JP2001061803A - Biomagnetic field analyzer - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the work load of inputting the conductance and to prevent failure in inputting the conductance. SOLUTION: This biomagnetic field analyzer for analyzing the magnetic field generated from an organism has a boundary data preparation part 13, a boundary inclusion relation editing part 16, an equation making calculation means 19 and a data analysis part 8. The boundary data preparation part 13 is to prepare a plurality of three-dimensional boundary data on a plurality of biological tissues based on a plurality of tomographic images of the organism. The boundary inclusion relation editing part 16 is to find the inclusion relation between the plurality of biological tissues based on the plurality of boundary data. The equation making calculation means 19 is to calculate a transfer matrix based on the inclusion relation and the conductance inside each boundary. The data analysis part 8 is to analyze the electrical activities of the organism based on the calculated transfer matrix and the measured value of the magnetic field generated from the organism.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a biomagnetic field analyzer for analyzing a measured value of a magnetic field or a potential (referred to as a biopotential or a biomagnetic field) generated with the activity of a living tissue and displaying the result.

[0002]

2. Description of the Related Art A variety of biomagnetic field analysis techniques have been discussed in the past. Typical examples are, for example, Reference 1: ACLBamard, IMDuck, MSLynn: The Applica
tion of Electromagnetic Theory to Electrocardiolog
y. Byophysical Journal, Vol. 7, pp. 443-462, 1967 Reference 2: B. Milan Horacek: Digital Model for Studies
in Magnetocardiograply.IEEE Trans.Mag., Vol.MAG-9, N
o.3, pp.440-444,1973 Reference 3: David B. Geeselowitz; On the Magnetic Field G
eneratod Outside an Inhomogeneous Volume Conductor
by Internal Current Sources.IEEE Tmas.Mag., Vol.MA
G-6, No.2, pp.346-347, 1970 Reference 4: Jukka Sarvas: Basic mathematical and electr
onagnetic conceptsof the biomagnetic inverse probl
em.Phys.Med.Biol., Vol.32, No.1, pp.11-22,1987 Reference 5: Jukka T. Nenonen: Solving the Inverce Proble
m in Magnetocardiography. IEEE Eng.Med. And Biol., P
p.487-496, 1994 Reference 6: RCBarr, TCPilkington, JPBoineau, MSSp
ach: Determining SurfacePorface Potentials from Cur
ent Dipoles, with Application to Electrocardiograph
y.IEEE Trans.Biomed.Eng., Vol.BME: 13, No.2, pp.88-92,
It is described in 1966.

The above-mentioned documents 1, 2, 4, 5, and 6 disclose a method of calculating a potential generated by a current source distribution in a living body. The calculation methods described in these documents are all the same, and use the calculation formula described below.

In a region divided by N boundaries S _j (j = 1... N) and having uniform conductivity in each segment, a current dipole J (r ′) (r ′) distributed in a region Ω ∈
Ω), the potential V (r) generated at the point r on the boundary satisfies the following boundary integral equation.

[0005]

(Equation 1)

Here, r is a point on the boundary S, σ _k ′ and σ _k ″ are the conductivity inside and outside the boundary S _k , and n (r ′) is the normal direction of the boundary at r′∈S _jベ ク ト ル Ω′dv is the integral over r ′ 積分 Ω, _∫s′jds
Represents integral with respect to r′∈S _j . If equation (1) is discretized, simultaneous linear equations can be obtained,

[0007]

(Equation 2)

Can be derived. Here, v is a vector in which the potentials V (r) at many points on the discretized boundary are arranged vertically, and v ^∞ is the potential in an infinite uniform medium having unit conductivity.

[0009]

(Equation 3)

H is a vector in which values at respective points are arranged vertically, and H is a matrix formed by discretizing equation (1).
The potential on the boundary can be obtained by solving equation (1.1) using a computer.

Reference 3 discloses a technique for calculating a magnetic field generated at a point outside the living body. The outline is as follows. If permeability is equal to the permeability mu ₀ of vacuum throughout the system, the secondary magnetic field generated at the observation point r _s by the current dipole J is

[0012]

(Equation 4)

[0013] Consider a piecewise uniform system composed of _N partial regions ｛Ω _n ｝ _{n = 1, N.} If the conductivity in the region Ω _n is σ _n , then in each partial region,

[0014]

(Equation 5)

Since the following holds, Geslowiz's expression:

[0016]

(Equation 6)

Is obtained. Here, the gamma _mn a boundary surface between Omega _m and Omega _n, n _mn is the unit normal vector on the boundary surface gamma _mn toward the Omega _n from Omega _m. Also, (m,
n) means the sum regarding the combination of all the areas touching each other. By discretizing equation (3),

[0018]

(Equation 7)

An equation of the form F is a vector in which the values of the secondary magnetic field measured by the plurality of magnetic sensors are arranged vertically, and A is a matrix obtained by discretizing the equation (3).
By using the equation (3.1), the secondary magnetic field can be calculated from the potential on the boundary.

[0020]

(Equation 8)

Is added to the discretized vector F _p , a magnetic field F = F _s + F _p generated outside the body can be obtained.

The magnetic field calculation method as described above is applied to a technique for estimating the electrical activity of a living tissue as described in References 5 and 6. A biomagnetic analysis method using the technique will be described below. First, a biological image of a subject is photographed using a magnetic resonance imaging apparatus (MRI) or the like. The region to be imaged is, for example, a region covering the chest including the heart if the target tissue is the heart, or the head if the target tissue is the brain. Generally, imaging is performed using a so-called multi-slice at a distance of 5 mm to 1 cm. . The photographed biological image is transferred to a computer system for analysis via a network, a magnetic tape, or the like.

In the computer system for analysis, first,
Three-dimensional biometric data is constructed from a multi-slice biotomographic image using a technique such as interpolation between slices, and the three-dimensional biometric data is displayed on a screen. As shown in FIG. 16, the operator traces the outline of the tissue on the displayed tomographic image and inputs the outline data to the computer.
Next, the traced contour data is connected, and the obtained tissue boundary is divided into a number of triangular elements. Data relating to the tissue boundary represented by the multiple triangular elements is referred to as boundary data. For example, to create a four-layer model of the scalp, skull, cerebrospinal fluid, and brain parenchyma, contour tracing and triangulation are applied to each tissue boundary and repeated four times. At the same time, for each of the four boundaries, the conductivity (typical value) outside and inside is input.

From the boundary data thus created,
(1) A simultaneous linear equation obtained by discretizing the equation is created. When creating the matrix H of simultaneous linear equations, the positions of vertices of a large number of triangular elements are referred to, and the integral of the boundary S _j is multiplied by the difference between σ _j ′ and σ _j ″.

For this reason, it is necessary to input in advance not only the inside of the boundary input for each boundary but also the conductivity outside the boundary.

The data analyzing means uses the matrix H and the matrix A when analyzing the electrical activity of the living tissue from the measured magnetic field value. The specific method is described in References 4 and 5.

In such a conventional biomagnetic field analysis, as described below, problem 1 “it is difficult to handle the boundary of an arbitrary structure” and problem 2 “the number of nodes increases when the structure becomes complicated” Increase, the matrix H becomes large-dimensional, and the operation time increases rapidly ".

