JP2007061359A - Mammographic method using microwave and mammography apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an easy-to-handle and low-cost mammography apparatus with high sensitivity and discrimination ability which is safe and not causing any pain by using microwave but without using X-ray. <P>SOLUTION: The mammography apparatus which visualizes three-dimensionally the inside of tissue structure of the breast T by use of microwave is provided with: an irradiating means 1, 2 for irradiating the breast T with the microwave pulse; a receiving means 1, 2 for receiving scattered waves from the breast T; and an imaging means 3 for imaging the distribution of electrical constants in the breast T. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a method (mammography method) for visualizing a tissue structure inside a breast three-dimensionally using a microwave and a device (mammography device).

  In recent years, breast cancer is the most common cancer that affects women. In such a current situation, breast cancer is a cancer with a high cure rate due to early detection and early treatment, and thus mass screening is recommended.

  For example, in the United States, a combination of visual examination and X-ray mammography has been performed for breast cancer screening from an early stage. In Japan, the Ministry of Health, Labor and Welfare issued a guideline in 2000, stating that women over 50 years old should undergo X-ray mammography once every two years in addition to visual examination. Furthermore, in 2003, a research team of the Ministry of Health, Labor and Welfare presented basic guidelines for “reducing X-ray mammography subjects to those in their 40s”.

  X-ray mammography is currently said to be the most effective diagnostic method (diagnostic method in mass screening) as a method for detecting breast cancer that is clinically invisible.

  However, since the method using X-rays (X-ray mammography) is low in sensitivity to a tumor that does not show very early microcalcification or a breast tissue in which the mammary gland has developed, breast cancer by this method (X-ray mammography) Even after screening, there is a problem that about 15% of breast cancers are missed (false negatives). Moreover, since the X-ray photograph obtained by X-ray mammography is a two-dimensional projection image, an expert who has experience is necessary for interpretation.

  Furthermore, X-rays (ionizing radiation) used in the above-mentioned X-ray mammography have a strong influence on the fetus and are said to contribute to activation of oncogenes and inactivation of tumor suppressor genes. Therefore, X-ray mammography has a problem that it is not desirable to use it for women who are pregnant or have a possibility of pregnancy, or women who have a family history of breast cancer.

  In addition, when a mammography interpretation is not possible or when a more detailed examination is performed, a needle biopsy may be performed, but as a result of this needle biopsy, it is often not a tumor (false positive). Such needle biopsy is necessary for detailed examinations, but it is accompanied by physical pain such as piercing the breast in addition to the mental pain of performing breast cancer screening. It is where you want.

JP 08-186762 A

  As described above, X-ray mammography according to the prior art is (1) a tumor that is not initially microcalcified is less sensitive, (2) it is difficult to find a tumor in the breast with high breast density, and (3) ionization. There are problems such as being unsuitable for women who are pregnant or may be pregnant because of radiation, and (4) skilled in image diagnosis because of the two-dimensional image. Further, in the prior art, there is a problem that (5) a painful needle biopsy has to be performed when it cannot be read by X-ray mammography or when a detailed examination is performed.

  Therefore, the present invention has been made to solve the above-described problems of the prior art, and uses microwaves without using X-rays (that is, eliminating the problems (1) to (4) above). It is an object of the present invention to provide a mammography method and a mammography apparatus that are safer, painless, have high sensitivity and discrimination ability (eliminating the problem of (5) above), and are easy to handle at low cost. .

  The present invention has been made in order to solve the above problems, and is a mammography method for three-dimensional visualization of the tissue structure inside the breast using microwaves, and an irradiation step of irradiating the breast with a microwave pulse; A measurement step of measuring time domain data of the scattered wave from the breast, and an imaging step of imaging the electrical constant distribution in the breast based on the time domain data of the scattered wave. It is said.

  The mammography method according to the present invention includes an initial estimation value step for providing an appropriate initial estimation value for an electrical constant distribution to be estimated in the imaging step, and a numerical calculation for an object having the initial estimation value. Using the forward propagation calculation step for obtaining the received data, the measured reception data obtained in the measurement step, and the calculated reception data obtained in the forward propagation calculation step. A functional calculation step for calculating a value, a back-propagation calculation step for calculating an adjoining electromagnetic field using a difference between the measured received data and the calculated received data, and a gradient of the functional from the adjoining electromagnetic field A gradient calculating step for calculating the estimated value, an estimated value updating step for updating the estimated value of the electric constant distribution using the gradient, a convergence determination for the functional, and the If the result of the iteration number determination regarding the number of updates of the constant value satisfies a predetermined condition, the process proceeds to image the electrical constant distribution, and if the result does not satisfy the predetermined condition, A determination step for performing a determination process so as to advance the process to repeatedly perform the back propagation calculation step, the gradient calculation step, the estimated value update step, the forward propagation calculation step, and the functional calculation step. Preferably there is.

  The mammography method according to the present invention includes an initial estimation value step for providing an appropriate initial estimation value for an electrical constant distribution to be estimated in the imaging step, and a numerical calculation for an object having the initial estimation value. Forward propagation calculation step for obtaining the received data, measured reception data obtained in the measurement step, the calculated reception data obtained in the forward propagation calculation step, and an estimated distribution of electrical constants And a functional calculation step for calculating a functional value incorporating a regularization term, and an inverse for calculating an adjoining electromagnetic field using a difference between the measured reception data and the calculated reception data. A propagation calculating step; a gradient calculating step for calculating a functional gradient from the associated electromagnetic field; and an estimated value updating step for updating the estimated value of the electric constant distribution using the gradient. And the result of the convergence determination regarding the functional and the iteration number determination regarding the update number of the estimated value satisfy a predetermined condition, the process proceeds to image the electric constant distribution, and the result is If the condition given in advance is not satisfied, the process proceeds to repeat the back propagation calculation step, the gradient calculation step, the estimated value update step, the forward propagation calculation step, and the functional calculation step. And a determination step for performing a determination process.

  Further, the mammography method according to the present invention includes an initial estimation value step for giving an appropriate initial estimation value to an electrical constant distribution to be estimated in the imaging step, a filter setting step for setting a filter, and the measurement step. A first filtering step of passing the measured received data obtained in this way through a filter, a forward propagation calculating step of obtaining received data calculated by numerical calculation for the object having the initial estimated value, and the forward propagation calculating step A second filtering step of passing the calculated received data obtained in step through a filter, a measured received data after filtering obtained in the first filtering step, and obtained in the second filtering step. Calculate the functional value using the received data after filtering A function calculation step, a back-propagation calculation step for calculating an adjoining electromagnetic field using a difference between the measured received data after filtering and the calculated received data after filtering; A gradient calculating step for calculating a gradient of the function; an estimated value updating step for updating the estimated value of the electric constant distribution using the gradient; and a convergence determination for the functional and an iteration count determination for the update count of the estimated value. When the result satisfies a predetermined condition, the process proceeds to the step of determining the number of filter stages. When the result does not satisfy the predetermined condition, the back propagation calculation step, the gradient calculation step, the estimated value The update step, the forward propagation calculation step, the second filtering step, and the functional calculation step are repeated. In order to proceed as much as possible, a determination step for performing a determination process, and after the determination step, if the number of stages of the filter satisfies a predetermined condition, the process proceeds to image an electrical constant distribution, When the condition is not satisfied, a filter update step, the first filtering step, the forward propagation calculation step, the second filtering step, the functional calculation step, the determination step, the back propagation calculation step, the gradient It is preferable to have a configuration including a filter stage number determination step for performing a determination process so that the process proceeds to repeatedly perform the calculation step and the estimated value update step.

  Further, the mammography method according to the present invention includes an initial estimated value step for providing an appropriate initial estimated value for an electrical constant distribution to be estimated in the imaging step, a multigrid setting step for setting a multigrid, and the measurement Obtained in the third filtering step of passing the measured reception data obtained in the process through the filter, the fourth filtering step of passing the microwave pulse irradiated in the irradiation step through the filter, and the fourth filtering step A forward propagation calculation step of irradiating an object having the initial estimated value as an incident wave with the filtered microwave pulse and obtaining reception data calculated by numerical calculation; and the forward propagation calculation step. The calculated received data and obtained in the third filtering step. A functional calculation step of calculating a functional value using measured received data after filtering, and using the difference between the measured received data after filtering and the calculated received data, A back propagation calculating step for calculating an associated electromagnetic field, a gradient calculating step for calculating a functional gradient from the associated electromagnetic field, an estimated value updating step for updating an estimated value of the electric constant distribution using the gradient, When the result of the convergence determination regarding the functional and the determination of the number of iterations regarding the update count of the estimated value satisfy a predetermined condition, the process proceeds to a multigrid stage number determination step, and the result does not satisfy the predetermined condition In this case, the back propagation calculation step, the gradient calculation step, the estimated value update step, the forward propagation calculation step, and the general function If the determination step for performing the determination process and the number of stages of the multigrid satisfy a predetermined condition after the determination step so that the process proceeds to repeat the calculation step, the electrical constant distribution is imaged. If the process is not performed and the condition is not satisfied, an extended interpolation process step of grid intervals, a grid update step, the third filtering step, the fourth filtering step, a forward propagation calculation step, and a functional calculation step The multi-grid stage number determination step for performing the determination process is preferably performed so that the determination step, the back propagation calculation step, the gradient calculation step, and the estimated value update step are repeated.

  Furthermore, the present invention has been made to solve the above-described problem, and is a mammography apparatus for three-dimensionally visualizing a tissue structure inside a breast using a microwave, and irradiating means for irradiating a breast with a microwave pulse And receiving means for receiving the scattered wave from the breast, and imaging means for visualizing the electrical constant distribution in the breast based on the time domain data of the scattered wave.

