JPH10276996A - Estimating device for organism action current source - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an estimating device for an organism action current source, the device that can accurately estimate the position of an organism action current source. SOLUTION: This device functions as follows: to set a larger weighting factor to the outline point of a region of interest, which is nearer to a magnetic sensor (S2); to calculate a spheral model using a function, which is weighted by this weighting factor, of the square difference between a distance from the center point of the spheral model to the respective outline point and the radial of the spheral model (S3); to calculate a value of the coincidence evaluating function (the coincident extent), which shows the extent of coincidence of this spheral model and the outline point (S4); and to adjudge whether this value is smaller than a threshold value or not (S5). The spheral model is determined in a case where the coincident extent is smaller than the threshold value. On the other hand, in a case where the coincident extent is larger than the threshold value, processed S2 to S5 are taken repeatedly until the coincident extent value becomes smaller than the threshold value. Thus, a spheral model that is most conformable to the surface shape of the region of interest facing to the magnetic sensor is calculated.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a biological activity current source estimating apparatus for estimating the position, orientation, and size of a biological activity current source, and more particularly, to a biological activity current source for accurately estimating the position of a biological activity current source. Source device.

[0002]

2. Description of the Related Art For example, when a stimulus is applied to a living body, the polarization formed across the cell membrane is broken and a living activity current flows. This biological activity current appears in the brain and heart, and is recorded as an electroencephalogram and an electrocardiogram. The magnetic field generated by the biological activity current is recorded as a magnetoencephalogram or a magnetocardiogram.

[0003] Conventionally, three-dimensional morphological information can be obtained by photographing, for example, the head, which is a region of interest, using an MRI apparatus. For example, only the outline data of the brain is extracted from the morphological information, and a three-dimensional image of the brain is created by the MRI apparatus using the method described in Japanese Patent Application Laid-Open No. 4-226635. The biological activity current source estimating apparatus calculates the contour data of the stereoscopic image of the brain as M
Received from RI device. Based on the brain contour data, a sphere model that fits the entire stereoscopic image is calculated by a function using the least squares method.

Next, a SQUID (Supercon
A magnetic sensor using a ducting Quantum Interference Device (superconducting quantum interferometer) or the like is placed, and a magnetic field generated by a current dipole (hereinafter referred to as a “current source”), which is a biological activity current source generated in the brain, is not detected by the magnetic sensor. Measure invasively.
The position, size, and direction of the current source are estimated from the measurement data by using an estimation method using the minimum norm method shown in Japanese Patent Application Laid-Open No. 7-148132. Note that the position, size,
The direction is estimated based on the coordinate system of the spherical model. In addition, since the spherical model is calculated from the contour data of the brain, the position, size, and direction of the current source can be superimposed on the stereoscopic image of the brain. Thus, the position of the current source in the brain is estimated.

Here, using a function including the least squares method,
Conventional method of calculating a sphere model from contour data of a region of interest
The order will be described. The contour data described above consists of a large number of coordinate points (hereinafter
Below, it is called "contour point")).
The coordinates of the Guo point are P_i= (X_i, y_i, Z_i) (I = 1,2,
…, N). Here, the sphere model you want
Is P = (X, Y, Z) and the radius is r. Next
And each contour point P from the center point P of the sphere model_iThe distance to
R _i(i = 1,2,…, n), R_iIs expressed by the following equation (1).
It is.

R _i = {((X−x _i ) ² + (Y−y _i ) ² + (Z−z _i ) ² )} (1) Distance from center point P to each contour point P _i By calculating the coordinates P = (X, Y, Z) of the center point and the radius r of the sphere model to minimize the square difference between R _i and the radius r of the sphere model to be obtained, the surface shape of the region of interest is obtained. Can be set. That is, the least square difference between R _i and r is obtained. Here, the following equation (2) is used as the function e using the least square method.

[0007]

(Equation 1)

In the equation (2), the conditions for minimizing the function e are shown in the following equations (3) to (6).

[0009]

(Equation 2)

Here, the solution of the simultaneous equations given by the equations (3) to (6) is obtained to obtain the coordinates P of the center point of the spherical model.
= (X, Y, Z) and its radius r are determined, and a sphere model suitable for the surface shape of the site of interest is set. This is,
It is a calculation procedure of a sphere model using a function including a least squares method from contour data.