First, regarding the problem 1, in the conventional biomagnetic analysis apparatus, when a new boundary is added, not only the conductivity inside and outside the added boundary is input, but also the boundary changed by the addition is added. In accordance with the inclusion relation, it is necessary to change the conductivity outside of the other boundary that originally existed, and there is a problem that the operation is complicated and error-prone.
Furthermore, the conductivity outside a certain boundary and the conductivity inside the boundary just outside the boundary must naturally be aligned with the same value, but the value determines the inclusion relationship between a plurality of adjacent boundaries. Since the decision has to be made while checking each item, there is a problem that not only the operation is complicated but also errors easily occur.

As an example, in the case of a tissue structure as shown in FIG. 17A, the conductivity outside the tissue boundary S1 is σ1, the conductivity inside the tissue boundary S1 is σ2, and the conductivity outside the tissue boundary S2 is σ1. 0, the conductivity inside the tissue boundary S2 is set as σ1. As shown in FIG. 17B, when an attempt is made to add a boundary S3 with an inner conductivity of σ3 to this structure, σ1 is set to the conductivity outside the added boundary S3, and σ3 is set to the inner conductivity. , The conductivity outside the boundary S1 whose inclusion relation has changed is changed from σ1 to σ3.
Need to be changed to

It should be noted that the conductivity inside the boundary S2 and the conductivity outside the boundary S3 have the same value, the conductivity inside the boundary S3 and the conductivity outside the boundary S1. Must be set to the same value. If the correct value is not always input in consideration of the inclusion relation between boundaries, a large error will be included in the calculation result. This example is a comparatively simple example, but when dealing with a more complicated structure boundary, setting the conductivity between the inside and the outside tends to be very complicated and easy to be mistaken.

Next, regarding the problem 2, a large number of actual living tissues are complexly combined, and each tissue has a different conductivity. Therefore, in order to calculate a more accurate magnetic diagram (magnetic field distribution), a biological tissue model having a complicated structure having many boundaries is required. Since the order of the matrix H of the simultaneous linear equation (1.1) is equal to the number of points on all the boundaries, the order of the matrix H becomes enormous if the number of boundaries is to be increased.

Furthermore, the amount of operation required to solve the simultaneous linear equations is proportional to the cube of the order of the matrix H. Therefore, an increase in the number of boundaries results in an abrupt increase in the amount of operation. For this reason, there was a problem that a model having a very complicated boundary structure could not be adopted practically.

[0033]

SUMMARY OF THE INVENTION It is an object of the present invention to provide a biomagnetic field analyzer capable of reducing the workload of inputting conductivity and eliminating erroneous input of conductivity.
Another object of the present invention is to provide a biomagnetic field analyzing apparatus capable of realizing high-speed and high-accuracy biomagnetic field analysis processing.

[0034]

Means for Solving the Problems (1) The present invention relates to a biomagnetic field analyzing apparatus for analyzing a magnetic field generated from a living body.
Boundary data generating means for generating a plurality of three-dimensional boundary data relating to a plurality of living tissues from a plurality of tomographic images relating to the living body; and boundary inclusion for obtaining an inclusion relationship between the plurality of living tissues based on the plurality of boundary data. Relation editing means, transfer matrix calculating means for calculating a transfer matrix for deriving a magnetic map from a current distribution based on the inclusion relation and the conductivity inside each boundary, and the calculated transfer matrix and a magnetic field generated from a living body. Data analyzing means for analyzing the electrical activity of the living body based on the measured values.

(2) The present invention relates to a biomagnetic field analyzing apparatus for analyzing a magnetic field generated from a living body, wherein boundary data creation for creating a plurality of three-dimensional boundary data concerning a plurality of living tissues from a plurality of tomographic images concerning the living body. Means, a means for deleting any boundary data from the plurality of created boundary data, and adding any boundary data; and a magnetic map from a current distribution based on the plurality of boundary data obtained by the creation, deletion, and addition. Transfer matrix calculating means for calculating a transfer matrix for deriving the data, and data analysis means for analyzing the electrical activity of the living body based on the calculated transfer matrix and the measured value of the magnetic field generated from the living body. Features.

[0036]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The preferred embodiments of the present invention will be described below in detail with reference to the drawings. (First Embodiment) FIG. 1 shows a configuration of a biomagnetic field analyzing apparatus according to a first embodiment of the present invention, and FIG. 2 shows a configuration of a biological model creating unit of FIG. Magnetic resonance imaging (MRI) and X-ray computed tomography (X-ray C
For example, multi-slice tomographic image data captured by the biological image capturing apparatus 1 such as T) is taken into the computer system 4 via the biological image transfer unit 5 and temporarily stored in a storage unit (not shown). At the time of storage, an image inspection number is assigned in inspection units (a plurality of tomographic image data are collected in one inspection) so that they can be identified when recalled later.
All data handled here, such as three-dimensional image data and boundary data created later using the tomographic image data of the living body, are stored in association with this image examination number. For this association, a series of numbers called an image inspection number is used. As the image examination number, the same examination number as that already assigned when the image is taken by the biological image pickup apparatus 1 can be used, or a new examination number unique to the analyzer can be assigned. It is. Although called by the name of the image inspection number, it is actually composed of any characters such as English characters, kana, and kanji.

The stored tomographic image data is called by the three-dimensional imaging means 10, converted into three-dimensional image data using a technique such as inter-slice interpolation, and temporarily stored in storage means (not shown). The three-dimensional image display means 10 projects and displays the three-dimensional image data called from the storage means two-dimensionally. At the time of this projection, the three-dimensional image can be displayed as a diagram viewed from an arbitrary direction by changing the viewpoint and the viewing direction. Further, by cutting at a cross section having a plurality of arbitrary positions and directions, it is possible to project an arbitrary cross section. The image projected in this manner is displayed on display means (not shown).

The operation of this embodiment will be described below together with the operation procedure. FIG. 3 shows a front page of an operation screen for creating a biological model. The outer frame of the screen represents the entire operation screen for creating a ventricle model. Usually, this is the entire display screen, but it is also possible to display the entire screen as a part thereof. In the child window of the operation screen,
The image examination number related data list is displayed. This list contains image inspection numbers (for example, AA00CCDD)
The names of all data related to are listed.
Information recorded on the transferred tomographic image data such as patient ID, patient name, gender, age, and the like is automatically displayed, but can be changed on the image examination number related data list.

Below the operation screen, a list of identification numbers for each read tomographic image is displayed. One of the items in this list, for example, SS out of SS0D000000 / A
0D000000 is the examination number of the tomographic image, and A is a symbol representing one of the data sets. SS0
Click on the “Open” part displayed to the right of D000000 / A to see SS0D0000
A series of tomographic images related to 00 / A is displayed.

Below the tomographic image identification number, an operation screen for a three-dimensional image is displayed. The display of “3D image A” means that there is one 3D image that has already been created, and by clicking “Open” on the right side, the 3D image created by the 3D imaging means 10 is displayed. The displayed three-dimensional image is displayed. When "Create" on the right side of "New 3D image" is clicked, a new 3D image is created based on one or more of the tomographic images. Parameters necessary for creating a three-dimensional image, such as the voxel size of the three-dimensional image, are specified after selecting “create”.