  Further, in the mammography apparatus according to the present invention, the imaging means is configured by using a personal computer, and the initial estimated value calculating means for giving the personal computer an appropriate initial estimated value for the electrical constant distribution to be estimated, Forward propagation calculation means for obtaining reception data calculated by numerical calculation for an object having an initial estimated value, measured reception data obtained by the reception means, and general calculation using the calculated reception data A functional calculation means for calculating a value of the function, a back propagation calculation means for calculating an associated electromagnetic field using a difference between the measured received data and the calculated received data, and a gradient of the functional from the associated electromagnetic field A gradient calculating means for calculating the estimated value, an estimated value updating means for updating the estimated value of the electric constant distribution using the gradient, and a convergence determination and When the result of the iteration number determination regarding the update number of the estimated value satisfies a predetermined condition, the process proceeds to image the electric constant distribution, and the result does not satisfy the predetermined condition. Is configured to function as a determination unit that performs a determination process so that the process proceeds to repeatedly perform the back propagation calculation, the gradient calculation, the estimated value update, the forward propagation calculation, and the functional calculation. preferable.

  Further, in the mammography apparatus according to the present invention, the imaging means is configured by using a personal computer, and the initial estimated value calculating means for giving the personal computer an appropriate initial estimated value for the electrical constant distribution to be estimated, Forward propagation calculating means for obtaining received data calculated by numerical calculation for an object having an initial estimated value, measured received data obtained by the receiving means, the calculated received data, and an electrical constant A functional calculation means for calculating a functional value incorporating a regularization term using the estimated distribution, and an associated electromagnetic field is calculated using a difference between the measured received data and the calculated received data Back propagation calculating means, gradient calculating means for calculating a gradient of a functional from the accompanying electromagnetic field, and an estimated value update for updating the estimated value of the electric constant distribution using the gradient Means, and if the result of the convergence determination regarding the functional and the number of iterations determination regarding the number of updates of the estimated value satisfy a predetermined condition, the process proceeds to image the electrical constant distribution, and the result is If a predetermined condition is not satisfied, a determination process is performed so that the process proceeds to repeatedly perform the back propagation calculation, the gradient calculation, the estimated value update, the forward propagation calculation, and the functional calculation. It is preferable that the configuration function as a determination unit.

  Further, the mammography apparatus according to the present invention is such that the imaging means is configured using a personal computer, and the personal computer gives an initial estimated value calculating means for providing an appropriate initial estimated value to an electrical constant distribution to be estimated, a filter Filter setting means for setting the above, first filtering means for passing the measured reception data obtained by the receiving means through a filter, and receiving data calculated by numerical calculation for the object having the initial estimated value A forward propagation calculating means, a second filtering means for passing the calculated received data through a filter, the measured received data after filtering, and a functional value calculated using the filtered received data after filtering Functional calculation means for performing the filtering, the received data measured after filtering, and the file Back propagation calculation means for calculating an adjoining electromagnetic field using a difference from the calculated reception data after tulling, a gradient calculation means for calculating a gradient of a functional from the adjoining electromagnetic field, and the electric constant using the gradient If the result of the estimated value updating means for updating the estimated value of the distribution, the convergence determination regarding the functional and the iteration count determination regarding the update count of the estimated value satisfies a predetermined condition, the process proceeds to the determination of the number of filter stages, If the result does not satisfy a predetermined condition, processing is performed to repeatedly perform the back propagation calculation, the gradient calculation, the estimated value update, the forward propagation calculation, the second filtering, and the functional calculation. In order to proceed, the determination means for performing the determination process, and after the determination process, if the number of stages of the filter satisfies a predetermined condition, the electrical constant distribution is imaged In the case where the condition is not satisfied, the update process of the filter, the first filtering, the forward propagation calculation, the second filtering, the functional calculation, the determination process, the back propagation calculation, It is preferable to have a configuration that functions as a filter stage number determination unit that performs a determination process so that the process proceeds to repeatedly perform gradient calculation and update of the estimated value.

  In the mammography apparatus according to the present invention, the imaging means is configured using a personal computer, and the personal computer is provided with an initial estimated value calculating means for giving an appropriate initial estimated value to an electrical constant distribution to be estimated, A multi-grid setting means for setting a grid; a third filtering means for passing the measured reception data obtained by the receiving means through a filter; and a fourth filter for passing the microwave pulse irradiated by the irradiation means through the filter. Filtering means, irradiating an object having the initial estimated value as an incident wave with the filtered microwave pulse, forward propagation calculating means for obtaining received data calculated by numerical calculation, the calculated received data, The functional value is calculated using the measured received data after filtering. A functional calculation means, a back-propagation calculation means for calculating an associated electromagnetic field using a difference between the measured received data after filtering and the calculated received data, and a gradient of the functional from the associated electromagnetic field Gradient calculation means for calculating the estimated value, estimated value update means for updating the estimated value of the electrical constant distribution using the gradient, results of convergence determination regarding the functional and iteration count determination regarding the update count of the estimated value are given in advance. If the above condition is satisfied, the process proceeds to the determination of the number of multigrid stages, and if the result does not satisfy a predetermined condition, the back propagation calculation, the gradient calculation, the estimated value update, the forward propagation calculation, And a determination means for performing a determination process so that the process proceeds to repeat the functional calculation, and after the determination process, the number of stages of the multigrid is given in advance. If the condition is satisfied, the process proceeds to image the electrical constant distribution. If the condition is not satisfied, the grid interval extension interpolation process, the grid update, the third filtering, and the fourth filtering are performed. , Function as multi-grid stage number determination means for performing determination processing so as to advance the processing to repeatedly perform the forward propagation calculation, the functional calculation, the determination processing, the back propagation calculation, the gradient calculation, and the estimated value update. A configuration is preferred.

  According to the present invention, by using microwaves without using X-rays, a mammography method and a mammography apparatus that are safer, painless, have high sensitivity and discrimination capability, and are easy to handle at low cost. Can be obtained.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

<First embodiment>
FIG. 1 is a schematic view of a mammography apparatus according to the first embodiment of the present invention. As shown in FIG. 1, the mammography apparatus according to this embodiment irradiates the breast T with the microwave pulse and receives the scattered wave from the breast T, and this array antenna measurement. Functions as an image display unit that displays an image based on control of the pulse wave transmission / reception unit 2 that performs transmission / reception of pulse waves to / from the unit 1, control of the pulse wave transmission / reception unit 2, waveform analysis of the received signal, and the obtained signal And a personal computer (hereinafter referred to as “PC”) 3 or the like.

  The array antenna measurement unit 1 (corresponding to the “irradiation unit” and “reception unit” of the present invention) includes a cylindrical antenna installation unit 11 that can accommodate the breast T and a plurality of antenna installation units 11. The antenna element 12 and a connection line 13 that electrically connects the antenna element 12 and the pulse wave transmitting / receiving unit 2 are configured. The antenna installation part 11 is comprised using insulating materials, such as a tetrafluoroethylene resin glass fiber, a polystyrene glass, a polyimide, glass epoxy, polyester, TMM, for example. The antenna element 12 is configured using, for example, copper, silver, alumina, or the like.

  The pulse wave transmission / reception unit 2 (corresponding to “irradiation means” and “reception means” of the present invention) is controlled by the PC 3 and irradiates a microwave pulse at a predetermined timing from a predetermined antenna element 12. It is configured as follows. Further, the pulse wave transmitting / receiving unit 2 receives a transmission / scattering pulse from the breast T using a predetermined antenna element 12 after irradiating the microwave pulse (or while irradiating the microwave pulse). It is structured to do.

  Next, a measurement method in the mammography apparatus according to the present embodiment will be described using a two-dimensional model. Here, FIG. 2 shows a schematic diagram of a two-dimensional model of the array antenna measurement unit 1 constituting the mammography apparatus according to the present embodiment. In FIG. 2, the inner circle indicates the breast T, and the outer circle indicates the antenna installation unit 11. And the state by which the several antenna element (1st antenna element 12a, 2nd antenna element 12b, ... 8th antenna element 12h) was provided in this antenna installation part 11 is shown. 2A is a schematic diagram showing a case where the first antenna element 12a is a transmission antenna (indicated by “◯”), and FIG. 2B is a diagram showing the second antenna element 12b as a transmission antenna (“ It is the schematic which shows the case where it is set as (circle).

  In the present embodiment, among the plurality of antenna elements 12 constituting the array antenna measurement unit 1, microwave pulse irradiation is performed using one antenna element as a transmission antenna, and transmission / scattering is performed using the remaining antenna elements as reception antennas. It is configured to perform pulse measurement processing.

  Specifically, as shown in FIG. 2A, the first antenna element (# 1) 12a is used as a transmission antenna (indicated by “◯”), and a microwave is applied from the first antenna 12a to the breast T. Pulses are irradiated, and the remaining antenna elements 12b to 12h are used as receiving antennas (indicated by “●”), and transmission / scattering pulses from the breast T are received (measured) using these antenna elements 12b to 12h. . At this time, the microwave pulse irradiated from the first antenna element 12a and the transmitted / scattered pulses measured by the remaining antenna elements 12b to 12h are associated with the transmission / reception data ( Measurement data) is stored in the PC 3.

