually or simultaneously from the image plane memory. What is claimed is: 1. A tomographic imaging apparatus using nuclear magnetic resonance, which is capable of displaying. 2) The nuclear magnetic resonance tomographic image diagnostic apparatus according to claim 1, characterized in that a plurality of the display units are provided, and the image data from each image plane memory can be displayed independently. A tomographic imaging device using nuclear magnetic resonance.