When scrolling downward, an operation screen relating to boundary data is displayed as shown in FIG. Here, the operator can delete arbitrarily selected boundary data by using a deletion function called by operating a keyboard or the like in a state where some boundary is selected. When you select "Create" to the right of "New Boundary", a new screen is displayed,
The selected three-dimensional image is displayed by selecting the three-dimensional image data on which the boundary data is created. further,
By selecting “viewpoint” and “section”, the viewpoint and cutting of the three-dimensional image are changed, and an arbitrary section is displayed on the screen. It is possible to display another section parallel to the designated section by changing the "section interval", the "number of sheets", and the "display section".

Here, when "input" is selected, the contour input means 12 is activated. As shown in FIGS. 5A and 5B, the operator traces the outline of the target tissue on the tomographic image using a mouse or a touch pen provided on the outline input unit 12. Thereby, the contour data of the tissue is input.

Further, when "edit" is selected, it is possible to adjust the shape of the contour after input. Next, when the division roughness is input and “division” is selected, based on the continuous contour obtained based on the plurality of tomographic images shown in FIG.
The boundary data creating means 13 recognizes the boundary of the tissue, and divides this boundary into a number of triangular elements as shown in FIG. Various methods have been proposed for the triangulation, and the triangulation is realized by using those methods. The above boundary creation operation does not need to be performed for all the boundaries at once. After creating some boundaries, it is possible to create another boundary even after performing other operations such as transfer matrix calculation It is.

Below the boundary operation screen, a list of the boundary inclusion relation table is displayed as shown in FIG.
The inclusion relation between the boundaries indicates the relation of which boundary is inside which boundary or which boundary surrounds which boundary, and in this embodiment, this inclusion relation Is automatically calculated by the inclusion relation editing means 16 based on the boundary data. Here, when "Create" is clicked on the right of "New inclusion relation", an inclusion relation table screen is displayed. Here, an arbitrary boundary is selected from the boundary list of the image inspection number related data list (click ▽ next to the boundary name), and the electric conductivity inside each boundary is input to the electric conductivity column. The conductivity inside the boundary can be selected from typical values prepared in advance for each tissue such as “ventricular muscle” and “lung”. In this case, information corresponding to the electric conductivity, for example, “ventricular muscle” or “lung” is displayed in the electric conductivity column instead of a specific numerical value. The specific value of the conductivity represented by the “ventricular muscle” or “lung” can be set and changed separately.

Here, the check box on the left side of the boundary name selects the outermost boundary among a plurality of boundaries, that is, the boundary surrounding all the other boundaries. This is equivalent to setting the outer conductivity to zero. In principle, it is also meaningful to specify a plurality of boundaries at the outermost position. Such a specification is also permitted, but in the range of normal use, the outermost boundary is one, so When the outermost boundary is specified, a warning screen for calling attention is displayed. When the setting of the boundary inclusion relation is completed without specifying any outermost boundary, the boundary is automatically detected and set.

The automatic determination of the outermost boundary uses the following method. First, a solid angle is obtained between a plurality of boundaries. The resulting solid angle table is shown in FIG. FIG. 8 shows a tissue structure corresponding to the solid angle table of FIG. As the solid angle, for example, the solid angle of the boundary S1 is 0 when viewed from a certain point on the boundary S2, which means that the boundary S2 is not inside the boundary S1, that is, the boundary S2 is included in the boundary S1. It is not done. Conversely, for example, when viewed from a point on the boundary S2, the solid angle of the boundary S3 is 4π.
The inside of the boundary S3, that is, the boundary S2 indicates that it is included in the boundary S3.

In this sense, if the solid angles when viewing another boundary from a point on a certain boundary are all 0, it is determined that the boundary is the outermost, and any one of the solid angles has a solid angle. If it is not 0, it can be determined that the boundary is not the outermost.
However, it can be said that it is practical to make a determination based on, for example, whether the absolute value is smaller than 0.1 in consideration of a calculation error. In addition, it is possible to select some of the boundaries in the boundary inclusion relation table and delete the selected boundaries from the table as not being used for analysis by a deletion function called by the keyboard and the mouse.

In the above operation, the boundaries other than the outermost boundary are provided with the inner conductivity, but are not provided with the outer conductivity (not specified or input). However, the potential calculation by equation (1.1) and
The magnetic field calculation by the equation (3.1) requires not only the conductivity inside the boundary but also the value of the conductivity outside the boundary. In the present embodiment, the inclusion relation is automatically determined by the inclusion relation editing means 16 instead of the manual input as in the related art, and each boundary is determined in accordance with the inclusion relation and the inner conductivity already input for each boundary. Automatically determine (set) outer conductivity
Then, the result is stored in the boundary inclusion relation storage means 17.

As an initial setting of the process of automatically determining the outer conductivity, first, a solid angle when a boundary other than the above-mentioned boundary is viewed from a point set on a certain boundary is calculated. . Next, an “exclusion list” of boundaries is created. The process of determining the inclusion relationship is to sequentially determine boundaries from the outside to the inside, and the “exclusion list” refers to a boundary that has been determined as outside in the process of this determination process.
This is a list for excluding from the target of the determination processing of the next loop. All items in the "exclusion list" are initially set to non-exclusion.
The variable “outer conductivity” used in the procedure
Is initially set to 0. After these initializations, give the "exclusion list" and "outer conductivity" and call the following procedure.

In this procedure, based on the "exclusion list" L, all boundaries that have not been excluded, that is, all remaining boundaries that have not yet been determined as outside, are recognized. When looking at all other boundaries recognized from a certain boundary, search for a boundary Si that satisfies the condition that all solid angles at which the boundary is viewed become 0, and determine the boundary as outside in the immediately preceding process. And the conductivity outside the boundary is set to the value of the variable “outside conductivity”, that is, the value of the conductivity inside the boundary determined to be outside in the immediately preceding process. If no boundary satisfying such a condition is found, this procedure ends. Otherwise, this procedure is called recursively for each boundary Si where the solid angles are all 0. At the time of this call, "outer conductivity"
Is given the value of the conductivity inside the boundary Si, and L + Si is given to the "exclusion list".

This process will be described in detail with reference to FIGS. 7 and 8. First, for all boundaries, the boundary where all solid angles to all other boundaries become 0 is determined in S4. The boundary S4 is determined as the outermost boundary. Next, among the boundaries where only the solid angle to the boundary S4 is not 0, that is, among the boundaries S1, S2, S3, and S5 other than the boundary S4 that has just been determined and is listed in the exclusion list, another boundary is defined. Search for a boundary where all the observed solid angles are 0.
In this case, two boundaries S3 and S5 correspond. It is determined that these boundaries S3 and S5 are boundaries immediately inside the boundary S4 that has been determined immediately before and has been excluded. Therefore, the conductivity outside these boundaries S3 and S5 is determined by the boundary S4
To the conductivity inside.

Next, paying attention to the boundary S5, a search is made for a boundary where the solid angle at which the other boundary is seen is zero among the boundaries S4 and the boundaries S1, S2 and S3 other than the boundary S5. Since this boundary does not exist, it is determined that there is no boundary inside the boundary S5. Similarly, paying attention to the boundary S3,
Among the boundaries S4 and the boundaries S1, S2, and S5 other than the boundary S3, a boundary in which the solid angles of the other boundaries are all 0 is searched. In this case, two boundaries S1 and S2 correspond. It is determined that these boundaries S1 and S2 are boundaries immediately inside the boundary S3 that has been determined immediately before and has been excluded.
Therefore, the conductivity outside these boundaries S1 and S2 is set to the conductivity inside the boundary S3. Here, the boundaries S4, S
Since there is no boundary where the solid angles viewed from other than 3, S1 or other than S4, S3, and S2 are all 0, this procedure is completed. At this point, the conductivity outside all boundaries is automatically set correctly.