  Next, as shown in FIG. 2B, the second antenna element (# 2) 12b is used as a transmission antenna (indicated by “◯”), and a microwave pulse is applied from the second antenna 12b to the breast T. Irradiation is performed, and the remaining antenna elements 12c to 12a are used as receiving antennas (indicated by “●”), and transmission / scattering pulses from the breast T are received (measured) using these antenna elements 12c to 12a. Also in this case, the microwave pulse irradiated from the second antenna element 12b and the transmission / scattering pulse measured by the remaining antenna elements 12c to 12a are associated with the transmission / reception data associated with each other via the pulse wave transmission / reception unit 2. It is stored in the PC 3 as (measurement data).

  In the measurement method according to the present embodiment, the same processing as that when the first antenna element 12a and the second antenna element 12b described above are used as transmission antennas is performed for the other antenna elements 12c to 12h. In other words, each antenna element is used as a transmitting antenna, and the remaining antenna elements are used as receiving antennas, and transmission / reception data (measurement data) in which microwave pulses and transmission / scattering pulses in each case are associated is stored in the PC 3. Let

  Next, a method for imaging a breast (image reconstruction algorithm) in the mammography apparatus according to the present embodiment will be described.

  The image reconstruction algorithm is a calculation procedure for visualizing the electrical constant (dielectric constant, conductivity) distribution inside the breast using the received data measured based on the method of FIG. 2 described above. In the microwave region, the contrast of the electrical constant between the normal tissue of the breast and the tumor is very high, so if the electrical constant distribution of the received data obtained as described above is visualized, the tissue structure inside the breast is tertiary. It is possible to visualize the original.

  In the image reconstruction algorithm according to the present embodiment, first, an appropriate distribution is given as the electrical constant distribution in the breast that is the target object. Then, for this distribution, the same procedure as the measurement method shown in FIG. 2 is numerically performed on the computer (PC3) to obtain the received data numerically (the obtained data is hereinafter referred to as “calculated "Received data"). Since the “appropriately distributed electrical constant distribution” is generally different from the “true electrical constant distribution” of an actual breast, “calculated received data” is actually measured data (hereinafter “measured "Received data"). The difference between “calculated received data” and “measured received data” includes information on how “appropriately given electrical constant distribution” differs from “true electrical constant distribution”. Therefore, it is conceivable to use this information to update the “appropriately given electrical constant distribution” to approach the “true electrical constant distribution”, and this updating method is an image reconstruction algorithm.

  The image reconstruction algorithm according to the present embodiment is a method for minimizing a functional defined based on a difference between “calculated received data” and “measured received data”. The functional arguments are a spatially varying dielectric constant distribution function and a conductivity distribution function. The functional is nonlinear, and a specific example is shown below.

  Here, first, in defining the functional, consider the reconstruction problem of the lossless dielectric D placed in the free space as shown in FIG. Assuming the current source J, the standardized Maxwell equation is expressed by the following equation 1 using the operator L.

  Here, the operator L is expressed as the following Expression 2.

  V, J, A, B, C, F, and G in Expression 1 and Expression 2 are expressed as the following Expression 3.

C (= 1 / (√ (ε ₀ μ ₀ ))) is the speed of light in free space, η (= √ (μ ₀ / ε ₀ )) is the intrinsic impedance in free space, and ε _r , μ _{r 1} and σ are the relative permittivity, relative permeability, and conductivity of the dielectric, respectively. The incident pulse is excited at time t = 0, and the electromagnetic field is zero when t ≦ 0. In other words, the forward problem is solved by Maxwell's equation (Equation 1) under the following equation 4, which is an initial condition at t = 0.

  Next, the functional F (p) is defined by the following equation (5).

Here, v _m ′ (r _n ^r , t) is an electromagnetic field that is irradiated (emitted) with an electromagnetic wave from a wave source located at the m-th transmission point r _m ^t and observed (received) at the n-th observation point r _n ^r. in ^{_{and, v m (p; r n}} r, t) is the calculated estimated electromagnetic field using the estimated medium vector function (below 6).

T is an observation time length, K _mn (t) is a non-negative weight function that becomes zero when t = T, and the superscript “t” represents transposition.

  In an ideal case where there is no noise in the measurement data, the functional F (p) has a minimum value of zero if the estimated medium vector function p matches the true medium vector function. Therefore, the inverse scattering problem considered here can be regarded as a minimization problem of the functional F (p).

In this embodiment, the functional F (p) is minimized using a gradient method. The Fréchet derivative F _p ′ δp of the functional F (p) is given by the following equation (7).

  Here, δp is a derivative of the estimated medium vector function p, and is expressed by the following equation (8).

Further, δv _m (p; r _n ^r , t) is a Frechet derivative of v _m (p; r _n ^r , t) at p, and simultaneous differential equations represented by the following equation 9 and This is the value at the observation point r = r _n ^r of the solution δv _m (p; r, t) satisfying the initial condition at t = 0 represented by 10.

  Furthermore, δF and δG in Equation 9 are expressed as the following Equation 11.

The Fréchet differential F _p ′ δp can be represented by the inner product of δp and the gradient vector g (r) according to the Reese's representation theorem.

Here, the inner product of the three-dimensional vector function a = (a ₁ , a ₂ , a ₃ ) ^t and b = (b ₁ , b ₂ , b ₃ ) ^t is expressed by the following equation (13). However, V is a reconstruction area.

  The gradient g (r) is given by the following equations (14), (15), and (16).

Here, the 6-dimensional adjoint vector function W _m (p; r, t) = (W _m1 , W _m2 , W _m3 , W _m4 , W _m5 , W _m6 ) ^t is the equation of Expression 17 and This is the solution when solved below.

However, the adjoint operator L ^* is given by the following equation (19).

Further, δ (r−r _n ^r ) is a delta function, and u _m is the following equation (20).

The adjoint vector function W _m can be numerically calculated by using the FDTD method (finite difference time domain method) in Equation 17 by tracing the time back from t = T.

  In the present embodiment, the functional F is defined based on the above-described idea, and the functional F is minimized by the conjugate gradient method. Specifically, in order to minimize the functional F, various calculations based on the measurement data (finally, imaging of electric constant distribution) are performed.

  The algorithm of this embodiment will be described below based on the flowchart of FIG. FIG. 4 is a flowchart showing an image reconstruction algorithm according to the present embodiment.

  As shown in FIG. 4, first, in this embodiment, “measured received data” stored in a memory or the like in the PC 3 is calculated using an input device (such as a keyboard) such as the PC 3. Part (CPU or the like). That is, an input process related to “measured reception data” is performed (step S401).

  Next, using an input device (such as a keyboard) such as a PC 3, an incident pulse (a microwave pulse irradiated to the breast T) stored in a memory or the like in the PC 3 is converted into an arithmetic processing unit (such as a CPU). Given to. That is, input processing related to the incident pulse is performed (step S402). Specifically, the shape (waveform) of the incident pulse is input.

  Next, an appropriate initial estimated value is given to the electric constant distribution to be estimated (step S403) (corresponding to the “initial setting value step” of the present invention). That is, an appropriate initial estimated value is given to the breast T. Here, as an appropriate initial estimated value, for example, an estimated value given by an extended Born approximation method, a back propagation method, a simple estimation method of an electric constant using a pulse reception time, or the like is used.

  Next, the scattered electromagnetic field in the receiving antenna is calculated by numerical calculation for an object having an estimated electrical constant distribution (here, breast T). That is, on the PC 3, it is assumed that the breast T to which an appropriate initial estimated value is given is irradiated with the microwave pulse (assuming that the irradiation similar to the actual measurement is performed). To obtain the “calculated received data” (step S404) (corresponding to the “forward propagation calculation step” of the present invention).

Next, “calculated received data” (corresponding to “v _m (p; r _n ^r , t)” in equation 5) and “measured received data” (“v _m ′ (r _n ^{r in} equation 5) , T) ") is used to calculate the value of the functional F (step S405) (corresponding to the" functional calculation step "of the present invention).

  Next, a determination as to whether or not the functional F is equal to or less than a predetermined value (convergence determination), and a determination as to whether or not the number of updates of the electrical constant estimation value has been a predetermined number (number of iterations) (Determination) is performed (step S406) (corresponding to the “determination step” of the present invention). If the functional F is equal to or less than a predetermined value, or if the number of updates of the electrical constant estimated value has reached a predetermined number (“Yes” in S406), then The process after step S410 is performed. If the functional F is not less than a predetermined value and the number of updates of the electrical constant estimated value has not reached the predetermined number (in the case of “No” in S406), then the step The process after S407 is performed.

As a result of the determination of the number of convergence and iteration (S406), when the functional F is not less than a predetermined value and the update number of the electrical constant estimated value has not reached the predetermined number ("No" in S406) ”), The associated electromagnetic field is calculated using the difference between“ calculated received data ”and“ measured received data ”(step S407) (“ back propagation calculation step of the present invention ”). ”). That is, the electromagnetic wave having the residual at the receiving point (u _{m in} Expression 17) as an equivalent wave source is propagated back from time t = T to zero.

  After the back propagation calculation in step S406 is performed, the gradient of the functional F is then calculated from the associated electromagnetic field (step S408) (corresponding to the “gradient calculation step” of the present invention). That is, the gradient vector of the functional F in the estimated electrical constant is calculated.

  After the calculation of the gradient in step S408, the estimated value of the electrical constant distribution is then updated using the calculated gradient by an update method determined by the gradient method (for example, the conjugate gradient method) (search). The electric constant is updated along the direction) (step S409) (corresponding to the “estimated value update step” of the present invention), and the processing after step S404 is repeated.

