Diagnostic device, diagnostic method, determination device, determination method, and imaging method

Abstract

This determination device comprises: a sensor that has a plurality of electrodes; a current / voltage application and measurement unit that applies a current or voltage between the electrodes and measures inter-electrode impedance; a relaxation time distribution function calculation unit that calculates an inter-electrode relaxation time distribution function from the inter-electrode impedance; and a determination unit that, on the basis of feature information about the inter-electrode relaxation time distribution function, determines the presence or absence of an abnormal part in which a lesion is possibly present.

Description

Diagnostic device, diagnostic method, assessment device, assessment method, and imaging method

[0001] This application claims priority to Japanese Patent Application No. 2024-71767, filed April 25, 2024, the contents of which are incorporated herein by reference.

[0002] Breast cancer is the uncontrollable growth of breast cells, most of which originate in the milk ducts and some in the lobules. There are several forms of breast cancer, and the type of breast cancer depends on which breast cells develop into cancer. Ductal carcinoma is the most common type of breast cancer, which begins in the milk ducts and spreads to surrounding breast tissue, lymph nodes, and other organs, accounting for approximately 80% of all breast cancer cases.

[0003] Ductal carcinoma is typically detected using mammography, a low-dose X-ray imaging diagnostic technique, and radiologists analyze these X-ray images to detect abnormalities or signs of cancer. However, mammography is not a simple test, making early detection difficult in the case of advanced breast cancer. Furthermore, because mammography uses X-rays, it is not minimally invasive and places a significant burden on patients. Therefore, a non-invasive and simple diagnostic method is needed.

[0004] As a non-invasive method for detecting the presence or absence and location of breast cancer, Patent Document 1 proposes a method using electrical impedance tomography (EIT).

[0005] Japan Special Table No. 2022-506804

[0006] However, the method of Patent Document 1 has the problem that it is difficult to accurately diagnose cancer tissue such as breast cancer.

[0007] The present invention has been made in consideration of the above circumstances, and an object of the present invention is to provide a diagnostic device, a diagnostic method, a determination device, a determination method, and an imaging method that are capable of diagnosing cancerous tissue more precisely.

[0008] In order to solve the above problems, the present invention proposes the following means. (1) A diagnostic device according to Aspect 1 of the present invention comprises: a sensor having a plurality of electrodes; a current / voltage application / measurement unit that applies a current or voltage between the electrodes and measures impedance; and a relaxation time distribution function prediction unit that predicts a relaxation time distribution function from the impedance using a regression model. (2) Aspect 2 of the present invention relates to the diagnostic device of Aspect 1, wherein the relaxation time distribution function prediction unit may predict a relaxation time distribution image from the impedance using the regression model. (3) Aspect 3 of the present invention relates to the diagnostic device of Aspect 1 or 2, wherein the sensor comprises: a support having a dome-shaped internal space; and a plurality of electrodes arranged inside the support, wherein the electrodes constitute a plurality of annular electrode groups arranged in a circle around the central axis of the internal space, and the annular electrode groups may be arranged at predetermined intervals in the height direction of the internal space. (4) A diagnostic method according to a fourth aspect of the present invention includes an impedance measurement step of applying a current or a voltage between the electrodes, and measuring impedance based on a current application / voltage measurement pattern when applying the current, or measuring the impedance based on a voltage application / current measurement pattern when applying the voltage, and a relaxation time distribution function prediction step of predicting a relaxation time distribution function from the impedance using a regression model. (5) An imaging method according to a fifth aspect of the present invention includes an impedance measurement step of measuring the impedance of a measurement object, and a relaxation time distribution prediction step of obtaining a relaxation time distribution image from the impedance using a regression model. (6) A determination device according to a sixth aspect of the present invention includes a sensor having a plurality of electrodes, a current / voltage application measurement unit that applies a current or voltage between the electrodes and measures the inter-electrode impedance, a relaxation time distribution function calculation unit that calculates the inter-electrode relaxation time distribution function from the inter-electrode impedance, and a determination unit that determines the presence or absence of an abnormal region that may be a lesion based on characteristic information of the inter-electrode relaxation time distribution function. (7) A seventh aspect of the present invention is the determination device according to the sixth aspect, wherein the determination unit includes an abnormality detection unit that determines whether or not a peak value of the inter-electrode relaxation time distribution function exceeds a reference peak value.(8) Aspect 8 of the present invention is the determination device of Aspect 6, further comprising a feature information extraction unit that extracts peak values ​​of the inter-electrode relaxation time distribution function obtained from the relaxation time distribution function calculation unit, converts the peak values ​​into peak values ​​for each pixel, and supplies the extracted feature information to the determination unit. (9) Aspect 9 of the present invention is the determination device of Aspect 6 or 8, further comprising: a support having a dome-shaped internal space; and a plurality of electrodes arranged inside the support, wherein the electrodes constitute a plurality of annular electrode groups arranged in a circle around the central axis of the internal space, and the annular electrode groups are arranged at predetermined intervals in the height direction of the internal space. (10) A determination method of aspect 10 of the present invention includes an impedance measurement step of applying a current or a voltage between the electrodes, and measuring the inter-electrode impedance based on a current-application-voltage measurement pattern when applying the current, or measuring the inter-electrode impedance based on a voltage-application-current measurement pattern when applying the voltage, a relaxation time distribution function calculation step of calculating an inter-electrode relaxation time distribution function using the inter-electrode impedance, a determination step of determining the presence or absence of an abnormal site that may be a lesion based on characteristic information of the inter-electrode relaxation time distribution function, and a visualization display step of visually displaying the determination result. (11) An imaging method of aspect 11 of the present invention includes an impedance measurement step of measuring the inter-electrode impedance of a measurement target, a relaxation time distribution function calculation step of determining the inter-electrode relaxation time distribution function from the inter-electrode impedance, and a visualization display step of converting the characteristic information of the inter-electrode relaxation time distribution function into image information and displaying it. (12) Aspect 12 of the present invention is the imaging method of aspect 11, wherein the visualization display step converts graph or numerical information of the inter-electrode relaxation time distribution function as characteristic information into image information. (13) A thirteenth aspect of the present invention is the imaging method of the eleventh aspect, further comprising a feature information extraction step of extracting a peak value of the inter-electrode relaxation time distribution function obtained by the relaxation time distribution function calculation step, converting the peak value into a pixel-by-pixel peak value, and supplying the resulting feature information to the visualization display step.

[0009] According to the present invention, it is possible to provide a diagnostic device, a diagnostic method, a determination device, a determination method, and an imaging method that are capable of diagnosing cancer tissue more precisely.

[0010] FIG. 1 is a schematic diagram of a diagnostic device according to a first embodiment. FIG. 2 is a schematic diagram of the current / voltage application measurement unit of FIG. 1 and the target (A) breast of a living body, and (B) an example of an electrode structure. FIG. 3 is a flowchart of a diagnostic method according to a first embodiment. FIG. 4 is a schematic diagram of the current / voltage application measurement unit of FIG. 3. FIG. 5 is a diagram for explaining a measurement pattern based on an adjacent method. FIG. 6 is a flowchart of an imaging method according to a second embodiment. FIG. 7 is a schematic diagram of an evaluation device according to Example 1. 11(a) shows the results of γ for the first peak in the vicinity of the ductal cancer tissue (IDC), the vicinity of normal breast gland tissue (nGBT), and the vicinity of the adipose tissue (adipose) of a breast sample. FIG. 11(b) shows the results of γ for the second peak in the vicinity of the ductal cancer tissue (IDC), the vicinity of normal breast gland tissue (nGBT), and the vicinity of the adipose tissue (adipose) of a breast sample. FIG. 11(a) shows the results of γ for the second peak in the vicinity of the ductal cancer tissue (IDC), the vicinity of normal breast gland tissue (nGBT), and the vicinity of the adipose tissue (adipose) of a breast sample. FIG. 11(b) shows the results of γ for the second peak in the breast sample. 21A and 21B are relaxation time distribution images and mammography images of the vicinity of ductal cancer tissue (IDC) in Sa2. FIG. 22A are relaxation time distribution images and mammography images of the vicinity of ductal cancer tissue (IDC) in each breast sample Sa3. FIG. 23B are relaxation time distribution images and mammography images of the vicinity of ductal cancer tissue (IDC) in each breast sample Sa4. FIG. 23C are relaxation time distribution images and mammography images of the vicinity of ductal cancer tissue (IDC) in each breast sample Sa5. FIG. 23D are relaxation time distribution images and mammography images of the vicinity of ductal cancer tissue (IDC) in each breast sample Sa6. FIG. 24A is a functional block diagram of a determination device according to a third embodiment. FIG. 24B is a functional block diagram of a determination device according to a fourth embodiment. FIG. 24C is a functional block diagram of the current / voltage application measurement unit of FIG. 21A and the breast of a living body, which is the diagnostic target site, and (B) an example of an electrode structure. FIG. 24D is a flowchart of a determination method according to a fourth embodiment. FIG. 24E is a functional block diagram of a determination device according to a fifth embodiment. FIG. 24F is a functional block diagram of a determination device according to a fifth embodiment. FIG. 24F is a functional block diagram of a current / voltage application measurement unit of FIG. 21A and the diagnostic target site, (A) a living body breast, and (B) an example of an electrode structure. FIG. 24G is a flowchart of a determination method according to a fourth embodiment. FIG. 24G is a functional block diagram of a determination device according to a fifth embodiment. FIG. 24H is a functional block diagram of a current / voltage application measurement unit of FIG. 21A and the diagnostic target site, (A) a living body breast, and (B) an example of an electrode structure. FIG. 24H is a flowchart of a determination method according to a fourth embodiment.Fig. 25 is a schematic diagram of the current and voltage application and measurement unit in Fig. 24. Fig. 26 is a flowchart of an imaging method according to a fifth embodiment.

[0011] (First Embodiment) (Diagnostic Apparatus) A diagnostic apparatus 100 according to one embodiment of the present invention will now be described with reference to the drawings. As shown in Fig. 1, the diagnostic apparatus 100 includes a current / voltage application / measurement unit 1 and a diagnostic calculation unit 50. The diagnostic calculation unit 50 includes a relaxation time distribution function prediction unit 2, a diagnosis unit 5, and an output unit 6.