An operation screen relating to the channel arrangement is displayed below the boundary inclusion relation table. The SQUID driven by the SQUID drive circuit 3 as shown in FIG.
The ID magnetic sensor 2 is a device equipped with a plurality of coils using a so-called superconducting quantum interference device capable of detecting a minute magnetic field generated in a living body (here, one coil constitutes one channel). Of these channels (coils), those actually used are some channels selected by the operator. Here, it is necessary to align the selected channel with the body of the subject, that is, finally with respect to the three-dimensional image and the tomographic image. This alignment will be described below.

As for the channel arrangement (the arrangement relation of the selected coils), the channel arrangement normally used has already been initialized. If the user wants to change the already set channel arrangement, the user can change the channel arrangement by selecting “change” on the right of any item in the list. Items that can be changed include the use / non-use of each channel of the magnetic sensor 2, the position and orientation of the magnetic sensor 2 with respect to the body, and "body characteristic points" representing the positions of markers in the tomographic image.

FIGS. 9 (a) and 9 (b) are screens for changing the "body characteristic point", in which the + mark indicates the "body characteristic point".
Represents the position of. The “fuselage feature points” are superimposed and displayed on the three-dimensional image and the tomographic image of the torso, and their positions can be arbitrarily changed using a keyboard and a mouse. Also, these small points in FIG.
The characteristic points of the torso of the subject and the positions of points on the body surface measured together with the magnetocardiogram at the time of measurement of the magnetocardiogram are `` torso index points '',
These are stored in advance in channel arrangement storage means (not shown). "Body index point" is the magnetic sensor 2
By moving and rotating the position of the sensor 2 in parallel using the keyboard and the mouse, the positions of the sensor 2 and the marker positions of all the “body index points” are linked on the screen. It is configured to move. Because of this configuration, the position of the marker on the tomographic image and the position of the torso feature point “torso index point” measured during the magnetocardiogram measurement are aligned on the screen, whereby the position of the torso and the sensor 2 can be accurately aligned. The position and orientation of the aligned sensor 2 and the positions of the plurality of "body index points" are stored in a channel arrangement storage unit (not shown).

Next, below this “channel arrangement”, a list of transfer matrices is displayed. Clicking "Create" to the right of "New Transfer Matrix" displays a transfer matrix calculation setting screen as shown in FIG. 10, where the inclusion relationship and channel arrangement are set, and "Execute Calculation" is selected.
The matrices H and A of the formula (1.1) and the formula (2.1) are created, the equations are solved, and a transfer matrix necessary for creating a magnetic map from the current distribution is calculated by the equation creating operation means 19, It is stored in the transfer matrix storage means 20. The stored transfer matrix is used by the data analysis means 8 as basic data for analyzing the electrical activity in the living body by, for example, a magnetic diagram.

If the image inspection number related data list as described above and the information in the screen called up therefrom are the same type of information, a copy can be created by calling the copy function with a keyboard or a mouse. . In particular, in the image inspection number related data list,
In the “tomographic image list”, “3D image list”, “boundary list”, “boundary inclusion relation table list”, “channel arrangement list”, and “transfer matrix list”, a plurality of similar types of information are set. . When a copy function is called after selecting a copy source item in these lists, a new item having the same information as the copy source is created.

According to the present embodiment, once, a biological model (a three-dimensional boundary structure represented by three-dimensional boundary data)
After creating the data and analyzing it, add arbitrary boundary data to the boundary data created or duplicated based on the tomographic image, or delete boundary data that is considered unnecessary for analysis. Is possible, it is possible to perform analysis using a real biological model irrespective of a definitive biological model, so that more useful and practical analysis can be performed. In addition, since unnecessary boundary data can be deleted, an unnecessary increase in the amount of calculation can be prevented. Furthermore, since the inclusion relation can be automatically determined and the conductivity outside each boundary can be automatically and systematically set without contradiction based on the inclusion relation, the work load is reduced and erroneous input is prevented. be able to.

(Second Embodiment) In the second embodiment, an arbitrary boundary can be handled, and a boundary where no current source exists can be eliminated, and equations (1.1) and (3.1) are used. Using a new method different from the conventional potential calculation method
The calculation of the potential and the calculation of the magnetic field are performed. The method is
Compared with the first embodiment, the size of the transfer matrix can be reduced, and the boundary can be handled by an arbitrary biological model as in the first embodiment. In the method of analyzing electrical activity in a living body by an iterative method, the number of magnetic field calculations is performed many times in one analysis, so that the amount of calculation can be reduced, and the boundary of a complicated structure according to the user's request is considered There is an effect that highly accurate magnetic field calculation becomes possible.

In the first embodiment, in the setting processing of the boundary inclusion relation table, processing is performed so as to automatically set the electrical conductivity outside and inside the boundary. In addition, it is configured to set a list of boundaries that are adjacent when outside the boundary and a list of boundaries that are adjacent inside. This setting process is performed as follows.

As an initial setting, first, the position of a certain point on the boundary is obtained, and the solid angle when all other boundaries are viewed from that point is calculated. Next, an “exclusion list” of boundaries is created. All items in the "exclusion list" are initially set to non-exclusion.
Also, 0 for "outside conductivity" and "infinity" for "outside boundary"
And initial settings. After these initializations, the following procedure is called with an "exclusion list", "outside conductivity", and "outside boundary".

Procedure: Based on the given “exclusion list” L, search for a boundary Si where all solid angles at which the other boundary is seen are 0 among boundaries that are not excluded, and outside the boundary Si The value of “outside conductivity” is set to 0 for the conductivity, and “infinity” is stored as the “outside boundary” of the boundary Si. Further, an "inner boundary list of boundaries set as" outer boundaries "" is created, and all found Si are stored. Of course, even "infinity" has "inner boundary list"
Is set. If no boundary Si is found, this procedure ends. Otherwise, this procedure is recursively called for each boundary Si. In this call, the value of the conductivity inside the boundary Si is given to the “outside conductivity”, and L + Si is given to the “exclusion list”.

At the end of the above procedure, for all boundaries, "outer conductivity (of each boundary)", "inner conductivity (of each boundary)", "outer boundary (of each boundary)", "Inner boundary list (for each boundary)" is set.

The potential calculation method of this embodiment is based on a boundary integral equation in a single region. If there are multiple boundaries, they are connected by boundary conditions to derive the entire equation. The boundary integral equation for a region with a single boundary is expressed as:

[0065]

(Equation 9)

Here, r ′ is a point on the boundary S, and σ is a point on the boundary S
, And θ (r ′) is the solid angle of the boundary S viewed from r ′. Boundary unit normal vector n (r)
Is always defined outward of the boundary. Ψ ^電位 is a potential generated in the infinite uniform medium of unit conductivity by the electromotive force in the region Ω, and is expressed by the following equation.

[0067]

(Equation 10)

When the above equation (2-1) is discretized, the following equation is obtained.