  As a result of the determination of the convergence and the number of iterations (S406), when the functional F is less than or equal to a predetermined value, or when the number of updates of the electrical constant estimated value has reached a predetermined number (in S406) In the case of “Yes”, the obtained estimated electrical constant distribution is imaged (step S410), and a series of algorithms according to the present embodiment is terminated.

  As described above, the mammography apparatus according to the present embodiment is configured to be capable of three-dimensional visualization of the tissue structure inside the breast T by using microwaves instead of X-rays. Therefore, according to the present embodiment, since there is no concern about exposure to ionizing radiation, it is possible to obtain a mammography apparatus that is highly safe and can be received by the subject with peace of mind. In addition, since there is no concern about exposure to ionizing radiation, it is safe for an inspection engineer, and the inspection engineer can operate the apparatus beside the subject during measurement.

  In addition, since the mammography apparatus according to the present embodiment employs the time domain method using microwaves, it has the following effects.

  For example, compared with the frequency domain method, the frequency domain method is only a method using a single frequency (sine wave) or several kinds of frequencies, but the time domain method using the microwave according to the present embodiment is All frequencies within the spectral bandwidth of the pulse used are used. The frequency domain method only irradiates a target object (for example, a breast) with a sine wave and measures a sine wave-like scattered field, but the time domain method according to the present embodiment irradiates a target object with a pulse wave. In order to measure a pulsed scattered field, information on the target object is included in the pulse size, delay time, and waveform distortion of the scattered field. That is, the scattered field data in the time domain method using the microwave according to the present embodiment has more information on the target object than the scattered field data in the frequency domain method.

From this, according to this embodiment, since the number of transmission / reception antennas used for measurement of scattered field data can be configured with a smaller number, a mammography apparatus can be obtained at low cost.
Moreover, the measurement time can be shortened.
Furthermore, since a lot of data can be obtained, the algorithm is further stabilized and a clearer three-dimensional image can be obtained. Therefore, the position, shape, and size of the tumor can be easily read from various directions. In addition, since a three-dimensional image with a clear electrical constant distribution can be obtained in this way, tissue identification (specification of normal tissue or tumor) can be easily performed without being an expert. Furthermore, since a clear three-dimensional image can be obtained, it is possible to avoid psychological and physical disadvantages due to false positives (misdiagnosis) and missed malignant tumors due to false negatives.
Further, by applying the time gate, it is possible to remove clutter from the outside of the measuring device (mammography device).

  That is, according to the present embodiment, since it has the above-described configuration and the like, by using microwaves without using X-rays, it is safer, painless, has high sensitivity and discriminating ability, and is low in cost. A mammography apparatus that is easy to handle can be obtained.

<Second embodiment>
Next, a mammography apparatus according to the second embodiment of the present invention will be described. The basic configuration of the mammography apparatus according to the second embodiment (the configuration shown in FIG. 1 and the measurement method described with reference to FIG. 2) is the same as the mammography apparatus according to the first embodiment described above. Here, the specific configuration is omitted, and description will be made with reference to FIG.

  The difference between the mammography apparatus according to the second embodiment and the mammography apparatus according to the first embodiment is an imaging method (image reconstruction algorithm) in the breast. That is, the image reconstruction algorithm used when the inside of the breast T is visualized using the obtained data (measured received data) is different.

  Hereinafter, description of the same parts (configuration and method) as those of the first embodiment will be omitted, and mainly the parts different from the first embodiment (characteristic parts of the second embodiment) will be specifically described.

  Since the problem of estimating the electrical constant distribution inside the target object (here, the breast T) from the scattering data is an inappropriate inverse scattering problem, it is numerically unstable. In order to avoid this, it is necessary to give a condition that the electric constant distribution is a somewhat smooth change. A method for giving such a condition is called a “regularization method”.

  However, when a normal regularization method is used, a boundary (edge) where the tissue suddenly differs (for example, a boundary between the breast T and the air) is blurred, and thus the sharpness of such an “edge” is preserved. An edge-preserving regularization method is known. Therefore, in the present embodiment, an “edge-preserving regular term” (hereinafter simply referred to as “regularization term”) in the “edge-preserving regularization method” is incorporated to constitute an image reconstruction algorithm.

The functional F _TOTAL (p) incorporating the regularization term according to the present embodiment is given by the following equation (21).

Here, F _RES (p) is a functional defined based on the difference between “calculated received data” and “measured received data”, and F _REG (p) is a regularization term. It is.

The gradient of the functional F _TOTAL (p) incorporating the regularization term is given by the sum of the gradient of F _RES (p) and the gradient of F _REG (p). As the regularization parameter included in the regularization term F _REG (p) increases or decreases, the amount of contribution to the functional F _TOTAL (p) increases or decreases. In this embodiment, when the functional F _TOTAL (p) is successively minimized by the gradient method, the regularization term F _REG (p) is approximately the same as the residual functional F _RES (p) in the initial stage. The regularization parameter is gradually reduced as the iterative process proceeds. At the convergence stage, arithmetic processing is performed to eliminate the contribution of the regularization term F _REG (p).

  The algorithm of this embodiment will be described below based on the flowchart of FIG. FIG. 5 is a flowchart showing an image reconstruction algorithm according to the present embodiment.

  As shown in FIG. 5, in the second embodiment, as in the first embodiment described above, the input device (keyboard or the like) such as the PC 3 is used and stored in the memory or the like in the PC 3. “Measured received data” is given to an arithmetic processing unit (CPU or the like). That is, an input process related to “measured reception data” is performed (step S501).

  Next, using an input device (such as a keyboard) such as a PC 3, an incident pulse (a microwave pulse irradiated to the breast T) stored in a memory or the like in the PC 3 is converted into an arithmetic processing unit (such as a CPU). Given to. That is, input processing related to the incident pulse is performed (step S502). Specifically, the shape (waveform) of the incident pulse is input.

  Next, an appropriate initial estimated value is given to the electrical constant distribution to be estimated (step S503) (corresponding to the “initial estimated value step” of the present invention). That is, an appropriate initial estimated value is given to the breast T. Here, as an appropriate initial estimated value, for example, an estimated value given by an extended Born approximation method, a back propagation method, a simple estimation method of an electric constant using a pulse reception time, or the like is used.

  Next, the scattered electromagnetic field in the receiving antenna is calculated by numerical calculation for an object having an estimated electrical constant distribution (here, breast T). That is, on the PC 3, it is assumed that the breast T to which an appropriate initial estimated value is given is irradiated with the microwave pulse (assuming that the irradiation similar to the actual measurement is performed). To obtain the “calculated received data” (step S504) (corresponding to the “forward propagation calculation step” of the present invention).

  Next, using the “calculated received data”, “measured received data”, and the estimated distribution of electrical constants, the value of the functional (see Equation 21) incorporating the regularization term is calculated (step 21). S505) (corresponding to the “functional calculation step” of the present invention).

Next, a determination as to whether or not the functional F _TOTAL (p) represented by _Equation 21 is equal to or less than a predetermined value (convergence determination), and the number of times the number of updates of the electrical constant estimated value is given in advance. Is determined (repetition count determination) (step S506) (corresponding to the “determination step” of the present invention). If the functional is less than or equal to a predetermined value, or if the electrical constant estimated value has been updated a predetermined number of times (“Yes” in S506), then the step The process after S510 is performed. If the functional is not less than or equal to a predetermined value and the number of updates of the electrical constant estimated value has not reached the predetermined number (“No” in S506), then step S507 is performed. Subsequent processing is performed.

  As a result of the determination of the number of convergence and iteration (S506), when the functional is not less than a predetermined value and the number of updates of the electrical constant estimated value has not reached the predetermined number ("No" in S506) In the case of (1), an adjoining electromagnetic field is calculated using a difference between “calculated received data” and “measured received data” (step S507) (“back propagation calculation step” of the present invention). Equivalent).

  After the back propagation calculation in step S507 is performed, the functional gradient is then calculated from the associated electromagnetic field (step S508) (corresponding to the “gradient calculation step” of the present invention). That is, the functional gradient vector in the estimated electrical constant is calculated.

  After the calculation of the gradient in step S508, the estimated value of the electrical constant distribution is then updated using the calculated gradient by an updating method determined by a gradient method (for example, conjugate gradient method) (search). The electric constant is updated along the direction) (step S509) (corresponding to the “estimated value update step” of the present invention), and the processing after step S504 is repeated.

  As a result of the determination of the number of convergence and iteration (S506), when the functional is less than or equal to a predetermined value, or when the number of updates of the electrical constant estimated value reaches a predetermined number (in S506, “ In the case of “Yes”), the obtained estimated electrical constant distribution is imaged (step S510), and a series of algorithms according to the present embodiment is terminated.

  As described above, the mammography apparatus according to the present embodiment is configured to be capable of three-dimensional visualization of the tissue structure inside the breast T by using microwaves instead of X-rays. Therefore, according to this embodiment, all the various effects explained in the first embodiment can be obtained.

  In addition, according to the second embodiment, a functional calculation including a regularization term is performed in step S505. In addition, regular additions also contribute in steps S507 and S508. Therefore, according to the present embodiment, a stable calculation result can be obtained as compared with the first embodiment.

<Third embodiment>
Next, a mammography apparatus according to the third embodiment of the present invention will be described. The basic configuration of the mammography apparatus according to the third embodiment (the configuration shown in FIG. 1 and the measurement method described with reference to FIG. 2) is the same as that of the mammography apparatus according to the first embodiment described above. Here, the specific configuration is omitted, and description will be made with reference to FIG.

  The difference between the mammography device according to the third embodiment and the mammography devices according to the first and second embodiments is an imaging method (image reconstruction algorithm) in the breast. That is, the image reconstruction algorithm used when the inside of the breast T is visualized using the obtained data (measured received data) is different.