[0012] The diagnostic calculation unit 50 of the diagnostic device 100 includes, for example, a central processing unit (CPU), a read-only memory (ROM), a random access memory (RAM), and a hard disk drive (HDD) / solid state drive (SSD). The relaxation time distribution function prediction unit 2, the diagnostic unit 5, and the output unit 6 are realized by the CPU executing a predetermined program. The diagnostic calculation unit 50 may also control the current / voltage application measurement unit 1. The program may be acquired via a recording medium or via a network. A dedicated hardware configuration may also be used to realize the configuration of the diagnostic device 100. Each unit will be described below.

[0013] (Current-Voltage Application Measurement Unit) The current-voltage application measurement unit 1 applies a current or voltage between the electrodes 20 and measures the impedance of the measurement object. The measurement object may be, for example, biological tissue, specifically, a living breast or a resected breast. The current-voltage application measurement unit 1 will be described using FIG. 2. FIG. 2(A) is a schematic diagram of the current-voltage application measurement unit 1 in contact with the subject, and FIG. 2(B) is a plan view illustrating an example of an electrode. The drawings used in the following description may show characteristic portions enlarged for ease of understanding, and the dimensional ratios of each component may differ from the actual. The materials, dimensions, etc. exemplified in the following description are merely examples, and the present invention is not limited thereto. Appropriate modifications can be made within the scope of the present invention.

[0014] As shown in FIG. 2A, the current / voltage application / measurement unit 1 includes a sensor 10 and a control unit 30. The sensor 10 may be one or more, and the shape of the sensor 10 is not particularly limited and may be rectangular, circular, or the like. The sensor 10 may be provided with a support and placed on the breast surface. The shape of the support for the sensor 10 is not particularly limited and may be flat or cover the breast surface. The sensor 10 includes electrodes 20, and the placement positions and spacing of the electrodes 20 are not particularly limited. For example, when using the four-terminal method for the electrode structure, as shown in FIG. 2B, a sensor may include four electrodes: a low-current electrode (LC), a low-voltage electrode (LP), a high-current electrode (HC), and a high-voltage electrode (HP). For example, the four peripheral electrodes may serve as the LC electrode and the LP electrode, and the central electrode may serve as the HC electrode and the HP electrode. The electrode 20 may be placed on the object to be measured by making the support adhesive and attaching the support to the object to be measured, or by wrapping the support around the object to be measured, or by having the subject wear a support in the shape of a bra to place the electrode 20 on the object to be measured (measurement location).

[0015] The electrodes 20 are electrically connected to the control unit 30. There are no particular limitations on the material or shape of the electrodes 20 as long as they can apply a current or voltage to the breast as a living body of the subject or to a resected breast. Examples of the electrodes 20 include metals such as Au, Ag, and Cu, conductive polymers, fibers whose surfaces are coated with metals, and fibers whose surfaces are coated with conductive polymers.

[0016] There are no particular limitations on the method of electrically connecting the electrodes 20 and the control unit 30. As a connection method, for example, the electrodes 20 and the control unit 30 may be connected by a coaxial cable or a lead wire, or the electrodes 20 and the control unit 30 may be connected by wiring with conductive fibers woven into it.

[0017] (Control Unit) The control unit 30 includes, for example, a multiplexer for switching between current application electrodes (or voltage application electrodes) that apply a current and voltage measurement electrodes that measure a potential difference (or current measurement electrodes that measure a current), and an impedance analyzer that performs potential difference (or current measurement) and phase measurement. The impedance analyzer is a component that measures impedance, i.e., the ratio of a measured potential difference (applied voltage) to an applied current (measured current), and its phase, by changing the applied frequency and amplitude. The control unit 30 may perform impedance measurement (measurement of the ratio of a potential difference to a current, and its phase) by, for example, executing a predetermined program in a CPU and controlling the multiplexer and impedance analyzer. The control unit 30 may perform impedance measurement by controlling the control unit 30 only within the current / voltage application measurement unit 1, or may perform impedance measurement by controlling the control unit 30 according to a program executed in the diagnostic calculation unit 50. The results of the impedance measurement are sent to the relaxation time distribution function prediction unit 2. The method of transmitting information to the relaxation time distribution function prediction unit 2 is not particularly limited. The signal may be sent from the control unit 30 to the relaxation time distribution function prediction unit 2 of the diagnostic calculation unit 50 via a wired connection, or may be sent wirelessly to the relaxation time distribution function prediction unit 2 of the diagnostic calculation unit 50. In consideration of the effect on the living body and the simplicity of the device, the applied current value and its frequency are preferably, for example, an AC current of 1.0 mA or less in the Hz band to the MHz band, or even in the GHz band.

[0018] (Relaxation Time Distribution Function Prediction Unit) The relaxation time distribution function prediction unit 2 according to this embodiment predicts the relaxation time distribution function γ from the imaginary part Z″ of the measured impedance shown in the following formula (1). The present invention is not limited to the case where the relaxation time distribution function γ is predicted from the imaginary part Z″ of the impedance. The relaxation time distribution function γ can also be predicted from the real part and phase of the impedance, or the real part, imaginary part, and phase of the admittance. In the following formula (1), T means transpose, and R Ｍmeans a column vector consisting of M real number elements. Hereinafter, the current application frequency or voltage application frequency may be simply referred to as the application frequency. For example, M is an application frequency pattern, and when there are many electrodes 20 as described below, there are many application frequency patterns, and the number M is a value such as M=201. m is a real value in the range of 1≦m≦M. The physical relationship between the imaginary part Z″ of the impedance and the relaxation time distribution function γ is mathematically expressed by the following equation (2).

[0019]

[0020] In the above formula (2), f, τ, and ln are the applied frequency, relaxation time, and natural logarithm, respectively; γ is a relaxation time distribution function, which is a function of τ; and Z″ and γ are generally expressed in units of [Ω]. The present invention is characterized in that it does not directly use the value of the imaginary part Z″ of the measured impedance to detect ductal carcinoma (IDC) or the like, but predicts the relaxation time distribution function γ from the imaginary part Z″ and uses the predicted γ to detect ductal carcinoma (IDC) or the like. Hereinafter, when emphasizing "prediction," * may be added to the upper right of the physical symbol. Generally, it is not easy to calculate γ in the integral from the measured imaginary part Z″; in the present invention, γ is calculated from the imaginary part Z″. ＊ A significant feature of the method for predicting is the use of linear regression, nonlinear regression, Gaussian process, etc. For example, if we use the known M imaginary parts Z'' and the unknown (to be predicted) N γ ＊ (referred to as a predicted relaxation time distribution function) can be expressed as a joint probability distribution in a Gaussian process by the following equation (3).

[0021]

[0022] where γ ＊ can be expressed by the above formula (4). N in the above formulas (3) and (4) is, for example, the relaxation time distribution function γ ＊ , where N is an arbitrary number, such as 501, and N and M are independent numbers. GP in the above formula (3) is a function that generates a Gaussian distribution with a mean value of zero and a covariance matrix in ( ), and θ in ( ) １is the noise level, I is the unit matrix expressed by the above formula (5), and R Ｍ×Ｍ is a matrix consisting of M×M real number elements. K in the above equation (3) is a covariance matrix for M known impedances Z″, and is expressed by the following equation (6).

[0023]

[0024] K in the above formula (3) ＊ is a function of M known impedances Z″ and N predicted relaxation time distribution functions γ ＊ and is a covariance matrix between them, and is expressed by the following equation (7).

[0025]

[0026] K in the above formula (3) *＊ is the N predicted relaxation time distribution functions γ ＊ It is the covariance matrix of itself, which is a diagonal matrix expressed by the following equation (8).

[0027]

[0028] k(lnτ) in the above formula (8) ｉ , lnτ ｊ ) is a kernel function, and i, j are integers from 1 to N or M. k(lnτ ｉ， lnτ ｊ ) may use, for example, a Gaussian kernel shown in the following equation (9).

[0029]

[0030] In the above formula (9), θ ２ and θ ３ The transformation operator L(.) used in the covariance matrix of the above formula (3) is defined by the following formula (10), and LK is defined by the following formula (11), and is a transformation of the covariance matrix having the kernel function as an element by the transformation operator L. ２ K is defined by the following formula (12), which is obtained by further transforming formula (11) with a transformation operator L.

[0031]

[0032] From the joint probability distribution equation (3), when Z'' is given, γ ＊ The posterior probability distribution of the mean vector μ∈R is expressed by the following equation (13). Ｎ , covariance matrix Σ∈R Ｎ×Ｍ As γ ＊ The mean vector is expressed by the following equation (14), and the covariance matrix Σ is expressed by the following equation (15).

[0033]

[0034] Next, in the relaxation time distribution function prediction unit 2, the relaxation time distribution function γ ＊ In order to predict as a more specific numerical value, it is preferable to determine the hyperparameter θ of the following formula (17). For example, the hyperparameter θ can be determined using the logarithmic likelihood p expressed by the following formula (16). Here, lnτ in the following formula (16) is expressed by the following formula (18).

[0035]

[0036] The hyperparameters can be optimized using the log likelihood p by a known method, such as the Nelder-Mead method. From the above, N relaxation time distribution functions γ ＊ The relaxation time distribution function prediction unit 2 can predict the predicted relaxation time distribution function γ ＊ is sent to the diagnostic unit 5 or the output unit 6.

[0037] In the present invention, for a breast in which invasive ductal carcinoma exists, the relaxation time distribution function γ ＊ The value of shows a peak at a specific predicted relaxation time τ*, and the relaxation time distribution function γ ＊ It is noteworthy that the values ​​of are significantly different, and this is a significant feature in detecting invasive ductal carcinoma (IDC).

[0038] (Diagnosis unit 5) The diagnosis unit 5 calculates the predicted relaxation time distribution function γ sent from the relaxation time distribution function prediction unit 2.＊ For example, at a specific (peak) predicted relaxation time τ*, the predicted relaxation time distribution function γ ＊ The first peak value (intensity value of the first peak) showing the highest peak can be used for discrimination. For example, the diagnostic unit 5 compares the first peak value with values ​​in a database stored in a storage unit (not shown) to determine whether or not invasive ductal carcinoma (IDC) is present in the measurement object. The diagnostic unit 5 sends the discrimination result to the output unit 6. A specific predicted relaxation time τ* and a predicted relaxation time distribution function γ are calculated in advance from the impedance measurement results of a large number of patients. ＊ The database is prepared by storing the relaxation time distribution function γ ＊ The presence or absence of invasive ductal carcinoma (IDC) in the measurement target may be determined from the relaxation time distribution function γ ＊ Based on this, the relaxation time distribution function γ of the breast that is expected to be the affected area is calculated. ＊ The presence or absence of invasive ductal carcinoma (IDC) in the subject may be determined from the difference between these values.