[0069]

[Equation 11]

[0070] Here, H is a left-hand side of equation (2-1), G is a matrix obtained by discretizing the second term on the right side, v is the V (r), [psi ^∞ is discretized (2-2) below Vector. q is a vector obtained by discretizing the normal derivative ▽ V (r) · n (r) of V (r) at the boundary.

In the case of a problem having a plurality of regions Ω _i (i = 1... N), in each region,

[0072]

(Equation 12)

The following equation is obtained. Where σ
^{Ωi is} _the conductivity _{of the} ^{region Omega} _i, boundary number surrounding the M _i region Omega _i,
H _jk ^.OMEGA.i puts r 'on j-th boundary surrounding the region Ω _i, (2-1) is a discretized matrix left side of expression when placed r a k-th boundary. G _jk ^Ωi is similarly the second on the right side
This is a matrix in which terms are discretized. In addition, each boundary S _i (i = 1
Boundary conditions such as the following are present in ... N _B). v _j
^.OMEGA.i and q _j1 ^.OMEGA.i are potential and normal derivative on j-th boundary region Omega _i.

[0074]

(Equation 13)

[0075]

[Equation 14]

Here, σ _{i +} and σ _i− are the conductivity outside and inside the boundary S _i , respectively, and q _{i +} and q _i− are the normal derivatives as seen from the area outside and inside S _i , respectively.

When equation (2-4) is made simultaneous for all the regions and equation (2-4) is substituted, the following equation is obtained.

[0078]

(Equation 15)

Here, H _jk ^- and H _jk ⁺ are r ′ at the j-th boundary and k ′ at the k-th boundary, respectively, when the boundary adjoining the inside and outside of the j-th boundary is k or j = k. discretized matrix left side of the equation (2-1) when spaced r, otherwise a 0 matrix of the same size matrix, G _jk ^- is included k to "inner boundary list" of the j-th boundary Have been
Alternatively, when j = k, a matrix obtained by discretizing the second term on the right side of the equation (2-1) when r ′ is placed at the j-th boundary and r is placed at the k-th boundary. G _jk ⁺ is a matrix that becomes 0 matrix, and when the “outer boundary” of the j-th boundary is k, r ′ is placed on the j-th boundary and r is placed on the k-th boundary. A matrix G _jk that discretizes the second term on the right side,
When the boundary k is included in the “inner boundary list” of the “outer boundary” of the j-th boundary, a matrix _{obtained by multiplying} G _jk by σ _k − / σ _k +, and otherwise a 0 matrix of the same size Is a matrix.

[0080] Also, v _i is and q _i - the normal differentiation when viewed from the inside of the potential and boundaries on the i-th boundary, [psi _i ^∞ is the potential during Mugen homogeneous medium on the boundary i. The block matrix on the i-th row that constitutes the matrix on the right and left sides is composed of two sets of upper and lower matrices as shown in the first and last rows of the matrix unless the boundary i is the outermost boundary. only if the boundary i is the outermost boundary, as a portion surrounded by a broken line enclosing the (2-6a) equation, H _ij ^- or G _ij ^- only composed. In the above equations (2-6a) and (2-6b), (2-5
By substituting the equations a) and (2-5b), the following equation is obtained.

[0081]

(Equation 16)

L _i (i = 1... N _L ) is a non-outermost boundary,
N _L represents the number. The expression (2-7) is expressed as v ₁ .
_NB and q _L1 ^- ... q _LNL ^- a a simultaneous linear equations and unknowns, by solving this in a known manner, it is possible to obtain the following equation.

[0083]

[Equation 17]

Also, using the equation (3.1),

[0085]

(Equation 18)

Thus, there is obtained an equation that the magnetic field measured by the sensor is calculated by the potential in the infinitely uniform medium. When (1) Infinity current source to any one of the areas of potential V ^∞ is the plurality of regions of uniform medium is present, which element also may become a value other than zero, (2 -9) expression [Phi ^∞, except elements corresponding to a region where a current source is present, 0
become. In many applications of biomagnetism, only the current source in the myocardium or brain is considered, so even when various other tissues are modeled in detail, the area where the current source exists is small (often one). It is. For example, when the current source exists only in one region, the number s of the boundary surrounding the outside of the region is set as the number s ₁ ,..., S of Ns boundaries included in the “inner boundary list” of the boundary. _When expressed as _Ns ,

[0087]

[Equation 19]

The following equation can be obtained, and many boundaries where there is no current source can be eliminated. Instead of the equations (1.1) and (3.1) by the biomagnetic analysis means,
By using the expression (2-10), the amount of calculation can be reduced.

As an effect of the second embodiment, in particular, when an analysis is performed in consideration of the complex structure of a living body in detail, various tissues other than the brain and heart where electrical activity exists, such as blood It is necessary to consider many tissues such as skeletal muscle, fat, lung, and bone. However, considering such a complicated structure, it is necessary to consider the boundary surface of many regions. In such a case, the conventional method takes a long calculation time in the magnetic field calculation by the biomagnetic analysis means, so that it was not possible to substantially consider the complicated structure of the living body. On the other hand, according to the method of the present embodiment, many boundaries where no current source is present are erased, so that even when various tissues other than the tissue such as the brain and heart where the current source is present are considered, the method is substantially performed. Since the amount of calculation does not increase, a complicated structure can be considered.

Further, similarly to the first embodiment, the conductivity outside each boundary and the inclusion relation between the boundaries are automatically determined, and the operator need only specify the conductivity inside each boundary. Therefore, even when there are many boundaries, there is no need to consider complicated inclusion relations, and the operation is simplified, and there is an effect that errors are less likely to occur.

(Third Embodiment) In the first embodiment, as the boundary data, the boundary created by using the contour input means and the boundary data creating means is synthesized as it is, and used for creating equations relating to magnetic field calculation. . In order to perform highly accurate magnetic field calculation by this method, it is necessary to reduce the distance between cross sections when tracing a plurality of contours. However, in this case, the number of triangle elements created by the next boundary data creation means increases. When the number of triangular elements increases, the amount of computation of magnetic field calculation using the transfer matrix obtained by creating equations relating to magnetic field calculation and solving these equations increases, and eventually, analysis of bioelectric activity takes a lot of time .

When the distance between the cross sections is reduced, a large number of triangular elements generated by the boundary data generating means include those having a flat shape rather than a regular shape. The existence of a flat triangular element is
It degrades the accuracy of the magnetic field calculation and therefore
In some cases, the analysis accuracy of bioelectric activity also deteriorates.

In the third embodiment, as shown in FIG. 11, a boundary data improving means 21 for improving the triangulation of the boundary is provided by a method described later.
Close to regular triangle elements and reduce the number of triangle elements,
Improved boundary data is obtained.

As a result, the magnetic field calculation with high accuracy can be realized with a small amount of calculation, thereby shortening the time required for analyzing the bioelectric activity, improving the analysis accuracy, and making the analysis of the bioelectric phenomenon practical. It becomes possible to do.

The improvement of the boundary data is performed as follows. FIG. 12A schematically shows the boundary data before the improvement as it is created by the boundary data creating means 13. By performing each process of “inflation process”, “polyhedral projection process”, and “return process” on the boundary data before this improvement, regular triangle elements or triangle elements close thereto are increased, and the number of triangle elements is increased. Improve the reduced boundary data (biological model).

First, in the “inflation process”, forces repelling each other are applied to the respective nodes of the triangulation of the biological model. The repulsive force acting on the node i is calculated by the following equation, for example.