  More specifically, the mammography apparatus according to the present embodiment differs from the second embodiment in the “regularization” method. Hereinafter, the description of the same parts (configuration and method) as those of the second embodiment will be omitted, and mainly the parts different from the second embodiment (characteristic parts of the third embodiment) will be specifically described.

  In the mammography apparatus according to the third embodiment, “regularization” using a frequency filter is performed. Specifically, several low-pass filters with different cutoff frequencies are used.

  First, the high frequency part is cut through the time waveform of “measured received data” through the filter having the lowest cutoff frequency. Using this filtered received data not only improves the signal-to-noise ratio (signal-to-noise ratio), but also allows the image to be reconstructed more stably. The electrical constant distribution obtained at this time is a smooth image of spatial changes.

  When an image obtained as described above is used as an initial estimated value and image reconstruction is performed using received data that has passed through a filter having a higher cutoff frequency, the obtained image becomes clearer. The mammography apparatus according to this embodiment can obtain an image with high resolution by repeating this process using a frequency filter.

  Hereinafter, the algorithm of the present embodiment will be described based on the flowchart of FIG. FIG. 6 is a flowchart showing an image reconstruction algorithm according to the present embodiment.

  As shown in FIG. 6, in the third embodiment, similarly to the first and second embodiments described above, the input device (keyboard or the like) such as the PC 3 is used to store in the memory or the like in the PC 3. The “measured received data” is given to an arithmetic processing unit (CPU or the like). That is, an input process related to “measured received data” is performed (step S601).

  Next, using an input device (such as a keyboard) such as a PC 3, an incident pulse (a microwave pulse irradiated to the breast T) stored in a memory or the like in the PC 3 is converted into an arithmetic processing unit (such as a CPU). Given to. That is, input processing related to the incident pulse is performed (step S602). Specifically, the shape (waveform) of the incident pulse is input.

  Next, an appropriate initial estimated value is given to the electrical constant distribution to be estimated (step S603) (corresponding to the “initial estimated value step” of the present invention). That is, an appropriate initial estimated value is given to the breast T. Here, as an appropriate initial estimated value, for example, an estimated value given by an extended Born approximation method, a back propagation method, a simple estimation method of an electric constant using a pulse reception time, or the like is used.

Next, the filter is set (step S604) (corresponding to the “filter setting step” of the present invention). Specifically, the number of filter stages N and the cutoff frequency f _n (n = 1, 2,..., N) are set. The cut-off frequency f _n increases with n.

Next, the measured received data is filtered (step S605) (corresponding to the “first filtering step” of the present invention). Specifically, the measured reception data is passed through a filter having a cutoff frequency f _n (n = 1 at this time).

  Next, the scattered electromagnetic field in the receiving antenna is calculated by numerical calculation for an object having an estimated electrical constant distribution (here, breast T). That is, on the PC 3, it is assumed that the breast T to which an appropriate initial estimated value is given is irradiated with the microwave pulse (assuming that the irradiation similar to the actual measurement is performed). To obtain the “calculated received data” (step S606) (corresponding to the “forward propagation calculation step” of the present invention).

Next, filtering of the calculated reception data is performed (step S607) (corresponding to the “second filtering step” of the present invention). More specifically, the filter cut-off frequency f _n, calculated received data is passed through.

  Next, functional calculation is performed (step S608) (corresponding to “functional calculation step” of the present invention). Specifically, the functional value is calculated using the filtered “calculated received data” and the filtered “measured received data”.

  Next, a determination as to whether or not the functional is less than or equal to a predetermined value (convergence determination), and a determination as to whether or not the number of updates of the electrical constant estimated value has reached a predetermined number (number of iterations) (Determination) is performed (step S609) (corresponding to the “determination step” of the present invention). If the functional is less than or equal to a pre-given value (this value may vary from filter to filter), or the number of times the electrical constant estimated value has been updated in advance (this number may vary from filter to filter) ) Has been reached (in the case of “Yes” in S609), the processing from step S613 onward is then performed. If the functional is not less than or equal to a predetermined value and the number of updates of the electrical constant estimated value has not reached the predetermined number (“No” in S609), then step S610 is performed. Subsequent processing is performed.

  As a result of the determination process of convergence and iteration number (S609), when the functional is not less than or equal to a predetermined value and the update number of the electrical constant estimated value has not reached the predetermined number ("No" in S609) In the case of "", back propagation calculation is performed (step S610) (corresponding to the "back propagation calculation step" of the present invention). Specifically, the associated electromagnetic field is calculated using the difference between the filtered “calculated received data” and the filtered “measured received data”.

  After the back propagation calculation in step S610 is performed, the functional gradient is then calculated from the associated electromagnetic field (step S611) (corresponding to the “gradient calculation step” of the present invention). That is, the functional gradient vector in the estimated electrical constant is calculated.

  After the calculation of the gradient in step S611, the estimated value of the electric constant distribution is then updated using the calculated gradient by an updating method determined by a gradient method (for example, the conjugate gradient method) (search). The electric constant is updated along the direction) (step S612) (corresponding to the “estimated value update step” of the present invention), and the processing after step S606 is repeated.

  As a result of the determination process (determination step) (S609) of the number of convergence and iteration, when the functional is less than or equal to a predetermined value, or the number of updates of the electrical constant estimated value has reached the predetermined number. In the case (“Yes” in S609), the filter stage number determination process is then performed (step S613) (corresponding to the “filter stage number determination step” of the present invention). Specifically, it is determined whether or not the number of filter stages is the maximum (the cutoff frequency is the same as the maximum frequency of the signal pulse). If the number of filter stages is the maximum (in the case of “Yes” in S613), then the processing from step S614 is performed. If the number of filter stages is not maximized (“No” in S613), the filter is updated to a filter that is larger by 1 than the current filter number n (the filter having the next highest cutoff frequency) ( Step S615) (corresponding to the “filter update step” of the present invention), and further, the processing after step S605 is repeated.

  As a result of the filter stage number determination process (S613), when the number of filter stages is maximized ("Yes" in S613), the obtained estimated electrical constant distribution is imaged (step S614). A series of algorithms according to the embodiment ends.

  As described above, the mammography apparatus according to the present embodiment is configured to be capable of three-dimensional visualization of the tissue structure inside the breast T by using microwaves instead of X-rays. Therefore, according to this embodiment, all the various effects explained in the first embodiment can be obtained.

  In the third embodiment, as described above, first, the measurement electromagnetic field is passed through a low-pass filter having a cut-off frequency sufficiently lower than the maximum frequency of the signal component, and only the low frequency component is used. Make the configuration. Then, the result of reconfiguration using only the low frequency component is set as the initial value, the reconfiguration is performed through a filter with a high cutoff frequency, and this process is repeated until the filter with the highest cutoff frequency. Yes. That is, according to the third embodiment, the image reconstruction is performed after the filtering process (S605, S607, etc.) is repeatedly performed during the calculation. Therefore, according to the third embodiment, it is possible to perform image reconstruction more stably as compared with the first embodiment, and it is possible to obtain an “image” with clearer and higher resolution. it can.

<Fourth embodiment>
Next, a mammography apparatus according to the fourth embodiment of the present invention will be described. The basic configuration of the mammography apparatus according to the fourth embodiment (the configuration shown in FIG. 1 and the measurement method described with reference to FIG. 2) is the same as that of the mammography apparatus according to the third embodiment described above. Here, the specific configuration is omitted, and description will be made with reference to FIG.

  The difference between the mammography apparatus according to the fourth embodiment and the mammography apparatus according to the third embodiment is an imaging method (image reconstruction algorithm) in the breast. That is, the image reconstruction algorithm used when the inside of the breast T is visualized using the obtained data (measured received data) is different.

  More specifically, the mammography apparatus according to the present embodiment is different from the third embodiment in that an image reconstruction algorithm using “multigrid” is employed. Hereinafter, the description of the same parts (configuration and method) as those of the third embodiment will be omitted, and mainly the parts different from the third embodiment (characteristic parts of the fourth embodiment) will be specifically described.

  The inverse scattering problem is generally ill-conditioned, and image reconstruction becomes unstable when noise is present or the number of unknown parameters increases. Further, when the reconstruction area becomes large, not only the reconstruction becomes unstable, but also the analysis takes a long time. In this embodiment, in order to improve these problems, an image reconstruction algorithm using “multigrid” is employed. Hereinafter, an example in which the number of grid stages is two will be described with reference to FIGS. 7 and 8 using a two-dimensional model of image reconstruction.

The algorithm according to the present embodiment does not analyze the reconstruction area D divided by Δu as shown in FIG. 7A, and changes the division width to Δu ′ = 2Δu as shown in FIG. It is configured to perform analysis with a "rough grid". In this case a coarse (received data measured) observation field necessary for reconstruction of the lattice _{"(v m '(r n r} , t))''is the observation field relative to the incident pulse" v _m' (r _n ^{r 2} , t) ”can be obtained by passing through LPF 1 (low-pass filter 1) having a cutoff frequency f _max / 2. Here, f _max is the highest frequency of the incident pulse. Further, an incident electromagnetic wave (calculated reception data) “(v _m (p; r _n ^r ,) in a coarse grating is obtained by using a pulse having only a low frequency component obtained by passing the incident pulse through the same LPF 1 as an incident wave. t)) '"can be obtained.

In the present embodiment, these new observed electromagnetic fields “(v _m ′ (r _n ^r , t)) ′” and estimated electromagnetic fields “(v _m (p; r _n ^r , t)) ′” By using this, the functional F (p) in Equation 5 described above is minimized, and the reconstruction process is performed.