[0039] (Output Unit 6) The output unit 6 outputs the predicted relaxation time distribution function γ ＊ and outputting at least one of the diagnostic results. Furthermore, the relaxation time τ, the predicted relaxation time τ*, and the predicted relaxation time distribution function γ calculated from the imaginary part and real part of the measured impedance are output. ＊ Detailed data such as the above may be displayed. The output destination may be a display unit such as a liquid crystal display, or a storage device such as a HDD.

[0040] The diagnostic device 100 according to the first embodiment has been described above. The diagnostic device uses a regression model to predict a relaxation time distribution function based on the impedance obtained by measurement, making it possible to determine whether, for example, breast cancer tissue is present in the measured tissue. Breast cancer tissue is a general term for invasive breast cancer, non-invasive breast cancer, ductal carcinoma, and the like.

[0041] (Diagnostic Method) Next, a diagnostic method using the diagnostic device 100 will be described. Fig. 3 is a flowchart of the diagnostic method according to this embodiment. The diagnostic method according to this embodiment includes an impedance measurement step S1 of measuring the impedance of the object to be measured, a relaxation time distribution function prediction step S2 of predicting a relaxation time distribution function from the impedance using a regression model, and a diagnostic step S3 of diagnosing the object to be measured from the predicted relaxation time distribution function. Each step will be described below.

[0042] (Impedance measurement step S1) In the impedance measurement step S1, the impedance of the measurement object is measured. The sensor 10 is brought into contact with the measurement object, and a current or voltage is applied between the electrodes 20 to measure the impedance. There are no particular limitations on the method of applying the current or voltage, but considering the living body, it is desirable to apply an AC current or AC voltage of 1 mA or less in the Hz to GHz band.

[0043] (Relaxation time distribution function prediction step S2) In the relaxation time distribution function prediction step S2, a regression model is used to predict the relaxation time distribution function from the impedance obtained in the impedance measurement step S1. The relaxation time distribution function can be predicted from the impedance by the method described above. Examples of the regression model include linear regression, nonlinear regression, and Gaussian process model. The regression model is preferably a Gaussian process regression model.

[0044] (Diagnostic step S3) In the diagnostic step S3, it is diagnosed whether the measurement target contains cancerous tissue based on the predicted relaxation time distribution function. For example, in the diagnostic step S3, it is determined whether the measurement target contains cancerous tissue based on the predicted relaxation time distribution function and the value of the relaxation time distribution function in the database.

[0045] The above has described the diagnostic method according to the first embodiment. According to the diagnostic method according to this embodiment, it is possible to determine whether cancer tissue is present in the measured tissue from the value of the predicted relaxation time distribution function using a regression model based on the impedance obtained by measurement.

[0046] Second Embodiment Next, a diagnostic device 100A according to a second embodiment of the present invention will be described. The diagnostic device 100A creates a relaxation time distribution image using impedance tomography and a regression model. As shown in FIG. 4 , the diagnostic device 100A includes a current / voltage application / measurement unit 1A and a diagnostic calculation unit 50A. The diagnostic calculation unit 50A includes a relaxation time distribution function prediction unit 2A and an output unit 6. In this second embodiment, the same components as those in the first embodiment are denoted by the same reference numerals, and their description will be omitted, with only the differences being described.

[0047] The diagnostic calculation unit 50A can have the same configuration as the diagnostic calculation unit 50.

[0048] (Current / Voltage Application Measurement Unit) The current / voltage application measurement unit 1A applies a current or voltage between the electrodes 20 and measures the impedance of the measurement object. The measurement object is, for example, biological tissue, specifically, a living body such as a breast. The current / voltage application measurement unit 1A will be explained using Fig. 5. Fig. 5(A) is a schematic diagram of the current / voltage application measurement unit 1A, and Fig. 5(B) is a diagram showing the sensor 10A attached to the subject.

[0049] As shown in FIG. 5A, the current / voltage application measurement unit 1A includes a sensor 10A and a control unit 30A. The sensor 10A includes multiple electrodes 20 (number of electrodes Q) spaced apart in a three-dimensional arrangement, and a support 25 that holds the electrodes 20 and allows the electrodes 20 to be positioned on a portion of the subject. The sensor 10 of the second embodiment includes a support 25 having a dome-shaped internal space S, and multiple electrodes 20 positioned inside the support 25. Here, "positionable on at least a portion of the subject" means that when the subject wears the sensor 10A, as shown in FIG. 5B, the electrodes 20 are positioned on at least a portion of the subject. The portion of the subject is, for example, the breast.

[0050] The method of electrically connecting the electrode 20 and the control unit 30A is not particularly limited. For example, the electrode 20 and the control unit 30A may be connected by a coaxial cable or a lead wire, or the electrode 20 and the control unit 30A may be connected by wiring with woven conductive fibers. Using woven wiring improves the comfort of the support 25.

[0051] After the subject wears the sensor 10A, the current / voltage application / measurement unit 1A applies a predetermined current or voltage between the electrodes 20 and measures the potential difference or current. When applying a current, the potential difference is measured based on a predetermined current application / voltage measurement pattern (a pattern in which two electrodes are selected from a large number of electrodes in sequence, a current is applied, and the potential difference is measured sequentially). At this time, it is desirable to also measure the phase (the time lag between the applied current and the measured potential difference). When applying a voltage, the current is measured based on a predetermined voltage application / current measurement pattern (a pattern in which two electrodes are selected from a large number of electrodes in sequence, a voltage is applied, and the current is measured sequentially). At this time, it is desirable to also measure the phase (the time lag between the applied voltage and the measured current). Hereinafter, the description will focus on the case of applying a current, and detailed description of the case of applying a voltage may be omitted. Here, the number Q of the electrodes 20 is 4 or more, preferably 64 or more. By arranging the electrodes 20 three-dimensionally and having four or more electrodes 20, it is possible to determine the location near the cancerous tissue. To improve the accuracy of the calculation, it is preferable to have a large number of electrodes. The arrangement positions of the electrodes 20 are not particularly limited. The electrodes 20 are preferably arranged so as to surround a part of the person being measured. The arrangement of the electrodes 20 on the sensor 10A will be described below.

[0052] As shown in Fig. 5A, the electrodes 20 are arranged in a circle around the central axis C of the internal space S, constituting a group of multiple annular electrodes 35. By arranging the electrodes in this manner, a two-dimensional electrical characteristic distribution of a portion of the subject can be obtained. The spacing between the electrodes in the group of annular electrodes 35 is not particularly limited, and the electrodes may be arranged at equal intervals along a circle centered on the central axis C. The number of electrodes 20 included in the group of annular electrodes 35 is, for example, four or more.

[0053] The annular electrode groups 35 are arranged at predetermined intervals in the height direction h of the internal space S. In this way, by arranging the annular electrode groups 35 at different positions in the height direction h of the internal space S, a three-dimensional distribution of electrical properties can be obtained. The annular electrode groups 35 may or may not be arranged evenly.

[0054] The support 25 is not particularly limited as long as it can hold the electrode 20 and position the electrode 20 on at least a portion of the person being measured. It is preferable that the support 25 be capable of applying a predetermined pressure to at least a portion of the person being measured. Applying a predetermined pressure improves the adhesion between the electrode 20 and the person being measured, allowing for more accurate application of a current or potential difference and measurement of the potential difference or current. The support 25 also covers at least a portion of the person being measured. The material of the support 25 is preferably a dielectric material such as elastomer, leather, or cloth. The support 25 is not particularly limited, but examples include underwear such as a brassiere.

[0055] (Control Unit) The control unit 30A includes, for example, a multiplexer for switching between current application electrodes (or voltage application electrodes) that apply a current and voltage measurement electrodes that measure a potential difference (or current measurement electrodes that measure a current), and an impedance analyzer that performs voltage measurement (or current measurement) and phase measurement. The control unit 30A executes a predetermined program in, for example, a CPU and controls the multiplexer and the impedance analyzer to perform impedance measurement (measurement of the ratio of the potential difference to the current and its phase). The control unit 30 may be controlled only within the current / voltage application measurement unit 1 to perform impedance measurement, or the control unit 30A may be controlled according to a program executed in the diagnostic calculation unit 50 to perform impedance measurement. The results of the impedance measurement are sent to the relaxation time distribution function prediction unit 2A. The method of transmitting information to the relaxation time distribution function prediction unit 2A is not particularly limited. The information may be sent from the control unit 30A to the relaxation time distribution function prediction unit 2A of the diagnostic calculation unit 50A via a wired connection or wirelessly.

[0056] The control unit 30A applies a current between the electrodes 20 and measures the potential difference based on a predetermined current application / voltage measurement pattern (a pattern that determines which electrodes the current is applied between and which electrodes the potential difference between). Alternatively, the control unit 30 applies a voltage between the electrodes 20 and measures the current based on a predetermined voltage application / current measurement pattern. Whether applying a current or a voltage, the electrodes 20 to which the current (voltage) is applied and the potential difference (current) is measured are not particularly limited. However, it is preferable to "uniformly" apply the current (potential difference) to the three-dimensionally arranged electrodes 20 and measure the potential difference (current). "Uniformly applying the current (potential difference) and measuring the potential difference (current)" means that the current / potential difference is applied and measured so that all of the electrodes 20 are used at least once for the application or measurement of the current / potential difference. Here, the case of measurement within the annular electrode group 35 will be described as an example.

[0057] FIG. 6 is a diagram illustrating a measurement pattern based on the adjacent electrode method. Here, an example will be described in which four annular electrode groups 35 are arranged in the height direction. For example, a current is applied between electrodes No. 1 and No. 2 of the annular electrode group 35 of layer 1, and the potential difference is measured between electrodes No. 3 and No. 4. Measurements are then performed by shifting the electrodes, resulting in 208 data points. Similarly, a current is applied between electrodes No. 17 and No. 18 of the annular electrode group 35 of layer 2, and the potential difference is measured between electrodes No. 19 and No. 20. Similar measurements are then performed for layers 3 and 4, thereby obtaining the electrical property distribution for each layer. Of course, instead of applying a current only between electrodes of the same layer and measuring the voltage as in the above example, it is also possible to apply a current between electrodes across layers and measure the voltage, for example, to apply a current between the electrodes of layer 1 and layer 2 and measure the voltage between the electrodes of layer 3 and layer 4. By applying a current between electrodes across layers and measuring the voltage in this way, it is possible to obtain a more realistic three-dimensional stereoscopic image.