[0097]

(Equation 20)

Here, r _i (i = 1... N) is the coordinates of the node. In addition, an attractive force is applied to adjacent nodes. When the node j (j = 1... M) is adjacent to the node i, the attractive force acting on the node i is calculated by the following formula as an example.

[0099]

(Equation 21)

Next, each node is slightly moved based on the repulsive force and attractive force acting on each contact point. Moving distance and vector d _i representing the direction with respect to the node i is calculated by the following equation, for example.

[0101]

(Equation 22)

In this example, it is devised that the one-time moving distance does not exceed the distance d. In this way, if there are nodes that are too close to each other in the original triangulation, strong repulsion prevents the nodes from flying far and ensures that the inflation process is performed stably. When the above steps are repeated a plurality of times, the fine irregularities on the curved surface disappear, the repulsive force and the attractive force balance, and the node does not move much. Here, the inflated intermediate model formed after the repetition is completed is schematically shown in FIG.

Next, as [polyhedral projection processing], a point inside the inflated intermediate model is set as shown in FIG.
A polyhedron model such as This polyhedron model has a roughly spherical shape, and is divided by triangular elements having substantially the same shape as substantially equilateral triangular elements. The intersection of the half line passing through each vertex of the polyhedron model with the center of the polyhedron model as the starting point and the triangular element of the inflated intermediate model is determined, and all nodes of the polyhedron model are moved to the intersection. FIG. 12C schematically shows the intermediate model thus formed. Generally, the number of vertices and the number of triangular elements of the polyhedral model are smaller than those of the original biological model of FIG.

The inflated intermediate model shown in FIG. 12B becomes a convex polyhedron if there is no large projection or the like in the shape of the biological model before inflation shown in FIG. 12A. In this case, the intersection of the half line is always one. One. However, if the biological model before inflation in FIG. 12A has large protrusions or the topology is not the same as a spherical surface, there may be a plurality of intersections with a half line. In this case, it is possible to execute the “polyhedron projection process” by selecting a point closest or farthest from the center of the polyhedron. However,
In this case, it should be noted that the shape of the curved surface subdivided according to the present embodiment is significantly different from the original curved surface. This problem can be solved by applying a fourth embodiment or a fifth embodiment described later.

Next, as [return processing], information indicating which vertex of the projected polyhedron model is located at which triangular element of the intermediate model in FIG. 12B and at which position within the triangular element is located. Then, the position corresponding to the position in the original biological model shown in FIG. 12A is obtained, and each vertex of the intermediate model shown in FIG. 12C is moved to that position. FIG. 12D shows the improved model thus created. The position of a certain point in the triangular element is represented by using a known coordinate system called area coordinates. The boundary data of the biological model thus improved is registered in the boundary data storage unit 15 as a new boundary.

The boundary data of the improved model created by the above steps is generally smaller in the number of vertices and triangle elements than the boundary data of the original biological model in FIG. Can represent the shape of the boundary with the same precision as the boundary data of. In addition, it is divided by a triangle element close to an equilateral triangle element.

Therefore, if the magnetic field calculation is performed using the boundary data improved by the present embodiment, a highly accurate magnetic field calculation can be performed in a shorter calculation time than when the boundary data created by the boundary data creation means is used for the magnetic field calculation. It can be performed.

(Fourth Embodiment) As a fourth embodiment, a method of improving boundary data different from the above-described third embodiment will be described. The boundary data improvement method according to the present embodiment refers to two boundary data having substantially the same shape (here, referred to as “surface description boundary data” and “improvement target boundary data”) and performs new improvement. This is a technique for creating the boundary data.

At this time, “reference length” is referred to as an amount for specifying the fineness of division. The “reference length” is an amount indicating the approximate side length of the triangular element to be divided. As the amount to specify the fineness of division, it is also possible to adopt various criteria such as the area of the triangular element, the minimum value of the side length, the angle of the adjacent triangular element, in addition to the "reference length" is there. Here, the case where the “reference length” is adopted will be described. However, it is very easy to change to adopt another reference amount or to use a combination of a plurality of reference amounts.

When this method is applied, the “surface description boundary data” and the “improvement target boundary data” may have exactly the same shape.

First, as [initialization processing], each node of the “improvement target boundary data” is projected onto the surface of the “surface description boundary data” using exactly the same method as the polyhedral projection processing of the third embodiment. I do.

Next, as the “potential calculation process (1)”, first, distances are obtained for all combinations of nodes of “boundary data to be improved”. There are a method of obtaining the distance and a method of using the Euclidean distance in a three-dimensional space, and a method of obtaining the shortest distance along a curved surface represented by “surface description boundary data”. Next, a potential is obtained. When a three-dimensional Euclidean distance is used, a function m / which is inversely proportional to the distance is used as a potential created by one node.
The (4πr ^n), the logarithm of the reciprocal of the distance when obtaining the shortest distance along the curved surface - (m / 2π) lnr or m / (4
πr ⁿ⁾ is used. m is the mass of the node, and the same mass may be given to all the nodes, or a different value may be given to each node. By giving a different value to each node, it is possible to change the fineness of division depending on the location.
n is typically 1 but other values can be employed. In order to obtain the total potential energy, the potential φi, j created at the position of the node j by the mass m at one node i is multiplied by the mass m of the node j, and this is added for all combinations of ij. That is,

[0113]

(Equation 23)

Thus, the total potential energy is calculated.

In the case of the boundary shape as shown in FIG.
When the dimensional Euclidean distance is used, a large repulsive force is generated because the distance between the point A and the point B is short, and no nodes are arranged near the points A and B. However, by using the distance along the curved surface, Such a problem can be solved.

Next, as the [potential calculation process (2)], if the shortest distance along the curved surface is used in the above-described potential calculation process (1), the points C and C in FIG.
Due to the difference between the shortest path between D and the shortest path between points C and E, the gradient of the potential created by point C at point D or point E will be in different directions. This causes calculation instability in the later “partial pass / fail criterion minimizing process”. In order to avoid this problem, the potential created by one mass m is given by-(m / 2π) 1nr + mcr
^It is effective to use a method using ^k (k and c are appropriate real numbers) and a method of making the distance between nodes infinite except for nodes adjacent to each other when creating a distance matrix. Alternatively, the method described below can be used.

This method uses a boundary element method (or finite element method) to calculate the potential. That is, in each triangle element existing on the surface of the “surface description boundary”, the equation − ▽ ² φ = σ
Holds, and the flux q
= ∂φ / ∂n is added to determine the potential φ by adding a boundary condition. Here, σ is the density, and here, the density distribution is such that the mass m is concentrated at the position of the node of the “improvement target boundary”. The specific method of obtaining φ is almost the same as that shown in the second embodiment, and it should be noted that the divided area is simplified two-dimensionally and that the outermost boundary does not exist. It is preferable to use a two-dimensional version of-(m / 2π) 1nr as the Green function. Algebraic equation corresponding to equation (2-7) thus obtained

[0118]

(Equation 24)

In the following equation:

[0120]

(Equation 25)

The potential at the node on the “surface description boundary” can be obtained by solving the simultaneous equations. R is a suitable scalar. By interpolating the potential, the potential φ _{i, j} at the node of the “improvement target boundary” can be obtained. The total potential energy is obtained by equation (1.1).