Next, the result of the coarse grid having the division width Δu ′ is used as the initial value of the fine grid having the division width Δu, and a reconstruction process is performed on the fine grid. At this time, the observation field _{"v m '(r n r,} t) " and is passed through the LPF2 cutoff frequency f _max, and removing the high noise of the high frequency component than the maximum frequency of the incident pulse.

  Further, in the present embodiment, the following extended interpolation of Expression 22 is used in order to convert the reconstruction result of the coarse lattice having the division width Δu ′ into the division width Δu. FIG. 8 shows a diagram of extended interpolation.

  The algorithm of this embodiment will be described below based on the flowchart of FIG. FIG. 9 is a flowchart showing an image reconstruction algorithm according to the present embodiment.

  As shown in FIG. 9, in the fourth embodiment, similarly to the third embodiment described above, it is stored in a memory or the like in the PC 3 using an input device (such as a keyboard) such as the PC 3. The “measured received data” is given to an arithmetic processing unit (CPU or the like). That is, an input process related to “measured reception data” is performed (step S901).

  Next, using an input device (such as a keyboard) such as a PC 3, an incident pulse (a microwave pulse irradiated to the breast T) stored in a memory or the like in the PC 3 is converted into an arithmetic processing unit (such as a CPU). Given to. That is, input processing related to the incident pulse is performed (step S902). Specifically, the shape (waveform) of the incident pulse is input.

  Next, an appropriate initial estimated value is given to the electrical constant distribution to be estimated (step S903) (corresponding to the “initial estimated value step” of the present invention). That is, an appropriate initial estimated value is given to the breast T. Here, as an appropriate initial estimated value, for example, an estimated value given by an extended Born approximation method, a back propagation method, a simple estimation method of an electric constant using a pulse reception time, or the like is used.

Next, multigrid setting is performed (step S904) (corresponding to the “multigrid setting step” of the present invention). Specifically, the number N of multigrid stages, the grid interval Δu _n (n = 1, 2,..., N), the cutoff frequency f _{n of the} filter corresponding to the grid interval (n = 1, 2,..., N), etc. Is set.

Next, the measured received data is filtered (step S905) (corresponding to the “third filtering step” of the present invention). Specifically, the measured reception data is passed through a filter having a cutoff frequency f _n (n = 1 at this time).

Next, filtering of incident pulses (irradiated microwave pulses) is performed (step S906) (corresponding to the “fourth filtering step” of the present invention). Specifically, the incident pulse is passed through a filter having a cutoff frequency f _n (n = 1 at this time).

Next, forward propagation calculation is performed (step S907) (corresponding to the “forward propagation calculation step” of the present invention). Specifically, a filtered incident pulse is used as an incident wave to irradiate an object having an estimated electrical constant distribution (here, breast T), and a numerical value on the grid interval Δu _n (n = 1 at this time). The scattered electromagnetic field (transmitted / scattered pulse) at the receiving antenna is calculated to obtain “calculated received data” (S907).

  Next, functional calculation is performed (step S908) (corresponding to “functional calculation step” of the present invention). Specifically, the value of the functional is calculated using “calculated received data” in step S907 and “measured received data” filtered in step S905.

  Next, a determination as to whether or not the functional is less than or equal to a predetermined value (convergence determination), and a determination as to whether or not the number of updates of the electrical constant estimated value has reached a predetermined number (number of iterations) (Determination) is performed (step S909) (corresponding to “determination step” of the present invention). When the functional is less than a predetermined value (this value may be different for each grid), or the number of times the electrical constant estimated value is updated in advance (this number may be different for each grid) ) Has been reached (in the case of “Yes” in S909), the processing from step S913 is then performed. If the functional is not less than or equal to a predetermined value and the number of updates of the electrical constant estimated value has not reached the predetermined number (“No” in S909), then step S910 is performed. Subsequent processing is performed.

  As a result of the determination of the number of convergence and iteration (determination step) (S909), the functional is not less than a predetermined value and the number of updates of the electrical constant estimation value has not reached the predetermined number (in S909) In the case of “No”, back propagation is calculated (step S910). Specifically, the adjoining electromagnetic field is calculated using the difference between the “calculated received data” and the filtered “measured received data” (S910) (“back propagation calculation step of the present invention”). ”).

  After the back propagation calculation in step S910, the functional gradient is calculated from the associated electromagnetic field (step S911) (corresponding to the “gradient calculation step” of the present invention). That is, the functional gradient vector in the estimated electrical constant is calculated.

  After the calculation of the gradient in step S911, the estimated value of the electric constant distribution is then updated using the calculated gradient by an updating method determined by a gradient method (for example, the conjugate gradient method) (search). The electric constant is updated along the direction) (step S912) (corresponding to the “estimated value update step” of the present invention), and the processing after step S907 is repeated.

  As a result of the determination of the number of convergence and iteration (determination step) (S909), when the functional is less than or equal to a predetermined value, or when the number of updates of the electrical constant estimated value reaches a predetermined number Next, in the case of “Yes” in S909, determination processing of the number of multigrid stages is performed (step S913) (corresponding to “multigrid stage number determination step” of the present invention). Specifically, it is determined whether or not the grid interval is smaller than the minimum value set in step S904. Then, when the grid interval is smaller than the minimum value (in the case of “Yes” in S913), the processing subsequent to step S916 is performed. Further, when the grid interval is not smaller than the minimum value (in the case of “No” in S913), the processing subsequent to step S914 is performed.

  As a result of the multi-grid stage number determination process (S913), if the grid interval is not smaller than the minimum value ("No" in S913), an extended interpolation process is performed (step S914) (in the present invention). Equivalent to “grid interval extended interpolation processing step”). Specifically, the grid interval is halved, and the distribution of the electrical constant for the fine grid is obtained from the estimated electrical constant by the current grid by extended interpolation (see FIG. 8).

After the extension interpolation process (S914) is performed, a grid update process is performed (step S915) (corresponding to the “grid update step” of the present invention), and thereafter, the processes after step S905 are repeated. . In this step S915, updates the value of the cutoff frequency _{f n} is carried out.

  As a result of the multi-grid stage number determination process (S913), when the grid interval is equal to the minimum value (in the case of “Yes” in S913), the obtained estimated electrical constant distribution is imaged (step S916). The series of algorithms according to this embodiment is completed.

  As described above, the mammography apparatus according to the present embodiment is configured to be capable of three-dimensional visualization of the tissue structure inside the breast T by using microwaves instead of X-rays. Therefore, according to the present embodiment, all the various effects described in the other embodiments described above can be obtained.

  Further, in the mammography apparatus (image reconstruction algorithm) according to the fourth embodiment, it is possible to improve the reconstruction image and speed up the analysis by combining the multigrid with the filtering process. Specifically, the reconstructed image can be clarified, and the analysis time can be greatly shortened.

  Furthermore, in the mammography apparatus (image reconstruction algorithm) according to the fourth embodiment, it is not always necessary to halve the grid interval when updating the grid. When the grid interval is other than 1/2, the mammography apparatus (image reconstruction algorithm) according to the fourth embodiment can be executed as it is by applying extended interpolation according to the update of the grid interval and the cutoff frequency. .

  The present invention is not limited to the above-described embodiments, and can be implemented with various modifications as necessary within the scope that can meet the spirit of the present invention. Is also included in the technical scope of the present invention.

  In the above-described embodiment, as shown in FIG. 1, the array antenna measurement unit 1 has been described using the cylindrical antenna installation unit 11 that can accommodate the breast T. It is not limited to this configuration. Therefore, for example, various types of array antenna measurement units 1A, 1B, and 1C as shown in FIG. 10 may be employed as necessary. In FIG. 10, connection lines connected to each antenna element are not shown.

  FIG. 10A shows a parallel panel type array antenna measurement unit 1A configured using two plate-like antenna installation units 11A1 and 11A2. When this array antenna measurement unit 1A is used, microwave pulse irradiation processing and transmission / scattering pulse measurement processing by the antenna element 12 with the breast T sandwiched between the two antenna installation portions 11A1 and 11A2 Is done. As for the subsequent imaging method (image reconstruction algorithm) inside the breast, the method according to any one of the above-described embodiments is adopted as necessary.

  FIG. 10B shows a single-panel array antenna measurement unit 1B configured using a single plate-like antenna installation unit 11B. When the array antenna measurement unit 1B is used, the antenna element 12 performs microwave pulse irradiation processing and reflection / scattering pulse measurement processing with the antenna installation unit 11B pressed against the breast T. As for the subsequent imaging method (image reconstruction algorithm) inside the breast, the method according to any one of the above-described embodiments is adopted as necessary.

  FIG. 10C shows a close contact formed by using a cup-shaped first antenna installation portion 11C1 that can be in close contact with the breast and a plate-like second antenna installation portion 11C2 that can be in close contact with the breast continuous to the breast. This shows an array antenna measurement section 1C of a mold. When this array / antenna measurement unit 1C is used, the irradiation process of the microwave pulse and the measurement process of the transmission / scattering pulse are performed by the antenna element 12 with the breast wrapped in the cup-shaped first antenna installation unit 11C1. Is called. As for the subsequent imaging method (image reconstruction algorithm) inside the breast, the method according to any one of the above-described embodiments is adopted as necessary. In such a case, the cup-shaped first antenna installation portion 11C1 may be configured to be replaceable in accordance with the size and shape of the subject's breast. If it is the array antenna measuring section 1C on the close contact side shown in FIG. 10C, it can be attached to the subject like a normal brassiere. Since the array antenna measurement unit 1C on the close contact side is in close contact with the breast, a better reconstructed image can be obtained as compared with other methods.