[0058] (Relaxation Time Distribution Function Prediction Unit 2A) In the relaxation time distribution function prediction unit 2 of the first embodiment, when there is only one sensor 10, the imaginary part Z″εR of the measured impedance is calculated from the above equation (10). Ｍ From the relaxation time distribution function γ ＊ ∈R Ｎ In the second embodiment of the present invention, by adding spatial dimensions L and K to the dimensions of the applied frequency M and N, the imaginary part Z″∈R of the impedance measured in space can be predicted in the same manner as in the first embodiment. Ｍ×Ｌ From the spatial relaxation time distribution function γ ＊ ∈R Ｎ×Ｋ where R Ｍ×Ｌ means a matrix with M rows (measured frequency dimension) and L columns (measured space dimension), and R Ｎ×Ｋ means a matrix with N rows (predicted frequency dimension) and K columns (predicted spatial dimension). Of course, the spatial relaxation time distribution function γ may be predicted using a Jacobian matrix (a matrix that associates the relaxation time distribution function γ with the imaginary part Z'' of impedance), which is a conventional general image reconstruction method. This makes it possible to obtain a spatial relaxation time distribution image.

[0059] The relaxation time distribution function prediction unit 2A sends the predicted spatial relaxation time distribution image to the output unit 6.

[0060] (Output Unit 6) The output unit 6 outputs a spatial predicted relaxation time distribution image of the measurement target. The output destination may be a display unit such as a liquid crystal display, or a storage device such as a HDD.

[0061] The diagnostic device 100A according to the second embodiment has been described above. The diagnostic device 100A can generate a relaxation time distribution image by using electrical impedance tomography and a regression model. This can identify, for example, the presence and location of nearby cancerous tissue.

[0062] (Imaging Method) Next, a diagnostic method using the diagnostic device 100A will be described. Fig. 7 is a flowchart of an imaging method according to a second embodiment. The imaging method according to this embodiment includes an impedance measurement step S1A in which a current or a voltage is applied between the electrodes 20, and if a current is applied, the impedance is measured based on a current application / voltage measurement pattern, and if a voltage is applied, the impedance is measured based on a voltage application / current measurement pattern, and a relaxation time distribution prediction step S2A in which a regression model is used from the impedance to predict a spatial relaxation time distribution function (relaxation time distribution image). Each step will be described below.

[0063] (Impedance Measurement Step S1A) In the impedance measurement step S1, a current or a voltage is applied between the electrodes 20. When a current is applied, the impedance is measured based on a current application / voltage measurement pattern, and when a potential difference is applied, the impedance is measured based on a voltage application / current measurement pattern. The current application / voltage measurement pattern (voltage application / current measurement pattern) is not particularly limited, and examples that can be used include an adjacent electrode method, a counter electrode method, and a reference method.

[0064] (Relaxation time distribution prediction step S2A) In the relaxation time distribution prediction step S2A, a regression model is used to predict a relaxation time distribution from the multiple impedances obtained in the impedance measurement step S1, and a spatial relaxation time distribution image is obtained. The relaxation time distribution can be predicted from the measured impedances using the method described above. In the relaxation time distribution prediction step S2A, an image is created based on the relaxation time distribution.

[0065] The above has described the imaging method according to the second embodiment. According to the imaging method according to this embodiment, by using impedance tomography and a regression model to obtain a relaxation time distribution image, it is possible to identify, for example, the presence and location of nearby cancer tissue.

[0066] (Third Embodiment) As shown in the functional block diagram of FIG. 20 , the basic configuration of a determination device 100B of the present invention includes a sensor 10B having multiple electrodes for measuring interelectrode impedance, a current / voltage application / measurement unit 2B that applies an input current or input voltage to the electrodes of the sensor 10B and measures the interelectrode impedance based on the output voltage or output current obtained from the remaining electrodes, a relaxation time distribution function calculation unit 3B that converts the interelectrode impedance into an interelectrode relaxation time distribution function and outputs the resulting characteristic information, a determination unit 4B that determines the presence or absence of an abnormal area that may be a lesion based on various characteristic information included in the interelectrode relaxation time distribution function, and a visualization display unit 5B. Although not shown, the determination device 100B is provided with a switch for setting the display, allowing the user to select an image displaying the determination results and an image displaying the characteristic information. Furthermore, the determination method of the present invention implements the functional block diagram of the determination device 100B described above using software. Furthermore, the imaging method of the present invention visualizes and displays the characteristic information of the interelectrode relaxation time distribution function.

[0067] The determination calculation unit 50B of the determination device 100B includes, for example, a central processing unit (CPU), read-only memory (ROM), random access memory (RAM), and hard disk drive (HDD) / solid state drive (SSD). The relaxation time distribution function calculation unit 3B, determination unit 4B, and visualization display unit 5B are realized by the CPU executing a predetermined program. The determination calculation unit 50B may also control the current / voltage application measurement unit 2B. The program may be acquired via a recording medium or a network. A dedicated hardware configuration may also be used to realize the configuration of the determination device 100B. The determination method of the present invention configures the functional block diagram of the determination device 100B using software. Furthermore, the imaging method of the present invention is characterized in that it visualizes characteristic information of the inter-electrode relaxation time distribution function corresponding to the presence or absence of an abnormal area that may be a lesion, without including a judgment step, as compared with the judgment method.

[0068] (Relaxation Time Distribution Function Calculation Unit 3B) The following describes the relaxation time distribution function calculation unit 3B, which is one of the main parts of the present invention and is almost common to the two embodiments. This relaxation time distribution function calculation unit 3B calculates the inter-electrode relaxation time distribution function γ from the imaginary part Z'' of the measured inter-electrode impedance shown in the above formula (1). The present invention is not limited to the case where the inter-electrode relaxation time distribution function γ is calculated from the imaginary part Z'' of the inter-electrode impedance. The inter-electrode relaxation time distribution function γ can also be calculated from the real part and phase of the inter-electrode impedance, or the real part, imaginary part, and phase of the admittance. T in the above formula (1) means transpose, and R Ｍ means a column vector consisting of M real elements. Hereinafter, the current application frequency or voltage application frequency may be simply referred to as the application frequency. For example, M is an application frequency pattern. As described below, when a large number of electrodes 20 are present, a large number of application frequency patterns exist, and the number M is a numerical value such as M=201. m is a real value in the range of 1≦m≦M. The physical relationship between the imaginary part Z″ of the inter-electrode impedance and the relaxation time distribution function γ is mathematically expressed by the above formula (2). In the above formula (2), f, τ, and ln are the application frequency, relaxation time, and natural logarithm, respectively. γ is the inter-electrode relaxation time distribution function, which is a function of τ. Z″ and γ are generally expressed in units of [Ω].

[0069] The present invention has a major feature in that it does not directly use the value of the imaginary part Z'' of the measured inter-electrode impedance to detect ductal carcinoma (IDC) or the like, but calculates the inter-electrode relaxation time distribution function γ from the imaginary part Z'', and uses the calculated γ to detect ductal carcinoma (IDC) or the like. In the following explanation, when emphasizing "calculated", * may be added to the upper right of the physical symbol. Generally, it is not easy to calculate γ in the integral from the measured imaginary part Z'', and in the present invention, it is possible to calculate γ from the imaginary part Z''. ＊ As a method for calculating γ, linear regression, nonlinear regression, Gaussian process, non-regression model neural network, etc. can be used. For example, when M measured imaginary parts Z″ and N unknown (to be calculated) γ ＊(referred to as the calculated inter-electrode relaxation time distribution function) can be expressed as a joint probability distribution in a Gaussian process, as shown in the above formula (3). For example, ＊ (referred to as the calculated inter-electrode relaxation time distribution function) can be expressed as the joint probability distribution in a Gaussian process by the above formula (3).

[0070] where γ ＊ can be expressed by the above formula (4). N in the above formulas (3) and (4) is, for example, the inter-electrode relaxation time distribution function γ ＊ , where N is an arbitrary number, such as 501, and N and M are independent numbers. GP in the above formula (3) is a function that generates a Gaussian distribution with a mean value of zero and a covariance matrix in ( ), and θ in ( ) １ is the noise level, I is the unit matrix expressed by the above formula (5), and R Ｍ×Ｍ is a matrix consisting of M×M real number elements. K in the above equation (3) is a covariance matrix for M known inter-electrode impedances Z″, and is expressed by the above equation (6).

[0071] K in the above formula (3) ＊ is the M known inter-electrode impedances Z″ and the N calculated relaxation time distribution functions γ ＊ and is the covariance matrix between them, and is expressed by the above formula (7).

[0072] K in the above formula (3) *＊ is the N calculated inter-electrode relaxation time distribution functions γ ＊ It is the covariance matrix of itself, which is a diagonal matrix expressed by the above formula (8). ｉ , lnτ ｊ ) is a kernel function, and i, j are integers from 1 to N or M. k(lnτ ｉ , lnτ ｊ ) may use, for example, the Gaussian kernel shown in the above equation (9).

[0073] In the above formula (9), θ ２ and θ ３The transformation operator L(.) used in the covariance matrix of the above formula (3) is defined by the above formula (10), and LK is the transformation of the covariance matrix having the kernel function as an element by the transformation operator L, defined by the above formula (11). ２ K is defined by the above formula (12) and is obtained by further transforming formula (11) with the transformation operator L.

[0074] The posterior probability distribution of γ* when Z″ is given from the joint probability distribution equation (3) is expressed by the above equation (13). Ｎ , covariance matrix Σ∈R Ｎ×Ｍ As γ ＊ The mean vector is expressed by the above formula (14), and the covariance matrix Σ is expressed by the above formula (15).

[0075] Next, the relaxation time distribution function calculation unit 3B calculates the inter-electrode relaxation time distribution function γ ＊ In order to calculate as a more specific numerical value, it is preferable to determine the hyperparameter θ of the following formula (17). For example, the hyperparameter θ can be determined using the log-likelihood p expressed by the above formula (16). Here, lnτ in the above formula (16) is expressed by the above formula (18).

