JP2023097742A - Abnormal tissue detection method, abnormal tissue detection device and program - Google Patents

Abstract

To provide an abnormal tissue detection method, abnormal tissue detection device and program which can suppress the influence of a gap, a foreign matter or the like included in a confocal image and detect an abnormal tissue accurately.SOLUTION: An abnormal tissue detection method according to the present invention includes: bringing an antenna part having a transmission antenna and a reception antenna into contact with an inspection portion of a subject in a plurality of azimuths; generating a plurality of pieces of confocal image data on the basis of a reflection signal received by the reception antenna; and estimating a position of an abnormal tissue on the basis of the overlapping state of the region of the abnormal tissue estimated with the plurality of pieces of confocal image data.SELECTED DRAWING: Figure 6

Description

TECHNICAL FIELD The present invention relates to an abnormal tissue detection method and the like, and more particularly to an abnormal tissue detection method, an abnormal tissue detection apparatus, and a program for detecting abnormal tissue using microwave reflection.

An abnormal tissue detection device has been developed that transmits and receives microwaves in a plurality of combinations of transmitting antennas and receiving antennas with different positional relationships, and estimates the position of abnormal tissue in the body from the microwave reflected by the abnormal tissue. (for example, Patent Document 1).

Such an abnormal tissue detection apparatus using microwaves can reduce the influence on the human body as compared with an inspection apparatus using radiation such as X-rays. Moreover, there are advantages in that the operation is easy and the device can be miniaturized.

JP 2018-51303 A

The abnormal tissue detection device of Patent Document 1 includes an antenna array composed of a plurality of transmitting antennas and receiving antennas, and an antenna cover. Then, the abnormal tissue is detected by bringing the antenna cover into contact with the test site of the subject and transmitting/receiving microwaves at a plurality of antenna positions while rotating the antenna array and the antenna cover. The abnormal tissue detection device of Patent Document 1 detects the reflected waves generated on the surface of the abnormal tissue due to the difference in dielectric constant, calculates the length of the signal path, generates a confocal image, and estimates the position of the abnormal tissue. . Therefore, if there is an air gap or foreign matter between the antenna cover and the subject's body surface, microwaves may be reflected on the surface of the air gap or foreign matter, and noise may be detected.

The present invention has been made in view of the above circumstances, and provides an abnormal tissue detection method and an abnormal tissue detection method capable of suppressing the effects of voids, foreign matter, etc. contained in confocal images and accurately detecting abnormal tissue. The purpose is to provide a device and a program.

In order to achieve the above object, in the abnormal tissue detection method according to the first aspect of the present invention,
An antenna unit having a transmitting antenna and a receiving antenna is brought into contact with a test site of a subject in a plurality of orientations, and based on reflected signals received by the receiving antenna, a plurality of confocal image data are generated,
A position of the abnormal tissue is estimated based on overlapping states of regions of the abnormal tissue estimated by the plurality of confocal image data.

Also, the overlapping state is derived by calculating an inner product of the plurality of confocal image data,
You can do it.

Further, the overlapping state is derived by binarizing the plurality of confocal image data and calculating an inner product,
You can do it.

Further, calculating a correlation coefficient indicating the degree of similarity of the plurality of confocal image data,
Deriving the overlapping state by displacing at least one of the plurality of confocal image data so that the correlation coefficient is maximized and calculating an inner product of the plurality of confocal image data.
You can do it.

The abnormal tissue detection device according to the second aspect of the present invention comprises
An antenna unit having a transmitting antenna for transmitting a microwave impulse signal, a receiving antenna for receiving the impulse signal, and an antenna cover for contacting a subject;
a drive unit that rotates the antenna unit;
a confocal image generating unit that generates confocal image data based on the reflected signal of the impulse signal received by the receiving antenna;
An image processing unit that removes noise components from the confocal image generated by the confocal image generation unit,
The image processing unit
Based on the overlapping state of the region of the abnormal tissue estimated by a plurality of confocal image data generated based on the received data acquired by bringing the antenna unit into contact with the subject in a plurality of orientations, the abnormal tissue region is determined. Estimate location.

A program according to a third aspect of the present invention comprises
the computer,
Confocal image generation for generating a plurality of confocal image data based on received signals received by the receiving antenna by bringing an antenna unit having a transmitting antenna and a receiving antenna into contact with a test site of a subject in a plurality of orientations. part,
an image processing unit that estimates the position of the abnormal tissue based on the overlapping state of the regions of the abnormal tissue estimated by the plurality of confocal image data;
function as

According to the abnormal tissue detection method, abnormal tissue detection apparatus, and program of the present invention, a confocal image is generated based on the inner product of a plurality of confocal image data generated by contacting the subject with the antenna unit in different orientations. Therefore, it is possible to suppress the influence of a gap between the detection device and the subject, foreign matter, and the like, and to detect abnormal tissue with high accuracy.