  Further, in the above-described embodiment, the configuration for performing imaging inside the breast using the reception data obtained using the array antenna measurement unit 1 has been described, but the present invention is not limited to this configuration, If necessary, the positional relationship between each antenna element and the breast may be measured, and the data inside the breast may be visualized by adding these data. More specifically, for example, when the antenna installation part is configured using a deformable sheet or the like (for example, a microstrip antenna using a conductive cloth), the shape of the breast or the antenna installation Since the positional relationship between each antenna element changes depending on the mounting state of the part, the positional relationship between each antenna element and the breast is photographed by an imaging means such as a CCD camera, and the information obtained here (the shape of the breast) The position of the antenna element in the breast, etc.) is also added to the image reconstruction algorithm. With such a configuration, it is possible to reconstruct the electrical constant distribution image inside the breast in a state in which the three-dimensional information of the breast outline is incorporated, so that the image quality is improved and the calculation time is shortened. be able to.

  Further, as another embodiment, a configuration as shown in FIG. 11 may be adopted. FIG. 11 shows a schematic diagram of an array antenna measurement unit 1D provided with a matching liquid container.

  In order to efficiently cause a microwave pulse to enter the breast T from the antenna element 12 provided in the antenna installation part 11, it is preferable to cover the antenna installation part 11 in which the antenna element 12 is arranged with the matching liquid 16. . FIG. 11 discloses such a configuration. That is, FIG. 11 shows a configuration in which the matching liquid container 15 is provided around the antenna installation portion 11 so that the cylindrical antenna installation portion 11 shown in FIG. In the configuration shown in FIG. 11, the antenna installation unit 11 and the breast T are placed in a matching liquid container 15 filled with the matching liquid 16. At this time, in order to prevent the breast T from directly touching the matching liquid 16, the breast T is covered with a thin film, or the shape of the breast T is formed in a portion where the breast T of the antenna installation unit 11 is inserted. A deformable sheet may be provided in advance. Further, when the matching liquid 16 is used as shown in FIG. 11, it is necessary to adjust the amount of the matching liquid 16 in the matching liquid container 15 depending on the size of the breast T and the like. After the insertion, the matching liquid 16 is injected into the matching liquid container 15.

  According to the configuration shown in FIG. 11, the microwave pulse can be efficiently incident on the breast T, and the antenna element can be downsized as compared with the case where no matching liquid is used. In addition, for the configuration using the matching liquid 16 shown in FIG. 11, it is possible to use not only the cylindrical antenna installation portion 11 but also the close contact type antenna installation portion 11C shown in FIG. 10C, for example. is there.

  Further, as the matching liquid 16, for example, pure water, physiological saline, alcohol, or the like can be used. Further, as the matching liquid 16, for example, a mixed solution in which pure water, salt, sugar, glycerin, alcohol, or the like is mixed can be used.

  Further, the matching liquid container 15 may be configured so as to be entirely covered with a radio wave absorber sheet. With such a configuration, an effect of attenuating unnecessary radio waves from the outside can be obtained.

  Furthermore, in the present invention, as another embodiment, an array antenna measurement unit 1E having a configuration as shown in FIG. 12 may be used. In FIG. 12, connection lines connected to each antenna element are not shown.

  Here, FIG. 12 (a) shows a cross-sectional view of an array antenna measurement unit 1E according to another embodiment, and FIG. 12 (b) shows an enlarged view of a part X in FIG. 12 (a). Yes. As shown in these figures, the array antenna measurement unit 1E according to the present embodiment is formed in a cup shape as a whole, and a sectional view of the center portion of the array antenna measurement unit 1E formed in this cup shape is shown. FIG. 12 (a). The array antenna measurement unit 1E is covered with a metal layer 19 on the outside and covered with two high dielectric constant dielectric layers (first dielectric layer 17 and second dielectric layer 18) on the inside. Yes. Between the first dielectric layer 17 and the second dielectric layer 18, an antenna element 12E composed of a plurality of metal strips regularly arranged is provided. This antenna element 12E has a function similar to that of the other embodiments described so far.

  In the array antenna measurement unit 1E as shown in FIG. 12, since the antenna element unit 12E is provided between the two dielectric layers 17 and 18, the efficiency of pulse incidence into the breast is increased, and the antenna The element 12E can be downsized. Further, it is possible to avoid the antenna element from directly touching the breast. That is, in the present embodiment, the two dielectric layers 17 and 18 have the same function as the matching liquid described above with reference to FIG. In the present embodiment, an array antenna measurement unit 1E having different radii of curvature may be prepared according to the size of the breast.

  Further, in the above-described embodiment, as illustrated in FIG. 2, the case where each antenna element 12 functions as a transmission antenna or a reception antenna in order has been described, but the present invention is not limited to this configuration. Therefore, for example, it may be configured such that not only all antenna elements but only some of the preselected antenna elements have a function as a transmission antenna. Further, for example, not all antenna elements other than the antenna element used as the transmitting antenna may be used as the receiving antenna, but some appropriately selected antenna elements may be used as the receiving antenna. Furthermore, instead of using a single antenna element as a transmitting antenna, it may be configured such that several antenna elements are used as transmitting antennas and at the same time a pulse wave is emitted to irradiate the breast. Further, if necessary, the transmitting antenna and the receiving antenna may be set by appropriately combining the various methods described above.

1 is a schematic view of a mammography apparatus according to a first embodiment of the present invention. FIG. 2 is a schematic diagram of a two-dimensional model of an array antenna measurement unit constituting the mammography apparatus according to the present embodiment. It is the schematic which showed the lossless dielectric pillar etc. used for description at the time of defining the functional concerning this embodiment. It is the flowchart which showed the image reconstruction algorithm concerning 1st embodiment of this invention. It is the flowchart which showed the image reconstruction algorithm concerning 2nd embodiment of this invention. It is the flowchart which showed the image reconstruction algorithm concerning 3rd embodiment of this invention. It is a figure for demonstrating the "fine grating | lattice" and the "rough grating | lattice" of a multigrid in the image reconstruction algorithm concerning 4th embodiment of this invention. It is a figure for demonstrating the "extension interpolation" of the multigrid in the image reconstruction algorithm concerning 4th embodiment of this invention. It is the flowchart which showed the image reconstruction algorithm concerning 4th embodiment of this invention. The schematic of the array antenna measurement part concerning other embodiment of this invention is shown. The schematic of the array antenna measurement part (array antenna measurement part provided with the matching liquid container) concerning other embodiment of this invention is shown. FIG. 5 is a schematic view of an array antenna measurement unit (array antenna measurement unit having two dielectric layers) according to another embodiment of the present invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1, 1A, 1B, 1C, 1D, 1E ... Array antenna measurement part 2 ... Pulse wave transmission / reception part 3 ... PC (personal computer)
11, 11 A 1, 11 A 2, 11 B, 11 C 1, 11 C 2 ... Antenna installation part 12, 12 a to 12 h, 12 E ... Antenna element 13 ... Connection line 15 ... Matching liquid container 16 ... Matching liquid 17 ... First dielectric layer 18 ... Second dielectric Body layer 19 ... Metal layer

Claims (10)