[0076] The hyperparameters can be optimized using the logarithmic likelihood p by a known method, such as the Nelder-Mead method. From the above, N inter-electrode relaxation time distribution functions γ ＊ The relaxation time distribution function calculation unit 3B can calculate the calculated relaxation time distribution function γ ＊ is sent to the determination unit 4B or the visualization display unit 5B.

[0077] In the present invention, in the case of a breast in which invasive ductal carcinoma is present, the values ​​of the inter-electrode relaxation time distribution function γ* for ductal carcinoma tissue (IDC), normal breast tissue (nGBT), and adipose tissue (Adipose) show peaks at specific calculated relaxation times τ*, and the values ​​of the relaxation time distribution function γ* at these peaks differ significantly. This is a major feature of the present invention, which detects invasive ductal carcinoma (IDC).

[0078] In the fourth embodiment, the sensor 10C is brought into contact with any part of the subject's body to determine whether or not there is an abnormal part near the contact position, which may be a lesion. Therefore, in this embodiment, the position of the sensor 10C can be freely moved to search for the position of the abnormal part, which may be a lesion.

[0079] (Determination Device) Hereinafter, a determination device 100C according to this embodiment will be described with reference to the drawings. As shown in Fig. 21 , the determination device 100C includes a sensor 10C, a current / voltage application / measurement unit 2C, and a determination calculation unit 50C. The determination calculation unit 50C includes a relaxation time distribution function calculation unit 3C, a determination unit 4C, and a visualization display unit 5C.

[0080] (Sensor 10C) Each sensor 10C has a pair of input and output electrodes. A predetermined current (or voltage) obtained from the current / voltage application measurement unit 2C is applied to the input electrodes, and the voltage (or current) obtained from the output electrode is input to the current / voltage application measurement unit 2C as a measured value.

[0081] (Current-Voltage Application Measurement Unit 2C) The current-voltage application measurement unit 2C applies a current or voltage between the electrodes 20C and measures the inter-electrode impedance of the area to be evaluated. The area to be evaluated is, for example, biological tissue, specifically, a living breast or a resected breast. The current-voltage application measurement unit 2C will be described using FIG. 22. FIG. 22(A) is a schematic diagram of the current-voltage application measurement unit 2C in contact with a subject, and FIG. 22(B) is a plan view illustrating an example of an electrode. The drawings used in the following description may show characteristic portions enlarged for ease of understanding, and the dimensional proportions of each component may differ from the actual. The materials, dimensions, etc. exemplified in the following description are merely examples, and the present invention is not limited thereto. Appropriate modifications can be made within the scope of the present invention.

[0082] As shown in FIG. 22(A), the current / voltage application / measurement unit 2C includes a sensor 10C and a control unit 30C. The sensor 10C may be one or more. The shape of the sensor 10C is not particularly limited and may be rectangular, circular, or the like. The sensor 10C may be provided with a support and placed on the breast surface. The shape of the support for the sensor 10C is not particularly limited and may be flat or have a shape that covers the breast surface. The sensor 10C includes electrodes 20C, and the placement positions and spacing of the electrodes 20C are not particularly limited. For example, when using a four-terminal method for the electrode structure, as shown in FIG. 22(B), the sensor 10C includes four electrodes: a low-current electrode (LC), a low-voltage electrode (LP), a high-current electrode (HC), and a high-voltage electrode (HP). For example, four peripheral electrodes of one sensor 10C may serve as the LC electrode and the LP electrode, and a central electrode may serve as the HC electrode and the HP electrode. The electrode 20C may be placed on the area to be determined by making the support adhesive and attaching the support to the area to be determined, or the electrode 20C may be placed on the area to be determined by wrapping the support around the area to be determined.

[0083] Electrode 20C is electrically connected to control unit 30C. There are no particular limitations on the material or shape of electrode 20C as long as it can apply a current or voltage to the subject's breast or excised breast as the area to be assessed. Examples of electrode 20C include metals such as Au, Ag, and Cu, conductive polymers, fibers whose surfaces are coated with metal, and fibers whose surfaces are coated with conductive polymers.

[0084] There are no particular limitations on the method of electrically connecting the electrode 20C and the control unit 30C. For example, the electrode 20C and the control unit 30C may be connected by a coaxial cable or a lead wire, or the electrode 20C and the control unit 30C may be connected by wiring with woven conductive fibers.

[0085] (Control Unit) The control unit 30C includes, for example, a multiplexer for switching between current application electrodes (or voltage application electrodes) that apply a current and voltage measurement electrodes (or current measurement electrodes that measure a current) that measure a potential difference, and an impedance analyzer that performs potential difference (or current measurement) and phase measurement. The impedance analyzer is a component that measures interelectrode impedance, i.e., the ratio of the measured potential difference (applied voltage) to the applied current (measured current), and its phase, by changing the applied frequency and amplitude. The control unit 30C may perform interelectrode impedance measurement (measurement of the ratio of the potential difference to the current, and its phase) by, for example, executing a predetermined program in a CPU and controlling the interelectrode impedance analyzer. The control unit 30C may perform interelectrode impedance measurement by controlling only the current / voltage application / measurement unit 2C, or may perform interelectrode impedance measurement by controlling the control unit 30C according to a program executed by the determination calculation unit 50C. The results of the impedance measurement are sent to the relaxation time distribution function calculation unit 3C. The method of transmitting information to the relaxation time distribution function calculation unit 3C is not particularly limited. The control unit 30C may send the signal to the relaxation time distribution function calculation unit 3C of the determination calculation unit 50C via a wired connection, or may send the signal to the relaxation time distribution function calculation unit 3C of the determination calculation unit 50C wirelessly. In consideration of the effect on the living body and the simplicity of the device, the applied current value and its frequency are preferably, for example, an AC current of 1.0 mA or less in the Hz band to the MHz band, or even in the GHz band.

[0086] (Relaxation time distribution function calculation unit 3C) Relaxation time distribution function calculation unit 3C calculates an inter-electrode relaxation time distribution function based on the inter-electrode impedance obtained from current / voltage application measurement unit 2C, using the above-mentioned equations (1) to (18). The inter-electrode relaxation time distribution function output from relaxation time distribution function calculation unit 3C is input to determination unit 4C and feature information extraction unit 6C.

[0087] The characteristic information extraction unit 6C extracts the calculated inter-electrode relaxation time distribution function γ ＊ The characteristic information useful for the judgment is extracted from the imaginary part and real part of the measured inter-electrode impedance. * , the calculated inter-electrode relaxation time distribution function γ ＊ These include:

[0088] (Determination unit 4C) The determination unit 4C determines the calculated inter-electrode relaxation time distribution function γ ＊ Whether or not invasive ductal carcinoma (IDC) is present in the target area is determined based on the calculated relaxation time τ * In this case, the calculated inter-electrode relaxation time distribution function γ ＊ The first peak value (intensity value of the first peak) showing the highest peak can be used for the judgment. For example, the abnormal region detection unit 4Ca compares the first peak value with values ​​in a database stored in the reference peak value storage unit 4Cb, and judges whether the first peak value exceeds the reference peak value, thereby judging whether there is a possibility of invasive ductal carcinoma (IDC) being present in the region to be judged. The judgment unit 4C sends the judgment result to the visualization display unit 5C as the judgment result. A specific calculated relaxation time τ * The calculated inter-electrode relaxation time distribution function γ ＊ and are stored as reference peak values ​​in the reference peak value storage unit 4Cb, thereby preparing a database. ＊The presence or absence of invasive ductal carcinoma (IDC) in the target area may be determined from the inter-electrode relaxation time distribution function γ ＊ Based on this, the inter-electrode relaxation time distribution function γ ＊ The difference between these values ​​may be used to determine the possibility of the presence or absence of invasive ductal carcinoma (IDC) in the area to be evaluated.

[0089] (Visualization display unit 5C) The visualization display unit 5C inputs at least one of the determination result and the feature information obtained from the feature information extraction unit 6C to the image conversion unit 5Ca. The determination result and the extracted feature information are converted into a character pattern or an image pattern by the image conversion unit 5Ca and output to the display unit 5Cb. The display unit 5Cb may be a display device such as a liquid crystal display, a printer, or a storage device such as a HDD.

[0090] The above describes the determination device 100C according to the fourth embodiment. The determination device 100C uses a regression model to calculate an inter-electrode relaxation time distribution function based on the inter-electrode impedance obtained by measurement, making it possible to determine whether, for example, breast cancer tissue is present in tissue corresponding to a measurement site. Breast cancer tissue is a general term for invasive breast cancer, non-invasive breast cancer, ductal carcinoma, and the like.

[0091] (Determination Method) Next, a determination method using the determination device 100C will be described. FIG. 23 is a flowchart of the determination method according to this embodiment. The determination method according to this embodiment includes an impedance measurement step S2C for measuring the inter-electrode impedance of the determination target region, a relaxation time distribution function calculation step S3C for calculating an inter-electrode relaxation time distribution function from the impedance using a regression model, a display form identification step SC for identifying a display form, a determination step S4C for determining whether cancerous tissue is present near the determination target region (sensor contact portion) based on the inter-electrode relaxation time distribution function, a characteristic information extraction step S6C for extracting characteristic information effective for cancer determination from information included in the inter-electrode relaxation time distribution function, and a visualization display step S5C for displaying a character image or a graph based on the display form. Each step will be described below.

[0092] (Impedance measurement step S2C) In the impedance measurement step S2C, the inter-electrode impedance of the part to be determined is measured. The sensor 10C is brought into contact with the part to be determined, and a current or voltage is applied between each of the electrodes 20C to measure the inter-electrode impedance. The method for applying the current or voltage is not particularly limited, but considering the living body, it is desirable to apply an AC current or AC voltage of 1 mA or less in the Hz to GHz band.

[0093] (Relaxation time distribution function calculation step S3C) In the relaxation time distribution function calculation step S3C, an inter-electrode relaxation time distribution function is calculated using a regression model from the impedance obtained in the impedance measurement step S2C. The inter-electrode relaxation time distribution function can be calculated from the inter-electrode impedance by the method described above. Examples of the regression model include linear regression, non-linear regression, and Gaussian process model. The regression model is preferably a Gaussian process regression model.