Next, as [triangulation division local correction processing],
In the boundary data correction process, a process of rearranging triangulations of triangular elements adjacent to each other in the “improvement target boundary” is performed. The method of rearrangement is shown in FIGS.
Various methods shown in FIGS. 14B, 14C and 14D are used in combination. Specifically, which triangle element is rearranged and how it is rearranged depends on the optimization algorithm used in the “division pass / fail standard minimization process”.

Next, in the [node movement process], some nodes in the "improvement target boundary" are designated in a random or designated direction according to the value determined in the "partition pass / fail criterion minimization process". Move by the distance

For example, the triangle element T of “surface description boundary”
The node r _i with "improved target boundary" present on to move in the direction of the vector t by a distance d of, if it is determined at "split acceptability criteria minimization process", triangle vector t elements T After projecting on the plane of, it is converted into a unit vector and set as t ′. t Request 'a vector obtained by multiplying the distance d r _i addition to r _i'. If r _i ′ is inside the triangle element T, then end; otherwise, the edges of the triangle element and r _i
'R _i the intersection of the straight line connecting the "Distant, r _i' r _i and r _i" Distant, r _i upon the adjacent triangular elements T ''
Is moved again by the distance d 'in the direction of t'.

Next, as a [divided pass / fail criterion calculation process],
A division pass / fail criterion E is calculated by the following equation.

[0126]

(Equation 26)

Here, d _i is the length of the i-th side of the M sides of the “improvement target boundary”. D is a “reference length” that specifies a desired length of the side of the triangulation. n
Is typically 1, but may be any real number. Real number a,
b is potential energy, side length and "standard length"
This is a parameter for determining which of the deviations from the above is to be emphasized before the re-division. Either of them can be set to 0, and in this case, only one of the deviation of the potential energy from the length of the side and the “reference length” is reflected in the division pass / fail standard. In particular, when a = 0 and D = 0 and a negative value is used for n, it is not necessary to execute a potential calculation process, so that the effect of improving triangulation can be obtained with a simple configuration. In this case, there is no need to execute the triangulation local correction process.

Next, in [the division pass / fail criterion minimization process]
The above-described [initialization processing], [potential calculation processing (1), (2)], [triangulation local correction processing], [node movement processing], and [divided pass / fail reference calculation processing] are repeatedly executed. In this way, a triangulation is created such that the division pass / fail criterion obtained by the “division pass / fail criterion calculation process” is reduced. A known optimization technique can be applied to this. Typical examples include a steepest descent method, a quasi-Newton method, a conjugate direction method, an annealing method, and a genetic algorithm method. In particular, in the steepest descent method or the like, the gradient based on the division pass / fail standard is used in minimization, but the procedure for calculating this gradient is basically equivalent to the procedure for calculating the repulsive force in the second embodiment.

Here, as an application example, the boundary created by the boundary data creating means 13 is applied to “surface description boundary data”, and the boundary data improved in the third embodiment is applied to “improvement target boundary data”. In this case, the fineness of the division is the third
Since it can be changed by the method of the embodiment, in the “triangulation local correction processing” of the present embodiment, even if the correction such that the number of triangle elements, nodes, and sides is increased or decreased is not performed,
The fineness of division can be freely specified.

(Fifth Embodiment) When tracing the outline of a tissue on a tomographic image, if the brightness (brightness) of the image differs between the tomographic images, the shape of the outline to be traced becomes inconsistent between the tomographic images. In addition, there is a problem that unnecessary undulations occur between the slices of the created boundary data. In order to solve this, as shown in FIG. 15, in this embodiment, an automatic area dividing unit 22 for automatically generating boundary data from a three-dimensional image is provided.

The automatic area dividing means 22 automatically discriminates which tissue each voxel of the three-dimensional image corresponds to using a known tissue segmentation technique.
Next, it is determined whether or not one of the combinations of the eight voxels adjacent to each other in a cubic shape is a boundary of the tissue of interest. For example, if attention is paid to the boundary between the myocardium and the blood, it is determined whether the voxel of the myocardium and the voxel of the blood are adjacent to each other. If they are adjacent, triangular elements are assigned on these boundaries in a predetermined pattern. If this is performed on all sets of eight adjacent voxels, the boundary of interest can be divided into triangular elements.

Since the triangulation at this point has fine irregularities of about the voxel size, the boundary smoothing means 23
Further, by executing the inflation processing in the third embodiment alone, fine irregularities in the boundary shape are obtained, that is, the boundary is smoothed. Since the triangulation thus formed is divided by a very large number of triangular elements, it is preferable to improve the division so as to obtain an appropriate number of triangular elements by using the third and fourth embodiments.

The present invention is not limited to the above-described embodiments, but can be implemented with various modifications.

[0134]

(1) According to the present invention, the conductivity outside the boundary is automatically set from the conductivity inside the boundary according to the inclusion relation between the boundaries automatically recognized. There is no need for the troublesome work of inputting the conductivity inside and outside of each boundary one by one while taking into account the complex inclusion relationship of the boundaries, and does not consider the inclusion relationship as an operator. It is only necessary to input only the inside conductivity that is acceptable. For this reason, not only the work load is reduced, but also the situation of erroneous input of the outer conductivity can be solved.

(2) According to the present invention, a practical biological model can be created by using the function of deleting and adding boundary data.
Thereby, useful biomagnetic field analysis can be performed. Further, since a boundary unnecessary for calculation, for example, a boundary where no current source is present can be deleted, the amount of calculation can be reduced accordingly. As a result, it is possible to increase the speed and accuracy of the biomagnetic field analysis.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a biomagnetic field analyzing apparatus according to a first embodiment of the present invention.

FIG. 2 is a block diagram illustrating a configuration of a biological model creation unit in FIG. 1;

FIG. 3 is a view showing an example of an operation screen related to image information in the first embodiment.

FIG. 4 is a view showing an example of an operation screen related to a boundary in the first embodiment.

FIG. 5 is a view showing a processing procedure from a contour trace to creation of three-dimensional boundary data in the first embodiment together with a result thereof;

FIG. 6 is a view showing an example of an operation screen relating to the inclusion relation and the conductivity in the first embodiment.

FIG. 7 is a view showing an example of a solid angle table for obtaining a boundary inclusion relation by the inclusion relation editing means in FIG. 2;

FIG. 8 is a diagram showing a boundary structure diagram corresponding to the solid angle table of FIG. 7;

FIG. 9 is a diagram showing, on a three-dimensional image and a tomographic image, a torso feature point and a torso index point used for positioning the magnetic sensor and the subject in the first embodiment.

FIG. 10 is a diagram showing an example of an operation screen related to a transfer matrix in the first embodiment.

FIG. 11 is a block diagram showing a configuration of a biological model creation unit in the biomagnetic field analyzer according to the third embodiment.

FIG. 12 is a view showing the processing procedure of the boundary data improving means of FIG. 11 together with the results.

FIG. 13 is a conceptual diagram of a boundary data improvement process according to the fourth embodiment.

FIG. 14 is a diagram showing an example of a triangular rearrangement in the triangulation local correction processing according to the fourth embodiment.

FIG. 15 is a block diagram showing a configuration of a biological model creating unit in the biological magnetic field analyzer according to the fifth embodiment.

FIG. 16 is a diagram showing a procedure of a conventional three-dimensional boundary data creation process together with the result.