1A is a perspective view showing the appearance of an abnormal tissue detection device according to Embodiment 1, and FIG. 1B is an exploded perspective view showing the configuration of the abnormal tissue detection device according to Embodiment 1. FIG. (A) is a side view showing the configuration of an antenna array, and (B) is a plan view showing the configuration of the antenna array. It is a block diagram which shows the structure of an abnormal-tissue detection apparatus. 4 is a graph showing an example of an impulse signal output from a transmitter; (A) is a schematic diagram showing a transmitting/receiving antenna and a breast when abnormal tissue is detected, and (B) is a schematic diagram showing the principle of abnormal tissue detection. 4 is a flowchart of abnormal tissue detection processing according to Embodiment 1. FIG. FIG. 10 is a diagram showing a state in which the abnormal tissue detection device is brought into contact with a subject, (A) is a diagram when the apparatus is in contact with the subject in the first orientation, and (B) is a diagram when the apparatus is in contact with the subject in the second orientation. . 1 is an xy cross-sectional view of confocal image data according to Embodiment 1, where (A) is first confocal image data, (B) is second confocal image data, and (C) is (B). is rotated by -90°, and (D) is the confocal image data obtained by the inner product of (A) and (C). 1 is a yz cross-sectional view of confocal image data according to Embodiment 1, where (A) is first confocal image data, (B) is second confocal image data, and (C) is (B). is rotated by -90°, and (D) is the confocal image data obtained by the inner product of (A) and (C). 10 is a flowchart of abnormal tissue detection processing according to Embodiment 2. FIG. FIG. 10 is a diagram of confocal image data according to Embodiment 2, where (A) is the first confocal image data, (B) is the second confocal image data rotated −90°, and (C) (A) is binarized, (D) is binarized (B), and (E) is confocal image data obtained by inner product of (C) and (D). 10 is a flowchart of abnormal tissue detection processing according to Embodiment 3. FIG. FIG. 10 is a diagram of confocal image data according to Embodiment 3, (A) is a diagram in which the first confocal image data is displaced, and (B) is a diagram in which the second confocal image data is rotated −90°. FIG. (C) is the confocal image data obtained by the inner product of (A) and (B). FIG. 10 is a diagram of confocal image data according to a modification of Embodiment 3, (A) is a diagram in which the first confocal image data is displaced, and (B) is a diagram in which the second confocal image data is displaced by −90°. Rotated view, (C) is a binarized view of (A), (D) is a binarized view of (B), and (E) is a confocal image of the inner product of (C) and (D). image data. FIG. 10 is a diagram of confocal image data according to a numerical example when a target exists, (A) is first confocal image data, (B) is second confocal image data, and (C) is an embodiment. 10 is confocal image data by the inner product of (A) and (B) according to Embodiment 1, and (D) is confocal image data by the inner product of (A) and (B) according to the third embodiment. FIG. 10 is a diagram of confocal image data according to a numerical example when there is no target, (A) is the first confocal image data, (B) is the second confocal image data, and (C) is the embodiment. 10 is confocal image data by the inner product of (A) and (B) according to Embodiment 1, and (D) is confocal image data by the inner product of (A) and (B) according to the third embodiment.

Hereinafter, the abnormal tissue detection method and the abnormal tissue detection device according to the embodiments of the present invention will be described in detail with reference to the drawings, taking an abnormal tissue detection device for detecting breast cancer as an example.

(Embodiment 1)
As shown in FIGS. 1A and 1B, the abnormal tissue detection device 1 according to the present embodiment includes a rotating portion 9 including a housing 57, an antenna portion 3, etc., a fixed base 31, a driving portion 33 and a handle 51 .

A driving portion 33 for rotating the rotating portion 9 is fixed to the fixed base 31 . The drive unit 33 is, for example, a stepping motor. A rotating portion of the driving portion 33 is fixed to the lid 52 . Thereby, the abnormal tissue detecting device 1 rotatably supports the rotating section 9 including the lid 52, the housing 57, the antenna section 3, and the like. A handle 51 is attached to the upper portion of the fixed base 31 . A person who performs the examination holds the handle 51 and brings the rotating part 9 into contact with the breast, which is the part of the subject to be examined, to perform the examination.