[0094] (Display form identification step SC) In the display form identification step SC, it is identified whether the setting of the determination device 100C is set to the determination mode or the peak value display mode. If it is set to the determination mode (Y in the display form identification step SC), the process proceeds to the determination step S4C, and if it is set to the peak value display mode (N in the display form identification step SC), the process proceeds to the feature information extraction step S6C.

[0095] (Determination step S4C) In the determination step S4C, it is determined whether or not there is a possibility that cancerous tissue is contained near the determination target site (sensor contact portion) based on the calculated inter-electrode relaxation time distribution function. For example, in the determination step S4C, a peak value, which is one of the characteristic information of the calculated inter-electrode relaxation time distribution function, is compared with a reference peak value determined based on peak values ​​of inter-electrode relaxation time distribution functions stored in advance in a database, to determine whether or not there is a possibility that cancerous tissue (lesion) is contained in the tissue of the determination target site.

[0096] (Feature Information Extraction Step S6C) When the peak value display mode is set, feature information effective for determining the possibility of cancer is extracted from the information contained in the inter-electrode relaxation time distribution function in the feature information extraction step S6C. In the feature information extraction step S6C, peak values ​​of the inter-electrode relaxation time distribution function obtained in the relaxation time distribution function calculation step are extracted, and the feature information converted into peak values ​​for each pixel is supplied to the visualization display step S5C.

[0097] (Visualization display step S5C) In the visualization display step S5C, if the setting is set to the feature information display mode, the feature information extracted in the feature information extraction step S6C is converted into image information, and a character image or graph indicating the feature is displayed on the display unit. In the visualization display step S5C, the graph or numerical information of the inter-electrode relaxation time distribution function is converted into image information as feature information. On the other hand, if the setting is set to the judgment mode, information on the presence or absence of cancer near the sensor contact part is converted into an image based on the judgment result input in the judgment step S4C, and a character image or graph is displayed on the display unit.

[0098] The above has described the determination method according to the fourth embodiment. According to the determination method according to this embodiment, it is possible to determine whether cancer tissue is present in the measured tissue from the value of the inter-electrode relaxation time distribution function calculated using a regression model based on the inter-electrode impedance obtained by measurement.

[0099] Fifth Embodiment Next, a determination device 100D according to a fifth embodiment of the present invention will be described. The determination device 100D disperses electrodes 20D of a sensor 10D over the entire determination target region, such as the breast of a subject. The determination device 100D measures a large number of inter-electrode impedances while sequentially switching between electrodes that input an applied current (or voltage) and electrodes that output a measured voltage (current) using a multiplexer in a current / voltage application / measurement unit 2D. The device converts the measured inter-electrode impedances into an inter-electrode relaxation time distribution function, converts the peak value of the inter-electrode relaxation time distribution function into a peak value for each pixel, and compares the peak value for each pixel with a reference peak value to determine the presence or absence of an abnormal region that may be a lesion. The determination results are visualized and displayed, thereby displaying the position, size, and level of abnormality of an abnormal region that may be a lesion in the determination target region. In this embodiment, the peak value for each pixel is visualized and displayed as feature information.

[0100] 24 , a determination device 100D includes a sensor 10D, a current / voltage application / measurement unit 1, and a determination calculation unit 50D. The determination calculation unit 50D includes a relaxation time distribution function calculation unit 3D, a determination unit 4D, and a visualization display unit 5D. In this second embodiment, parts that correspond to the same components as those in the first embodiment are designated with the letter B instead of the letter A in the same reference numerals, and their description will be omitted. Only the differences will be described to clarify the relationship between the present invention and the first and second embodiments.

[0101] (Sensor 10D) The sensor 10D shown in FIG. 25 has a large number of electrodes arranged over the entire area to be evaluated to measure the impedance between the electrodes. To this end, a predetermined current (or voltage) is applied to the electrodes from a current / voltage application measurement unit 2D in a determination calculation unit 50D, and the voltage (or current) generated at the other electrodes is supplied to the current / voltage application measurement unit 2D.

[0102] As shown in FIG. 25A, the current / voltage application / measurement unit 2D includes a sensor 10D and a control unit 30D. The sensor 10D includes multiple electrodes 20D (number of electrodes Q) arranged three-dimensionally at intervals, and a support 25 that holds the electrodes 20D and allows the electrodes 20D to be placed on a target area of ​​the subject. The sensor 10D of the fifth embodiment includes a support 25 having a dome-shaped internal space S and multiple electrodes 20D arranged inside the support 25. Here, "able to be placed on at least a target area of ​​the subject" means that when the subject wears the sensor 10B as shown in FIG. 25B, the electrodes 20D are placed on at least a target area of ​​the subject. The target area of ​​the subject is, for example, the breast.

[0103] The method of electrically connecting the electrode 20D and the control unit 30D is not particularly limited. For example, the electrode 20D and the control unit 30D may be connected by a coaxial cable or a lead wire, or the electrode 20D and the control unit 30D may be connected by a wiring with woven conductive fibers. Using a woven wiring improves the comfort of the support 25.

[0104] After the subject wears the sensor 10D, the current / voltage application / measurement unit 2D applies a predetermined current or voltage between the electrodes 20D and measures the potential difference or current. When applying a current, the potential difference is measured based on a predetermined current application / voltage measurement pattern (a pattern in which two electrodes are selected from a large number of electrodes in sequence, a current is applied, and the potential difference is measured sequentially). At this time, it is desirable to also measure the phase (the time lag between the applied current and the measured potential difference). When applying a voltage, the current is measured based on a predetermined voltage application / current measurement pattern (a pattern in which two electrodes are selected from a large number of electrodes in sequence, a voltage is applied, and the current is measured sequentially). At this time, it is desirable to also measure the phase (the time lag between the applied voltage and the measured current). Hereinafter, the description will focus on the case of applying a current, and detailed descriptions of the case of applying a voltage may be omitted. Here, the number Q of the electrodes 20D is 4 or more, preferably 64 or more. By arranging the electrodes 20D three-dimensionally and having four or more electrodes 20B, it is possible to determine the location near the cancerous tissue. To improve the accuracy of the calculation, it is preferable to have a large number of electrodes. The arrangement position of the electrodes 20D is not particularly limited. It is preferable that the electrodes 20D are arranged so as to surround the part of the subject to be determined. The arrangement of the electrodes 20D on the sensor 10D will be described below.

[0105] As shown in Fig. 25(A), a plurality of annular electrode groups 35 are formed in which the electrodes 20D are arranged in a circle around the central axis C of the internal space S. By arranging the electrodes in this manner, a two-dimensional electrical characteristic distribution (peak value distribution for each pixel) can be obtained for each annular electrode of the region to be evaluated of the subject. The spacing between the electrodes in the annular electrode group 35 is not particularly limited, and the electrodes may be arranged at equal intervals along a circle centered on the central axis C. The number of electrodes 20D included in the annular electrode group 35 is, for example, four or more.

[0106] The annular electrode groups 35 are arranged at predetermined intervals in the height direction h of the internal space S. In this way, by arranging the annular electrode groups 35 at different positions in the height direction h of the internal space S, a three-dimensional electrical property distribution (peak value distribution for each pixel) can be obtained. The annular electrode groups 35 may or may not be arranged evenly.

[0107] The support 25 is not particularly limited as long as it can hold the electrode 20D and can position the electrode 20D at least on the subject's area to be assessed. It is preferable that the support 25 be capable of applying a predetermined pressure to at least the subject's area to be assessed. Applying a predetermined pressure improves the adhesion between the electrode 20B and the subject, allowing for more accurate application of a current or potential difference and measurement of the potential difference or current. The support 25 also covers at least the subject's area to be assessed. The material of the support 25 is preferably a dielectric material such as elastomer, leather, or cloth. The support 25 is not particularly limited, but examples include underwear such as a brassiere.

[0108] (Control Unit) The control unit 30D includes, for example, a multiplexer for switching between current application electrodes (or voltage application electrodes) that apply a current and voltage measurement electrodes (or current measurement electrodes that measure a current) that measure a potential difference, and an impedance analyzer that performs voltage measurement (or current measurement) and phase measurement. The control unit 30D executes a predetermined program in, for example, a CPU and controls the multiplexer and the impedance analyzer to perform interelectrode impedance measurement (measurement of the ratio of potential difference to current and its phase). The control unit 30D may be controlled only within the current / voltage application / measurement unit 2D to perform interelectrode impedance measurement, or may be controlled according to a program executed by the determination / calculation unit 50D to perform interelectrode impedance measurement. The results of the interelectrode impedance measurement are sent to the relaxation time distribution function calculation unit 3D. The method of transmitting information to the relaxation time distribution function calculation unit 3D is not particularly limited. The control unit 30D may send the relaxation time distribution function calculation unit 3D of the determination calculation unit 50D via a wired connection, or may send the relaxation time distribution function calculation unit 3D of the determination calculation unit 50D via a wireless connection.

[0109] The control unit 30D applies a current between the electrodes 20D and measures the potential difference based on a predetermined current application / voltage measurement pattern (a pattern that determines which electrodes the current is applied between and which electrodes the potential difference between). Alternatively, the control unit 30D applies a voltage between the electrodes 20D and measures the current based on a predetermined voltage application / current measurement pattern. Whether applying a current or a voltage, the electrodes 20D to which the current (voltage) is applied and the potential difference (current) is measured are not particularly limited. However, it is preferable to "uniformly" apply the current (potential difference) to the three-dimensionally arranged electrodes 20D and measure the potential difference (current). "Uniformly applying the current (potential difference) and measuring the potential difference (current)" means that the current / potential difference is applied and measured so that all of the electrodes 20D are used at least once for the application or measurement of the current / potential difference. Here, the case of measurement within the annular electrode group 35 will be described as an example.

[0110] FIG. 6 is a diagram illustrating a measurement pattern based on the adjacent electrode method. Here, an example will be described in which four annular electrode groups 35 are arranged in the height direction. For example, a current is applied between electrodes No. 1 and No. 2 of the annular electrode group 35 of layer 1, and the potential difference is measured between electrodes No. 3 and No. 4. Measurements are then performed by shifting the electrodes, resulting in 208 data points. Similarly, a current is applied between electrodes No. 17 and No. 18 of the annular electrode group 35 of layer 2, and the potential difference is measured between electrodes No. 19 and No. 20. Similar measurements are then performed for layers 3 and 4, thereby obtaining the electrical property distribution (peak value distribution for each pixel) for each layer. Of course, instead of applying a current only between electrodes of the same layer and measuring the voltage as in the above example, it is also possible to apply a current between electrodes across layers and measure the voltage, for example, to apply a current between the electrodes of layer 1 and layer 2 and measure the voltage between the electrodes of layer 3 and layer 4. By applying a current between electrodes across layers and measuring the voltage in this way, it is possible to obtain a more realistic three-dimensional stereoscopic image.