FIG. 17 is a boundary structure diagram for supplementing a conventional problem.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Biological image imaging device, 2 ... SQUID magnetic sensor, 3 ... SQUID drive circuit, 4 ... Computer system, 5 ... Biological image transfer means, 6 ... Biological model creation means, 7 ... Data collection means, 8 ... Data analysis means, 9: analysis result display means, 10: three-dimensional imaging means, 11: three-dimensional image display means, 12: contour input means, 13: boundary data creation means, 14: boundary data storage means, 15: boundary change means, 16 ... Boundary inclusion relation editing means, 17 ... Boundary inclusion relation storage means, 18 ... Boundary synthesis means, 19 ... Equation creation operation means, 20 ... Transmission matrix storage means.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Yoichi Takada 1385-1, Shimoishigami, Otawara City, Tochigi Prefecture Inside the Toshiba Nasu Plant (72) Inventor Satoshi Aida 1385-1, Shimoishigami, Otawara City, Tochigi Stock Company F-term in Toshiba Nasu factory (reference) 4C027 AA10 BB05 GG00 HH11 HH13 HH16 HH18

Claims (28)

[Claims] 1. A biomagnetic field analyzing apparatus for analyzing a magnetic field generated from a living body, comprising: a plurality of three-dimensional boundary data relating to a plurality of living tissues from a plurality of tomographic images relating to the living body; Boundary inclusion relation editing means for obtaining an inclusion relation between the plurality of living tissues based on the boundary data, and a transfer matrix for deriving a magnetic map from a current distribution based on the inclusion relation and the conductivity inside each boundary. And a data analysis unit for analyzing electric activity of a living body based on the calculated transfer matrix and a measured value of a magnetic field generated from the living body. . 2. A biomagnetic field analyzing apparatus for analyzing a magnetic field generated from a living body, wherein: a boundary data creating means for creating a plurality of three-dimensional boundary data concerning a plurality of living tissues from a plurality of tomographic images concerning the living body; Means for deleting arbitrary boundary data from the plurality of boundary data and adding arbitrary boundary data, and transmission for deriving a magnetic map from a current distribution based on the plurality of boundary data obtained by the creation, deletion, and addition. A transfer matrix calculating means for calculating a matrix; and a data analysis means for analyzing electrical activity of a living body based on the calculated transfer matrix and a measured value of a magnetic field generated from the living body. Analysis device. 3. The method according to claim 2, wherein the plurality of generated boundary data include:
The biomagnetic analysis apparatus according to claim 1, further comprising a unit configured to select boundary data to be applied to the analysis of the electrical activity. 4. The apparatus according to claim 1, wherein said boundary inclusion relation editing means includes: means for calculating a solid angle between said boundaries; and means for determining inclusion relation of said boundary based on said solid angle. 2. The biomagnetic field analyzer according to claim 1. 5. An examination in which the tomographic image data on the living body, the three-dimensional image data on the living body created from the tomographic image data, the boundary data, the inclusion relation data of the boundary, and the transfer matrix data are common. 3. The biomagnetic analyzer according to claim 1, wherein the data is stored in association with a number, and a name of each data is displayed in a list. 6. The biomagnetic analysis apparatus according to claim 1, further comprising: means for creating a three-dimensional image from the plurality of tomographic images; and means for displaying the three-dimensional image. 7. The boundary data creating unit creates an outline of a tissue in a plurality of tomographic images of the living body, and creates the three-dimensional boundary data by connecting the created outlines. The biomagnetic analysis apparatus according to claim 1 or 2, further comprising: means. 8. The method according to claim 1, wherein a plurality of boundary structures and a plurality of transfer matrices reflecting different boundary structures are created in association with a single inspection number. The biomagnetic analysis apparatus according to claim 1. 9. The biomagnetic analysis apparatus according to claim 8, further comprising a unit for changing at least one of the selected boundary data to a non-selected state. 10. The biomagnetic analysis according to claim 1, further comprising outermost boundary determining means for determining an outermost boundary from the plurality of generated boundary data. apparatus. 11. The biomagnetic analysis apparatus according to claim 1, further comprising means for inputting conductivity inside the boundary or information corresponding thereto. 12. The apparatus according to claim 11, further comprising means for setting the conductivity outside the boundary from the input conductivity inside the boundary according to the inclusion relation between the boundaries. Biomagnetic field analyzer. 13. A magnetic sensor having a plurality of channels for detecting a weak magnetic field from the living body, a means for arbitrarily selecting a channel to be used from among the channels of the magnetic sensor, and the tomographic image or a tomographic image created therefrom. 3. The biomagnetic analysis apparatus according to claim 1, further comprising: means for adjusting a position of the use channel with respect to a three-dimensional image. 14. The biomagnetic analysis apparatus according to claim 13, wherein the positioning is performed based on a position of a torso index point of the living body. 15. The magnetic field calculation according to claim 1, wherein a magnetic field is calculated by a calculation formula having a driving term obtained by dividing a potential generated by a current source existing inside the boundary by a conductivity of the boundary. Biomagnetic field analyzer. 16. The biomagnetic analysis apparatus according to claim 1, further comprising: means for improving the boundary data. 17. The boundary data improving means includes an inflating means for enlarging the boundary, a polyhedron projecting means for projecting a polyhedron model on the enlarged boundary, and a return means for reducing the boundary based on the projection of the polyhedron model. 17. The biomagnetic field analyzer according to claim 16, comprising: 18. The system according to claim 1, wherein the inflating means inflates the boundary by applying a repulsive force and a suction force to each node of the boundary to move the node repeatedly.
8. The biomagnetic field analyzer according to 7. 19. The polyhedral model projecting means may include, among intersections of a half line extending from the center of the polyhedron model to each vertex of the polyhedron model and boundary data, an intersection of the polyhedron model closest to the center of the polyhedron model. The biomagnetic analysis apparatus according to claim 17, wherein each node is projected. 20. The boundary data improving means includes means for moving nodes of the boundary data to be improved on a curved surface of the boundary data in which the nodes are described by a curved surface, and means for minimizing a potential. Item 17. A biomagnetic field analyzer according to Item 16. 21. The biomagnetic analysis apparatus according to claim 20, wherein said boundary data improving means has means for calculating said potential. 22. The biomagnetic field analyzer according to claim 20, wherein said boundary data improving means has means for calculating a derivative of said potential. 23. The biomagnetic field analyzing apparatus according to claim 20, wherein said boundary data improving means has means for calculating the potential based on a three-dimensional Euclidean distance. 24. The biomagnetic field analyzer according to claim 20, wherein said boundary data improving means has means for calculating the potential based on a distance along a curved surface. 25. The apparatus according to claim 20, wherein said boundary data improving means has means for specifying a reference length, and means for improving the boundary data with reference to the specified reference length. The biomagnetic analysis apparatus according to claim 1. 26. The biomagnetic analysis apparatus according to claim 20, wherein said boundary data improving means includes a triangulation local correction means for locally correcting said boundary data. 27. The apparatus further comprising means for automatically tracing a contour of a tissue from the plurality of tomographic images, wherein the boundary data creating means creates a boundary by connecting the traced contours, and The biomagnetic field analyzer according to claim 1, wherein the boundary data is created by dividing the boundary data by a triangular element. 28. The biomagnetic analysis apparatus according to claim 27, further comprising a smoothing unit that smoothes a boundary divided by the triangular element.