A mammography method that visualizes the tissue structure inside the breast using a microwave in three dimensions,
An irradiation step of irradiating the breast with a microwave pulse;
A measuring step of measuring time domain data of scattered waves from the breast;
A mammography method comprising: an imaging step of imaging an electrical constant distribution in the breast based on time domain data of the scattered wave. In the imaging process,
An initial estimate step that provides an appropriate initial estimate for the electrical constant distribution to be estimated;
Forward propagation calculation step of obtaining received data calculated by numerical calculation for the object having the initial estimated value;
A functional calculation step of calculating a functional value using the measured reception data obtained in the measurement step and the calculated reception data obtained in the forward propagation calculation step;
A back-propagation calculation step of calculating an adjoining electromagnetic field using a difference between the measured reception data and the calculated reception data;
A gradient calculating step of calculating a gradient of a functional from the associated electromagnetic field;
An estimated value update step of updating the estimated value of the electrical constant distribution using the gradient;
When the result of the convergence determination regarding the functional and the determination of the number of iterations regarding the number of updates of the estimated value satisfies a predetermined condition, the process proceeds to image the electric constant distribution, and the result is given in advance. If the above condition is not satisfied, the determination process is performed so that the back propagation calculation step, the gradient calculation step, the estimated value update step, the forward propagation calculation step, and the functional calculation step are repeated. The mammography method according to claim 1, further comprising: In the imaging process,
An initial estimate step that provides an appropriate initial estimate for the electrical constant distribution to be estimated;
Forward propagation calculation step of obtaining received data calculated by numerical calculation for the object having the initial estimated value;
Using the measured reception data obtained in the measurement step, the calculated reception data obtained in the forward propagation calculation step, and the estimated distribution of electrical constants, A functional calculation step for calculating the value of the function;
A back-propagation calculation step of calculating an adjoining electromagnetic field using a difference between the measured reception data and the calculated reception data;
A gradient calculating step of calculating a gradient of a functional from the associated electromagnetic field;
An estimated value update step of updating the estimated value of the electrical constant distribution using the gradient;
When the result of the convergence determination regarding the functional and the determination of the number of iterations regarding the number of updates of the estimated value satisfies a predetermined condition, the process proceeds to image the electric constant distribution, and the result is given in advance. If the above condition is not satisfied, the determination process is performed so that the back propagation calculation step, the gradient calculation step, the estimated value update step, the forward propagation calculation step, and the functional calculation step are repeated. The mammography method according to claim 1, further comprising: In the imaging process,
An initial estimate step that provides an appropriate initial estimate for the electrical constant distribution to be estimated;
A filter setting step for setting a filter;
A first filtering step of passing the measured reception data obtained in the measurement step through a filter;
Forward propagation calculation step of obtaining received data calculated by numerical calculation for the object having the initial estimated value;
A second filtering step of passing the calculated received data obtained in the forward propagation calculation step through a filter;
A functional value is calculated using the measured received data after filtering obtained in the first filtering step and the calculated received data after filtering obtained in the second filtering step. A function calculation step;
Back propagation calculation step of calculating an associated electromagnetic field using a difference between the measured received data after filtering and the calculated received data after filtering;
A gradient calculating step of calculating a gradient of a functional from the associated electromagnetic field;
An estimated value update step of updating the estimated value of the electrical constant distribution using the gradient;
If the result of the convergence determination regarding the functional and the iteration number determination regarding the number of updates of the estimated value satisfy a predetermined condition, the process proceeds to the filter stage determination step, and the result does not satisfy the predetermined condition In such a case, the determination is made so as to proceed to repeat the back propagation calculation step, the gradient calculation step, the estimated value update step, the forward propagation calculation step, the second filtering step, and the functional calculation step. A determination step for processing;
After the determination step, when the number of stages of the filter satisfies a predetermined condition, the process proceeds to image the electric constant distribution. When the condition is not satisfied, the filter update step, The process proceeds to repeatedly perform the first filtering step, the forward propagation calculation step, the second filtering step, the functional calculation step, the determination step, the back propagation calculation step, the gradient calculation step, and the estimated value update step. The mammography method according to claim 1, further comprising: a filter stage number determination step for performing a determination process. In the imaging process,
An initial estimate step that provides an appropriate initial estimate for the electrical constant distribution to be estimated;
Multi-grid setting step for setting multi-grid,
A third filtering step of passing the measured reception data obtained in the measurement step through a filter;
A fourth filtering step of passing the microwave pulse irradiated in the irradiation step through a filter;
A forward propagation calculation step of irradiating the object having the initial estimated value as an incident wave with the filtered microwave pulse obtained in the fourth filtering step, and obtaining received data calculated by numerical calculation;
Functional calculation for calculating a functional value using the calculated reception data obtained in the forward propagation calculation step and the measured reception data after filtering obtained in the third filtering step. Steps,
Back propagation calculation step of calculating an associated electromagnetic field using a difference between the measured received data after filtering and the calculated received data;
A gradient calculating step of calculating a gradient of a functional from the associated electromagnetic field;
An estimated value update step of updating the estimated value of the electrical constant distribution using the gradient;
When the result of the convergence determination regarding the functional and the determination of the number of iterations regarding the update count of the estimated value satisfy a predetermined condition, the process proceeds to a multigrid stage number determination step, and the result does not satisfy the predetermined condition In this case, a determination step for performing a determination process so as to proceed to repeat the back propagation calculation step, the gradient calculation step, the estimated value update step, the forward propagation calculation step, and the functional calculation step; ,
After the determination step, if the number of stages of the multigrid satisfies a predetermined condition, the process proceeds to image the electric constant distribution. If the condition is not satisfied, extended interpolation of the grid interval is performed. Processing step, grid update step, third filtering step, fourth filtering step, forward propagation calculation step, functional calculation step, determination step, back propagation calculation step, gradient calculation step, and estimated value update The mammography method according to claim 1, further comprising: a multi-grid stage number determination step for performing a determination process so as to advance the process to repeat the steps. A mammography device for three-dimensional visualization of tissue structure inside the breast using microwaves,
An irradiation means for irradiating the breast with a microwave pulse;
Receiving means for receiving scattered waves from the breast;
A mammography apparatus comprising: an imaging unit configured to image an electrical constant distribution in the breast based on time domain data of the scattered wave. The imaging means is configured using a personal computer,
The personal computer;
Means for calculating an initial estimated value that gives an appropriate initial estimated value for the electrical constant distribution to be estimated;
Forward propagation calculating means for obtaining received data calculated by numerical calculation for the object having the initial estimated value;
Functional calculation means for calculating a functional value using the measured reception data obtained by the reception means and the calculated reception data;
Back propagation calculating means for calculating an associated electromagnetic field using a difference between the measured received data and the calculated received data;
A gradient calculating means for calculating a gradient of a functional from the associated electromagnetic field;
Estimated value updating means for updating the estimated value of the electrical constant distribution using the gradient;
When the result of the convergence determination regarding the functional and the iteration number determination regarding the update number of the estimated value satisfies a predetermined condition, the process proceeds to image the electric constant distribution, and the result is given in advance. A determination unit that performs a determination process so as to advance the process to repeatedly perform the back propagation calculation, the gradient calculation, the estimated value update, the forward propagation calculation, and the functional calculation when the predetermined condition is not satisfied The mammography apparatus according to claim 6, which functions as a computer. The imaging means is configured using a personal computer,
The personal computer;
Means for calculating an initial estimated value that gives an appropriate initial estimated value for the electrical constant distribution to be estimated;
Forward propagation calculating means for obtaining received data calculated by numerical calculation for the object having the initial estimated value;
Functional calculation means for calculating a functional value incorporating a regularization term using the measured reception data obtained by the reception means, the calculated reception data, and an estimated distribution of electrical constants ,
Back propagation calculating means for calculating an associated electromagnetic field using a difference between the measured received data and the calculated received data;
A gradient calculating means for calculating a gradient of a functional from the associated electromagnetic field;
Estimated value updating means for updating the estimated value of the electrical constant distribution using the gradient;
When the result of the convergence determination regarding the functional and the iteration number determination regarding the update number of the estimated value satisfies a predetermined condition, the process proceeds to image the electric constant distribution, and the result is given in advance. A determination unit that performs a determination process so as to advance the process to repeatedly perform the back propagation calculation, the gradient calculation, the estimated value update, the forward propagation calculation, and the functional calculation when the predetermined condition is not satisfied The mammography apparatus according to claim 6, which functions as a computer. The imaging means is configured using a personal computer,
The personal computer;
Means for calculating an initial estimated value that gives an appropriate initial estimated value for the electrical constant distribution to be estimated;
Filter setting means for setting the filter,
First filtering means for passing the measured received data obtained by the receiving means through a filter;
Forward propagation calculating means for obtaining received data calculated by numerical calculation for the object having the initial estimated value;
Second filtering means for passing the calculated received data through a filter;
Functional calculation means for calculating a functional value using the measured received data after filtering and the calculated received data after filtering;
Back propagation calculation means for calculating an associated electromagnetic field using a difference between the measured received data after filtering and the calculated received data after filtering;
A gradient calculating means for calculating a gradient of a functional from the associated electromagnetic field;
Estimated value updating means for updating the estimated value of the electrical constant distribution using the gradient;
When the result of the convergence determination regarding the functional and the iteration number determination regarding the number of updates of the estimated value satisfy a predetermined condition, the process proceeds to determination of the number of filter stages, and the result does not satisfy the predetermined condition A determination means for performing a determination process so as to advance the process to repeatedly perform the back propagation calculation, the gradient calculation, the estimated value update, the forward propagation calculation, the second filtering, and the functional calculation,
And after the determination process, if the number of stages of the filter satisfies a predetermined condition, the process proceeds to image the electric constant distribution, and if the condition is not satisfied, the filter update process, The determination process is performed so that the first filtering, the forward propagation calculation, the second filtering, the functional calculation, the determination process, the back propagation calculation, the gradient calculation, and the estimated value update are repeated. The mammography apparatus according to claim 6, wherein the mammography apparatus is made to function as means for determining the number of filter stages to be performed. The imaging means is configured using a personal computer,
The personal computer;
Means for calculating an initial estimated value that gives an appropriate initial estimated value for the electrical constant distribution to be estimated;
Multigrid setting means for setting multigrid;
Third filtering means for passing the measured received data obtained by the receiving means through a filter;
A fourth filtering means for passing the microwave pulse irradiated by the irradiation means through a filter;
A forward propagation calculating means for irradiating an object having the initial estimated value as an incident wave with the filtered microwave pulse, and obtaining received data calculated by numerical calculation;
Functional calculation means for calculating a functional value using the calculated received data and the measured received data after filtering;
Back propagation calculating means for calculating an associated electromagnetic field using a difference between the measured received data after filtering and the calculated received data;
A gradient calculating means for calculating a gradient of a functional from the associated electromagnetic field;
Estimated value updating means for updating the estimated value of the electrical constant distribution using the gradient;
When the result of the convergence determination regarding the functional and the iteration number determination regarding the number of updates of the estimated value satisfy a predetermined condition, the process proceeds to determination of the number of multigrid stages, and the result does not satisfy the predetermined condition In this case, a determination unit that performs a determination process so as to proceed to repeat the back propagation calculation, the gradient calculation, the estimated value update, the forward propagation calculation, and the functional calculation;
After the determination process, if the number of stages of the multigrid satisfies a predetermined condition, the process proceeds to image an electrical constant distribution. If the condition is not satisfied, the grid interval is extended. Processing to repeatedly perform interpolation processing, grid update, the third filtering, the fourth filtering, the forward propagation calculation, the functional calculation, the determination processing, the back propagation calculation, the gradient calculation, and the estimated value update. The mammography apparatus according to claim 6, wherein the mammography apparatus functions as multi-grid stage number determination means for performing determination processing so as to proceed.

Images