[0111] (Relaxation time distribution function calculation unit 3B) The relaxation time distribution function calculation unit 3B inputs the inter-electrode impedance obtained from the current / voltage application measurement unit 2B and calculates the inter-electrode relaxation time distribution function. In this embodiment, by adding spatial dimensions L and K to the dimensions of the applied frequencies M and N in the above-mentioned equations (1) to (18), the imaginary part Z''∈R of the inter-electrode impedance measured in space can be calculated in the same way as in the first embodiment. Ｍ×Ｌ From the above, the spatial relaxation time distribution function γ*∈R Ｎ×Ｋ where R Ｍ×Ｌ means a matrix with M rows (measured frequency dimension) and L columns (measured space dimension), and R Ｎ×Ｋ means a matrix with N rows (calculated frequency dimension) and K columns (calculated spatial dimension). Of course, the spatial relaxation time distribution function γ may be calculated using a Jacobian matrix (a matrix relating the inter-electrode relaxation time distribution function γ and the imaginary part Z'' of the impedance), which is a conventional general image reconstruction method.

[0112] (Feature information extraction unit 6D) The feature information extraction unit 6D inputs the inter-electrode relaxation time distribution function to a peak value conversion unit 6Da, calculates the peak value (maximum value) of each inter-electrode relaxation time distribution function, and then converts the peak value of the inter-electrode relaxation time distribution function into a peak value for each pixel using the sensitivity determinant stored in the sensitivity determinant memory unit 6Db, and supplies the peak value for each pixel as feature information of the inter-electrode relaxation time distribution function to the determination unit 4D and the visualization display unit 5D.

[0113] (Determination unit 4D) The determination unit 4D, which receives the peak value for each pixel, compares the peak value for each pixel with the reference peak value stored in the reference peak value storage unit 4Db in the abnormal portion detection unit 4Da, thereby determining whether or not there is an abnormal portion or a non-normal portion for each pixel. Note that the reference peak value storage unit 4Db has previously determined and stored a reference peak value for each pixel for the same patient so that the abnormal portions and normal portions determined in the X-ray tomographic image correspond to the abnormal portions and normal portions determined by the determination device of the present invention.

[0114] (Visualization display unit 5D) The visualization display unit 5D inputs the peak values ​​for each pixel input from the peak value conversion unit 6Da to the image conversion unit 5Da, visualizes the peak values, supplies them to the display unit 5Db, and displays a distribution image of the peak values ​​on the display unit 5Db. The visualization display unit 5D visualizes the judgment results input from the judgment unit 4D, supplies them to the display unit 5Db, and displays a distribution image of abnormal areas and normal areas on the display unit 5Db. The display unit 5Db may be a liquid crystal panel or a printer.

[0115] The determination device 100D according to the fifth embodiment has been described above. The determination device 100D can display the position and size of an abnormal portion in a measurement target portion.

[0116] (Determination Method) Next, a determination method using the determination device 100D will be described. Fig. 26 is a flowchart of the determination method according to the fifth embodiment, and the reference numerals of the respective steps correspond to the components of the determination device of the present embodiment shown in Fig. 24.

[0117] In the impedance measurement step S2D, a current or a voltage is applied between the electrodes 20D, and the inter-electrode impedance is measured based on a current application / voltage measurement pattern when a current is applied, and the inter-electrode impedance is measured based on a voltage application / current measurement pattern when a potential difference is applied. The current application / voltage measurement pattern (voltage application / current measurement pattern) is not particularly limited, and examples that can be used include an adjacent electrode method, a counter electrode method, and a reference method.

[0118] (Relaxation Time Distribution Calculation Step S3D) In ​​the relaxation time distribution calculation step S3D, an inter-electrode relaxation time distribution function is calculated using a regression model from the plurality of inter-electrode impedances obtained in the impedance measurement step S2D.

[0119] In the peak value conversion step S6D, the peak value (maximum value) of the inter-electrode relaxation time distribution function is extracted and converted into a peak value for each pixel. In this embodiment, this peak value for each pixel corresponds to the characteristic information of the inter-electrode relaxation time distribution function of the present invention.

[0120] (Display form identification step SD) In ​​the display form identification step SD, it is identified whether the setting of the determination device 100D is set to the determination mode or the peak value display mode. If it is set to the determination mode (Y in the display form identification step SD), the process proceeds to the determination step S4D. If it is set to the peak value display mode (N in the display form identification step SD), the process proceeds to the visualized display step S5D.

[0121] (Determination Step S4D) When the determination mode is set, in the determination step S4D, the area is classified into an abnormal area and a normal area by comparison with the reference peak value.

[0122] (Visualization display step S5D) In ​​the visualization display step S5D, if the setting is set to the peak value display mode, the peak value for each pixel input from the peak value conversion step S6D is converted into an image and a peak value distribution image is displayed on the display unit. On the other hand, if the setting is set to the judgment mode, the judgment object is displayed divided into normal and abnormal parts based on the judgment result input from the judgment step S4D.

[0123] The imaging method according to the fifth embodiment has been described above. According to the imaging method according to the present embodiment, by using impedance tomography and a regression model to obtain an image of the relaxation time distribution image determination result, it is possible to identify, for example, the presence or absence, as well as the location and size of an abnormal area that may be cancerous tissue (lesion).

[0124] In the judgment methods of the two embodiments (fourth and fifth embodiments) described above, if the user wishes to display feature information rather than the judgment result, the judgment device can be set to the feature display side, allowing the feature information to be displayed without displaying the judgment result, and the flowchart excluding the judgment step corresponds to the image display method of the present invention.

[0125] The technical scope of the present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the spirit of the present invention. In addition, the components in the above-described embodiments can be replaced with well-known components as appropriate, and the above-described modifications can be combined as appropriate, without departing from the spirit of the present invention.

[0126] The present invention will be described below with reference to examples, but the present invention is not limited to the following examples. In the following examples, evaluations are performed using excised breast samples to clarify the relationship with cancer tissue, but the measurement target of the present invention is not limited to excised breasts.

[0127] Example 1: Actual breast samples were applied to the first embodiment. Ten breast samples were prepared from ten patients with ductal carcinoma, and an attempt was made to detect the presence or absence of breast cancer. Table 1 shows the patient data and cancer information.

[0128]

[0129] For the evaluation device shown in Figure 8, an impedance analyzer (model number IM357) was used. Based on the four-terminal method, a rectangular sensor (Figure 2(B)) consisting of five electrodes was used. The center of the sensor represented the high-current electrode (HC) and high-potential electrode (HP), and the other four electrodes represented the low-current electrode (LC) and low-potential electrode (LP). The applied frequency f of an AC current with an amplitude of 0.1 mA was varied from 4 Hz to 5 MHz at M = 201 points, and the impedance was measured between the HC and LC electrodes. Impedance measurements were performed three times at room temperature, and the average value was used. The breast sample primarily contained ductal cancer tissue, normal breast tissue, and adipose tissue. The locations of these tissues were determined visually and by touch, and the impedance was measured near the ductal cancer tissue, normal breast tissue, and adipose tissue. Here, "near" in the context of the vicinity of ductal cancer tissue, the vicinity of normal mammary gland tissue, and the vicinity of adipose tissue refers to the region that includes the boundary between the tissue in question and other tissues and also includes tissues other than the tissue in question. For example, the vicinity of ductal cancer tissue includes not only ductal cancer tissue but also normal mammary gland tissue and adipose tissue, and the measured impedance does not purely measure only ductal cancer tissue, normal mammary gland tissue, and adipose tissue.

[0130] FIG. 9 is a Bode plot of the applied current frequency and the imaginary part of impedance Z″ in the vicinity of ductal cancer tissue (IDC), normal mammary gland tissue (nGBT), and adipose tissue (adipose) of a breast sample measured using the evaluation device of FIG. 8 . In FIG. 9 , it is difficult to clearly distinguish between ductal cancer tissue, normal mammary gland tissue, and adipose tissue from these simple Bode plots.

[0131] Next, from the Bode plot of the current application frequency and the imaginary part of the impedance Z″ of the breast sample in FIG. 9, the relaxation time distribution function prediction unit 2 was used to predict the relaxation time distribution function γ* in the vicinity of ductal cancer tissue (IDC), in the vicinity of normal breast tissue (nGBT), and in the vicinity of adipose tissue (nGBT). The obtained results are shown in FIG. 10.

[0132] Figure 10 shows the relationship between the relaxation time τ and the relaxation time distribution function γ in the vicinity of ductal carcinoma tissue (IDC), normal breast tissue (nGBT), and adipose tissue (nGBT), comparing the values ​​of the relaxation time distribution function γ of the first peak indicated by the subscript 1. No * is specifically added to the horizontal and vertical axes in the figure. Two prominent peaks were confirmed in breast samples Sa1 to Sa10, as indicated by the subscripts γ1 and γ2.

[0133] Figure 11(a) plots the relaxation time distribution function γ of the first peak in the vicinity of ductal cancer tissue (IDC), normal breast tissue (nGBT), and adipose tissue (adipose) for breast samples Sa1 to Sa10. As shown in Figure 11(a), the γ of the first peak near ductal cancer tissue was found to be higher than the γ of the normal breast tissue and the γ of the adipose tissue. Figure 11(b) plots the relaxation time distribution function γ of the second peak. The γ of the second peak near ductal cancer tissue was found to be similar to the γ of the normal breast tissue and the γ of the adipose tissue.