[0011]

However, the prior art having such a configuration has the following problems. When measuring biomagnetism, a magnetic sensor is installed near a position where a current source is expected to be present. Here, the region of interest facing the magnetic sensor, for example, the surface shape of the brain,
It is set as an ideal spherical surface. However, the human brain is not a sphere but a distorted shape. The sphere model calculated based on the entire stereoscopic image of the brain has a part set outside from the surface of the brain and a part set inside. For example, the surface shape of the brain facing the magnetic sensor often does not match the set spherical surface. When a current source is estimated at the part where the two do not match, if the current source is superimposed on a three-dimensional image of the brain, the current source is shown at a position deviating from the three-dimensional image of the brain. Thus, the sphere model fitted to the entire head could not estimate and indicate the current source at the correct position.

SUMMARY OF THE INVENTION The present invention has been made in view of such circumstances, and provides a life activity current source estimating apparatus capable of obtaining a sphere model suitable for accurately displaying the position of the life activity current source. It is intended to provide.

[0013]

The present invention has the following configuration to achieve the above object. That is, the invention according to claim 1 sets a sphere model that matches the surface shape of the region of interest based on the contour data of the region of interest of the subject, and uses this sphere model to perform life activity in the region of interest. An apparatus for estimating a physical quantity such as a position, a size, and a direction of a current source, comprising: (a) a position of a magnetic sensor for measuring biomagnetism and a number of contour points represented by contour data of the subject; Weighting factor setting means for setting a larger weighting factor for a contour point closer to the magnetic sensor based on the positional relationship with each position of (b), (b) a distance from a center point of the sphere model to the contour point, and A sphere model calculating unit configured to calculate a sphere model that minimizes the value of the function by using a function weighted by the weighting coefficient of the square difference with the radius of the sphere model. .

According to a second aspect of the present invention, there is provided the apparatus according to the first aspect, wherein (c) a sphere model calculated by the sphere model calculation means and a surface of a region of interest of the subject facing the magnetic sensor. A coincidence evaluation function for expressing the degree of coincidence with the shape by a numerical value is determined in advance, and the information of the calculated sphere model and the information of each contour point are given to the coincidence evaluation function, and the coincidence evaluation function is given. (D) when the degree of coincidence calculated by the degree of coincidence calculation is equal to or less than a predetermined threshold value, the sphere model has an appropriate sphere. If the degree of coincidence is larger than a threshold value, the weight coefficient setting means resets a new weight coefficient, and the sphere model calculating means based on the new weight coefficient. Recalculation of the sphere model It is characterized in that the degree of coincidence and a determination means for repeatedly executed until it is determined that more than the threshold value.

[Operation] The operation of the first aspect of the invention is as follows. First, a magnetic sensor for measuring biomagnetism from a site of interest of a subject is positioned so as to face the surface of the site of interest where a current source is expected to be present. Further, since the biomagnetism is extremely weak, most of the current sources that can be actually measured are current sources near the surface of the site of interest. Therefore, the one to be set is good. Therefore, in the present invention, when setting the sphere model, the weighting factor setting means determines the positional relationship between the position of the magnetic sensor and each of a number of contour points represented by the contour data of the subject obtained in advance. , A larger weighting factor is set for a contour point closer to the magnetic sensor. The sphere model calculating means calculates a sphere model that minimizes the square difference between the distance from the center point of the sphere model to each contour point and the radius of the sphere model using a function weighted by the weight coefficient. I do. The weighting factor for weighting the squared difference is set to be larger for contour points closer to the magnetic sensor, so that the sphere model deviates from the surface shape of the site of interest on the side closer to the magnetic sensor,
The value of the function is larger. In other words, if a sphere model that minimizes the function is calculated, a sphere model that matches the surface shape of the region of interest close to the magnetic sensor can be obtained.

The operation of the invention described in claim 2 is as follows. Depending on the set value of the weight coefficient, a sphere model that is consistent with the surface shape of the region of interest facing the magnetic sensor is not always calculated. Therefore, the information of the calculated sphere model and the information of each contour point are given to a coincidence evaluation function for expressing numerically the degree of coincidence between the surface shape of the region of interest facing the magnetic sensor and the sphere model. Calculate the degree of coincidence. When the degree of coincidence is equal to or less than a preset threshold value, it is determined that the sphere model is the sphere model that most matches the surface shape of the region of interest. When it is determined that the degree of coincidence is larger than the threshold value, a new weight coefficient is set again. By repeating the processing by the sphere model calculating means and the determining means based on the weight coefficient,
An optimal sphere model can be calculated for the surface shape of the subject.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A schematic configuration of a living activity current source estimating apparatus according to an embodiment will be described with reference to FIG. In the figure, reference numeral 14 denotes a magnetically shielded room, and a bed 15 on which the subject M lies supine, and a site of interest of the subject M, for example, which is disposed close to the brain and is generated in the magnetically shielded room. And a magnetic sensor 2 for non-invasively measuring a minute magnetic field generated by a living activity current source. In the present embodiment, when the brain of the subject M is a sphere, the magnetic sensor 2 includes a plurality of coils for detecting a magnetic field component in the radial direction.