[0134] Example 2: With the cooperation of six actual patients with ductal carcinoma, breast samples after mastectomy were applied to the second embodiment. Figure 12 shows a determination device using electrical impedance tomography, consisting of an impedance analyzer, a multiplexer, a breast cup, and a PC. Figure 13(A) shows details of the breast cup, which is dome-shaped to fit the shape of a human breast. It has a diameter of 140 mm and a depth of 70 mm. As shown in Figure 6, the electrodes are arranged in four layers, with 16 5 mm diameter electrodes per layer, for a total of 64 electrodes. The multiplexer functions as a digital switch to adjust the current injection pattern to the 64 electrodes in the breast cup. 21 applied current frequencies ranging from 1000 kHz to 5 MHz were used for the 64 electrodes using the four-wire adjacent method. Data processing was performed using Python software on a PC in the final stage. Figure 13(B) shows an example of a breast sample after mastectomy used in Example 2. The nipple was positioned appropriately by placing it in a quadrant at the center of the breast cup. The image reconstruction method used the first embodiment to estimate the relaxation time distribution function γ* in four layers, and then imaged the electrical conductivity σ using the general Jacobian matrix and the Gauss-Newton method (Equation (19) below). In Equation (19), R is the regularization matrix, and λ is the relaxation coefficient scalar automatically determined by the L-curve method (Hansen and O'Leary 1993).

[0135]

[0136] To compare the location of ductal carcinomas obtained from the reconstructed images using the electrical impedance tomography screening device, the mammographic images were compared with images processed by 3D slicer software. As shown in Figure 13(A), the location of breast cancers within the breast was examined using four quadrants: (I) upper outer, (II) upper inner, (III) lower inner, and (IV) lower outer.

[0137] ​Figure 14(A) shows the reconstructed image of breast sample Sa1 using an evaluation device using electrical impedance tomography. Figure 14(B) shows the mammography image processed using 3D slicer software. The arrow in Figure 14(B) indicates the location of the IDC. Figure 14(C) is a photograph of breast sample Sa1 placed in a breast cup. The results show that, except for breast samples Sa2, Sa3, and Sa5, the IDC is present only in the first layer from the mammography image and the evaluation device using electrical impedance tomography.

[0138] The results of the mammography image (FIG. 14(B)) processed using 3D Slicer software are compared with the results of the electrical impedance tomography (FIG. 14(A)). The mammography image (FIG. 14(C)) of breast sample Sa1 shows that the IDC is present in the left breast, located between the inner inferior (III) and outer inferior (IV) quadrants, whereas the IDC quadrant location is (III) according to FIG. 14(A).

[0139] FIG. 15(A) shows the results of a reconstructed image (Layer 1 result) of breast sample Sa2 obtained by an evaluation device using electrical impedance tomography. FIG. 15(B) shows the results of a reconstructed image (Layer 2 result) of breast sample Sa2 obtained by an evaluation device using electrical impedance tomography. FIG. 15(C) shows the results of a mammography image processed using 3D Slicer software. The arrow in FIG. 15(C) indicates the location of the IDC. From FIGS. 15(A) and 15(B), it was found that breast sample Sa2 had an IDC in the right breast. FIG. 15(C) shows the results of processing using 3D Slicer. The IDC is clearly visible in the lower outer quadrant (IV).

[0140] Figure 16(A) shows the results of a reconstructed image (Layer 1 result) of breast sample Sa3 obtained by an evaluation device using electrical impedance tomography. Figure 16(B) shows the results of a reconstructed image (Layer 2 result) of breast sample Sa3 obtained by an evaluation device using electrical impedance tomography. Figure 16(C) shows the results of a mammography image processed using 3D slicer software. The arrow in Figure 16(C) indicates the position of the IDC.

[0141] Figures 16(A) and 16(B) show the quadrant locations of the IDC in the left breast. As shown in Figure 16(C), the results of 3D processing of breast sample Sa3 showed a slight tendency for the IDC quadrant locations to be visible in the superior outer (I) and superior inner (II) regions.

[0142] Figure 17(A) shows the results of a reconstructed image (Layer 1) of breast sample Sa4 using an evaluation device using electrical impedance tomography. Figure 17(B) shows the results of a mammography image processed using 3D Slicer software. The arrow in Figure 17(B) indicates the location of the IDC.

[0143] Figure 17(A) shows the quadrant location of the IDC in the region between (I) and (II). Figure 17(B) shows the quadrant location of the IDC in the right breast. The IDC in the (II) upper inner region was clearly identified. Conversely, the results of the determination device using electrical impedance tomography of breast sample Sa4 (Figure 17(A)) accurately showed the quadrant location of the cancer in the (II) upper outer region of the right breast.

[0144] FIG. 18(A) shows the results of a reconstructed image (Layer 1 result) of breast sample Sa5 obtained by an evaluation device using electrical impedance tomography. FIG. 18(B) shows the results of a reconstructed image (Layer 2 result) of breast sample Sa5 obtained by an evaluation device using electrical impedance tomography. FIG. 18(C) shows the results of a reconstructed image (Layer 3 result) of breast sample Sa5 obtained by an evaluation device using electrical impedance tomography. FIG. 18(D) shows the results of a mammography image processed using 3D slicer software. The arrow in FIG. 18(D) indicates the position of the IDC.

[0145] As shown in Figure 18(D), breast sample Sa5 has an IDC in the outer inferior quadrant (IV) of the left breast. Furthermore, as shown in Figures 18(A)-18(C), the IDC is present across three different layers, designated as Layers 1-3. Notably, the IDC is consistently located in the outer inferior quadrant (IV) across all layers, consistent with the mammographic findings.

[0146] Figure 19(A) shows the results of a reconstructed image (Layer 1) of breast sample Sa6 using an evaluation device using electrical impedance tomography. Figure 19(B) shows the results of a mammography image processed using 3D slicer software. The arrow in Figure 19(B) indicates the location of the IDC.

[0147] The results in Figure 19(B) indicate that breast sample Sa6 has an IDC in the left breast, and the exact location of this IDC is in the lower outer quadrant (IV) of the left breast. However, the results in Figure 19(A) indicate that the IDC is located in (III). This differs from the results from the 3D slicer (Figure 19(B)). Nevertheless, as shown in Figure 11, the results from the assessment device using electrical impedance tomography are in good agreement with the breast sample placement in (III). Despite the differences in interpretation of mammography images, it was found that the assessment device using electrical impedance tomography at this frequency provides information corresponding to the true location of the IDC in (III). From the above, it was found that imaging the relaxation time distribution function γ not only determines the presence or absence of cancer, but also allows the location of the ductal carcinoma tissue (IDC) in each breast sample Sa1-Sa6 by dividing the breast into four quadrants: the upper outer quadrant (I), the lower outer quadrant (II), the upper inner quadrant (III), and the lower inner quadrant (IV).

[0148] The determination device of the present disclosure is capable of more precisely determining cancerous tissue, and therefore has high industrial applicability.

[0149] 1 Current / voltage application measurement unit, 2 Relaxation time distribution function prediction unit, 5 Determination unit, 6 Output unit, 10 Sensor, 20 Electrode, 100 Determination device

Claims

1. A diagnostic device comprising: a sensor having a plurality of electrodes; a current / voltage application measurement unit that applies a current or voltage between the electrodes and measures impedance; and a relaxation time distribution function prediction unit that predicts a relaxation time distribution function from the impedance using a regression model.

2. The diagnostic device according to claim 1, wherein said relaxation time distribution function predicting section predicts a relaxation time distribution image from said impedance using said regression model.

3. A diagnostic device as claimed in claim 1 or 2, wherein the sensor comprises: a support having a dome-shaped internal space; and a plurality of electrodes arranged inside the support, wherein each of the electrodes constitutes a plurality of annular electrode groups arranged in a circle around the central axis of the internal space, and each of the annular electrode groups is arranged at a predetermined interval in the height direction of the internal space.

4. A diagnostic method comprising: an impedance measurement step of applying a current or a voltage between electrodes, and measuring the impedance based on a current application / voltage measurement pattern when the current is applied, or measuring the impedance based on a voltage application / current measurement pattern when the voltage is applied; and a relaxation time distribution function prediction step of predicting a relaxation time distribution function from the impedance using a regression model.

5. An imaging method comprising: an impedance measuring step of measuring the impedance of a measurement object; and a relaxation time distribution prediction step of obtaining a relaxation time distribution image from the impedance using a regression model.

6. A determination device comprising: a sensor having a plurality of electrodes; a current / voltage application / measurement unit that applies a current or voltage between the electrodes and measures the inter-electrode impedance; a relaxation time distribution function calculation unit that calculates an inter-electrode relaxation time distribution function from the inter-electrode impedance; and a determination unit that determines the presence or absence of an abnormal area that may be a lesion based on characteristic information of the inter-electrode relaxation time distribution function.

7. The determination device according to claim 6, wherein the determination section includes an abnormal portion detection section that determines whether or not the peak value of the inter-electrode relaxation time distribution function exceeds a reference peak value.

8. The determination device according to claim 6, further comprising a feature information extraction unit that extracts peak values ​​of the inter-electrode relaxation time distribution function obtained from the relaxation time distribution function calculation unit, converts the peak values ​​into peak values ​​for each pixel, and supplies the resulting feature information to the determination unit.

9. A determination device as described in claim 6 or 8, wherein the sensor comprises: a support having a dome-shaped internal space; and a plurality of electrodes arranged inside the support, wherein each of the electrodes constitutes a plurality of annular electrode groups arranged in a circle around the central axis of the internal space, and each of the annular electrode groups is arranged at a predetermined interval in the height direction of the internal space.

10. A determination method comprising: an impedance measurement step of applying a current or a voltage between each electrode, and measuring the inter-electrode impedance based on a current application / voltage measurement pattern when applying the current, and measuring the inter-electrode impedance based on a voltage application / current measurement pattern when applying the voltage; a relaxation time distribution function calculation step of calculating an inter-electrode relaxation time distribution function using the inter-electrode impedance; a determination step of determining the presence or absence of an abnormal area that may be a focus of disease based on characteristic information of the inter-electrode relaxation time distribution function; and a visualization display step of visually displaying the determination results.

11. An imaging method comprising: an impedance measuring step of measuring the inter-electrode impedance of a measurement target; a relaxation time distribution function calculating step of calculating an inter-electrode relaxation time distribution function from the inter-electrode impedance; and a visualization display step of converting characteristic information of the inter-electrode relaxation time distribution function into image information and displaying it.

12. An imaging method according to claim 11, characterized in that said visualization display step converts the graph or numerical information of said inter-electrode relaxation time distribution function into image information as characteristic information.

13. The imaging method according to claim 11, characterized in that it includes a feature information extraction step of extracting peak values ​​of the inter-electrode relaxation time distribution function obtained by the relaxation time distribution function calculation step, converting the peak values ​​into peak values ​​for each pixel, and supplying the resulting feature information to the visualization display step.

Images