The magnetic field data detected by the magnetic sensor 2 is provided to a data conversion unit 4 and converted into digital data, and then collected by a data collection unit 5. The stimulating device 6 is for giving an electrical stimulus (or a sound, a light stimulus, or the like) to the subject M. The positioning unit 7 is a device for grasping a positional relationship of the subject M with respect to a three-dimensional coordinate system based on the magnetic sensor 2. For example, small coils are attached to a plurality of locations on the subject M, and power is supplied to these small coils from the positioning unit 7. Then, the magnetic sensor 2 detects the magnetic field generated by each coil, thereby grasping the positional relationship of the subject M with respect to the magnetic sensor 2. In addition,
Other methods for grasping the positional relationship of the subject M with respect to the magnetic sensor 2 include a method of irradiating the subject M with a light beam to grasp the positional relationship between the two, or Japanese Patent Application Laid-Open No. Hei 5-2370065. And various methods disclosed in Japanese Patent Application Laid-Open No. 6-788925.

The sphere model setting unit 8 includes a contour data storage unit 9
Is used to set a sphere model based on the contour data stored in. The contour data storage unit 9 stores, for example, contour data of a stereoscopic image of a brain obtained by an MRI apparatus as three-dimensional coordinate data. In addition, the current source estimating unit 16 provided in association with the sphere model setting unit 8 includes the sphere model set by the sphere model setting unit 8 and the data collection unit 5.
This is for estimating the current source in the region of interest of the subject M based on the magnetic field data collected in the step S1. The current source estimated by the current source estimating unit 16 is superimposed on the stereoscopic image of the brain stored in the contour data storage unit 9 and
, Or printed out on the color printer 12. Note that the contour data of the stereoscopic image of the brain obtained by the MRI apparatus may be directly transmitted to the sphere model setting unit 8 via the communication line 13 shown in FIG.

The sphere model setting process executed by the sphere model setting section 8 will be described below with reference to the flowchart of FIG. 2 and FIG.

First, three-dimensional morphological information can be obtained by photographing the head 3 using an MRI apparatus. Only the outline data of the brain is extracted from this morphological information, and a stereoscopic image 1 is created. The bioelectric current source estimating device stores the three-dimensional image 1 in the contour data storage unit 9. The three-dimensional image 1 is composed of triangular elements, and the coordinates of each vertex of the triangular element (hereinafter, referred to as “contour point”) are represented by contour data P _i = (x _i ,
y _i , z _i ) (i = 1, 2,..., n). The sphere model setting section 8 reads out the contour data P _i (step S
1).

For example, when measuring the occipital region, FIG.
The magnetic sensor 2 outside the back of the head 3
You. Center of the sensor surface of the magnetic sensor 2 with the center of the magnetic sensor 2
The magnetic sensor coordinates S are set as points. Where the magnetic
Sensor center point S and each contour point P _iFunction f
(Α_i). Where the variable α_iIs a relative angle
Can be represented by a relative distance, but in this embodiment, the variable α
_iIs represented by a relative distance. f (α_i) Is the expression (8) described later.
tan^-1Since it is used as a function variable, 0 ≦ f (α_i)
≦ 180, f (α_i) Is the following formula
It can be expressed by (7).

[0023]

(Equation 3)

In equation (7), α _i indicates the distance from the center point S of the magnetic sensor to the contour point P _i . α _max is
Shows the distance to the farthest contour point P _i from the central point S of the magnetic sensor. α _min indicates the distance from the center point S of the magnetic sensor to the nearest contour point P _i . Where f (α _i )
Is as contour point P _i is closer to the center point S of the magnetic sensor, the following weighting coefficients w _i with respect to contour point P _i, is a function of nature, such as larger.

Next, when the tan ^-1 function including the weighting variable f (α _i ) is used as the weighting coefficient w _i , w _i can be expressed by the following equation (8).

[0026]

(Equation 4)

In equation (8), parameters t, θ, s
Set any value to. Here, the explanation of the manner of change of the weighting coefficients w _i when each changing the parameter with reference to Figures 4-6. In FIG.
= 0, θ = 60, and the case where t is changed from 0 to 10 is shown. As is apparent from the graph, w _i to changes in as f (α _i) to increase the value of t is understood that change rapidly. In FIG. 5, t = 5 and s = 0.
When θ is changed from 0 to 180 while fixing θ, it is understood from the graph that θ is a parameter for setting a range in which the weight is to be increased. In FIG. 6, when s is changed from 1 to 2 while fixing t = 5 and θ = 60, as is clear from the graph, if s is increased, the value of f (α _i ) is increased. It can be seen that w _i changes significantly. In consideration of the above, arbitrary values are given to the parameters t, θ, and s in the equation (8), and the weight coefficient w _i is set. (Step S2).

A function E for calculating a sphere model including the above-mentioned w _i will be described. Here, similarly to the conventional example, the coordinates P = (X,
Y, Z) and its radius is r. Next, assuming that the distance from the center point P to each contour point P _i is R _i (i = 1, 2,..., N),
R _i is represented by the following equation. _{R i = √ ((X-} x i) 2 + (Y-y i) 2 + (Z-z
_i ) ² ) By calculating P = (X, Y, Z) and r so as to minimize the square difference between the distance R _i and the radius r of the sphere model to be obtained, the region of interest facing the magnetic sensor ( In this embodiment, a sphere model that matches the surface shape of the occiput can be calculated. That is, the least square difference between R _i and r is obtained. Here, a function E using the least squares method including the weight coefficient w _i is expressed as
The following expression (9) shows.

[0029]

(Equation 5)

In the equation (9), the condition for minimizing the function E is represented by the following equations (10) to (13).

[0031]

(Equation 6)

Here, by solving the simultaneous equations given by the equations (10) to (13), the coordinates P = (X, Y, Z) of the center point of the spherical model and its radius r are obtained. A sphere model that matches the surface shape of the region of interest facing the magnetic sensor is calculated (step S3).

Next, it is checked whether or not the spherical model sufficiently matches the surface shape of the region of interest facing the magnetic sensor 2. Whether the values match is determined by comparing a value obtained from the matching score evaluation function F with a predetermined threshold value. The following equation (14) shows the coincidence evaluation function F.

[0034]

(Equation 7)

In the coincidence evaluation function F, a larger weighting factor w _i is added to the contour point P _i closer to the magnetic sensor 2.
The value of the coincidence evaluation function F greatly changes depending on the difference between R _i and r. However, at the contour point P _i located far from the magnetic sensor 2, since the weight coefficient w _i is small, R
The value of the coincidence evaluation function F does not change much depending on the difference between _i and r. When the sphere model matches the surface shape of the region of interest facing the magnetic sensor 2, the matching score evaluation function F having such a characteristic is the value of the matching score evaluation function F (hereinafter, “matching score”). ) Becomes smaller, and when they do not match, the matching degree becomes larger. Here, by substituting the w _i and R _i and r to calculate the degree of coincidence in coincidence evaluation function F (step S4).

Steps S2 to S5 are repeated until the coincidence is equal to or less than the preset threshold. If the coincidence table function F is equal to or less than the threshold, the spherical model and the magnetic sensor 2 are faced. This sphere model is determined on the assumption that the surface shapes of the interested portions match each other (step S5).

The sphere model calculated by the above-described means is a sphere model most suitable for the surface shape of the subject facing the magnetic sensor 2. In addition, by performing current source estimation based on the spherical model, the current source can be estimated at an accurate position with respect to the stereoscopic image of the brain.

[Simulation] FIGS. 7 and 8 are views of a three-dimensional image 10 of the brain and spherical models 20 and 21 viewed from three directions. Here, (a) is a diagram of the brain viewed from the left side,
(B) is a diagram of the brain viewed from the back, and (c) is a diagram of the brain viewed from the upper side. The three-dimensional image in the area surrounded by the two-dot chain line corresponds to the surface of the brain facing the magnetic sensor 2. In order to confirm the effectiveness of the above-described method, a simulation was performed in which a magnetic sensor was placed on the back of the head. Also, in this simulation, in order to compare the conventional sphere model calculating means with the sphere model calculating means of the present invention, weighting coefficients under three conditions are set in advance without using the coincidence determining means. First,
FIG. 7 shows a sphere model 20 calculated based on the contour data of the triangular element three-dimensional model 10 by using the conventional method and superimposed on the triangular element three-dimensional model 10. Table 1 shows the center coordinates, radius, and coincidence evaluation function values of the sphere model 20. Next, a sphere model 21 weighted based on the contour data of the same triangular element three-dimensional model 10 is calculated. Each parameter of the equation (8) is set for each of the conditions 1 to 3. Condition 1: t = 1, θ = 30, s = 0 Condition 2: t = 10, θ = 30, s = 10 Condition 3: t = 10, θ = 90, s = 10 Table 1 shows the result of calculation of Equation (9). FIG. 8 shows a sphere model 21.

[0039]

[Table 1]

As is apparent from Table 1, the coincidence evaluation function value under the condition 2 is the smallest. 7 and 8
FIG. 8 shows a sphere model for the occipital region shown in FIG.
It can be said that the sphere model of the present invention is more suitable for the occipital region.

[0041]

As is clear from the above description, the apparatus for estimating a biological activity current source according to the first aspect of the present invention assigns a weight to a contour point of a site of interest closer to a magnetic sensor with a larger weighting factor. The sphere model is calculated using a function in which the square difference between the distance from the center point of the model to each contour point and the radius of the sphere model is weighted by a weight coefficient. Therefore, according to the present invention, unlike the case where the weight coefficient is not set, the sphere model is not set to be deviated with respect to the surface shape of the region of interest facing the magnetic sensor, and the interest of facing the magnetic sensor is not set. A sphere model adapted to the surface shape of the site can be obtained. By performing magnetic field measurement and current source estimation based on this sphere model, the position of the current source with respect to the subject can be accurately estimated with respect to the stereoscopic image of the site of interest.

According to the biological activity current source estimating device of the present invention, whether or not the surface shape of the region of interest facing the magnetic sensor and the sphere model match is determined by the value of the matching evaluation function. If they match, the sphere model is fixed.On the other hand, if they do not match, a new weighting factor is set, and the calculation of the sphere model based on the weighting factor is performed until the sphere model matches. repeat. Therefore, according to the present invention, it is possible to calculate a sphere model most suitable for the surface shape of the region of interest facing the magnetic sensor. By performing magnetic field measurement and current source estimation based on this sphere model,
The position of the current source can be more accurately estimated with respect to the stereoscopic image of the site of interest.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram of a biocurrent source estimation device.

FIG. 2 is a flowchart of a sphere model calculating unit of the embodiment.

FIG. 3 is a diagram showing a three-dimensional image of a brain and a magnetic sensor according to the embodiment.

FIG. 4 is a diagram showing a change in a weight coefficient when only a parameter t is changed.

FIG. 5 is a diagram illustrating a change in a weight coefficient when only a parameter θ is changed.

FIG. 6 is a diagram showing a change in a weight coefficient when only a parameter s is changed.

FIG. 7 is a diagram in which a sphere model calculated by a conventional technique is superimposed on a three-dimensional image of the brain and viewed from three directions of (a) a left side surface, (b) a rear surface, and (c) an upper surface of the brain.

FIG. 8 is a diagram in which a sphere model calculated according to the present invention is superimposed on a three-dimensional image of the brain and viewed from three directions of (a) a left side surface, (b) a rear surface, and (c) an upper surface of the brain.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... 3D image of brain 2 ... Magnetic sensor s ... Center coordinate of magnetic sensor surface _Pi ... Contour point 10 ... 3D image of brain 20 ... Spherical model calculated by conventional technology 21 ... Spherical model calculated by this invention

Claims (2)

[Claims] 1. A sphere model that matches the surface shape of the region of interest based on the contour data of the region of interest of the subject is set, and the position and size of the bioactivity current source in the region of interest are set using the sphere model. An apparatus for estimating a physical quantity such as a direction, wherein (a) the position of a magnetic sensor for measuring biomagnetism and the position of each of a number of contour points represented by the contour data of the subject Weighting factor setting means for setting a larger weighting factor for a contour point closer to the magnetic sensor based on the relationship; and (b) calculating a distance from a center point of the sphere model to the contour point and a radius of the sphere model. A biological activity current source estimating apparatus, comprising: a sphere model calculating unit that calculates a sphere model that minimizes the value of the function by using a function obtained by weighting the squared difference by the weight coefficient. 2. The apparatus according to claim 1, wherein (c)
A coincidence evaluation function for expressing numerically the degree of coincidence between the sphere model calculated by the sphere model calculation means and the surface shape of the region of interest of the subject facing the magnetic sensor is determined in advance. A coincidence calculating means for calculating the coincidence, which is a value of a coincidence evaluation function, by giving the information of the calculated sphere model and the information of each contour point to a function, and (d) the coincidence calculating means If the calculated degree of coincidence is equal to or less than a predetermined threshold value, it is determined that the sphere model is an appropriate sphere model, while if the degree of coincidence is greater than the threshold value, the weight is determined. The re-setting of a new weight coefficient by the coefficient setting means and the re-calculation of the sphere model by the sphere model calculating means based on the new weight coefficient are repeatedly executed until the degree of coincidence is determined to be equal to or less than a threshold value. Judgment means Bioelectric current source estimation apparatus, characterized in that was e.