The rotating section 9 includes a lid 52 , a control board 55 , a high frequency board 56 , a housing 57 and an antenna section 3 .

The antenna section 3 includes an antenna base 37 , an antenna mounting section 39 , an antenna array 38 and an antenna cover 41 . The antenna base 37 is a mounting member for fixing the antenna mounting portion 39 to the housing 57 .

An antenna array 38 composed of antennas for transmitting and receiving microwave radio signals is attached to the antenna attachment portion 39 . The antenna mounting portion 39 is fixed to the antenna base 37 . The antenna mounting portion 39 is made of resin, for example, and has a dome-shaped central portion.

The antenna array 38 is composed of transmitting antennas SA1-SA8 and receiving antennas RA1-RA8. Each antenna constituting the antenna array 38 is attached to the lower surface of the antenna attachment portion 39, that is, the surface facing the dome-shaped portion of the antenna cover 41, which will be described later. Wiring and the like for connecting each antenna to the switch section 2, which will be described later, are arranged on the upper surface side of the antenna attachment section 39 (inside the housing 57). As shown in FIGS. 2A and 2B, the transmitting antennas SA1 to SA8 and the receiving antennas RA1 to RA8 are arranged linearly in the radial direction from the center of the antenna mounting portion 39, forming four antenna arrays A1. , an antenna array 38 consisting of A4.

The antenna cover 41 is arranged so as to cover the outer surface of the antenna mounting portion 39 and comes into contact with the subject's body. The antenna cover 41 is screwed and fixed to the housing 57 via the sleeve 42 . The central portion of the antenna cover 41 is formed in a dome shape concentric with the antenna mounting portion 39, and has a shape that facilitates close contact with the breast, which is the examination site. Also, a plurality of antenna mounting portions 39 and antenna covers 41 having different sizes may be prepared in advance so that one suitable for the breast size of the subject can be selected.

The space between the antenna mounting portion 39 and the antenna cover 41 is filled with a dielectric having a dielectric constant close to that of the subject's body and that of the antenna cover 41, such as vaseline or anhydrous glycerin. As a result, generation of unnecessary radiation between the antenna mounting portion 39 and the antenna cover 41 can be suppressed, so that the detection accuracy of abnormal tissue can be improved.

The abnormal tissue detection device 1 is provided with an orientation marker M indicating the direction of the rotating part 9 as a reference. More specifically, it is possible to visually recognize in which direction the rotating portion 9 including the antenna portion 3 is arranged.

The shape of the azimuth marker M is not particularly limited, and may be a notch, a protrusion, a print, or the like, as long as the angle of rotation of the antenna unit 3 can be visually recognized from the outside when the test subject is in contact with the subject. The azimuth marker M is, for example, a marker printed on the outer peripheral surface of the antenna cover 41 .

The control board 55 is a printed wiring board on which hardware circuits for realizing the functions of the control unit 14, the storage unit 15, and the drive control unit 34 shown in the block diagram of FIG. ing.

The drive control unit 34 is a motor driver that controls the rotation of the stepping motor that is the drive unit 33 . As shown in FIG. 3 , the drive control section 34 rotates the antenna section 3 including the antenna array 38 by rotating the drive section 33 according to the control signal from the control section 14 .

The control unit 14 performs overall control of the abnormal tissue detection apparatus 1, image processing, and the like, and includes a CPU (Central Processing Unit), a memory such as a ROM (Read Only Memory), a RAM (Random Access Memory), and an external storage device. , input/output I/O, a crystal oscillator, and the like. The control unit 14 realizes each function of the control unit 14 by reading various operation programs and data stored in the ROM of the control unit 14, the storage unit 15, etc. into the RAM and operating the CPU. As a result, the control unit 14 includes a confocal image generating unit 141 (not shown) that generates a confocal image showing the estimated position of the abnormal tissue, and an image processing unit 142 (not shown) that removes noise components from the generated confocal image. shown).

The storage unit 15 is a non-volatile memory such as flash memory or hard disk. As shown in FIG. 3, the control unit 14 transmits the received signal RS received from the receiving unit 12 to the storage unit 15 for storage. Also, the confocal image generation unit 141 of the control unit 14 reads out the reception signal RS stored in the storage unit 15, and generates a confocal image.

As shown in FIG. 3, the control unit 14 outputs a timing signal indicating the timing of outputting the transmission signal TS to the transmission unit 11 and outputs a control signal CS1 for the CMOS switch 17 by executing the program. The control section 14 outputs a control signal CS2 for the CMOS switch 19 and receives waveform data of the reception signal RS from the reception section 12 .

The high-frequency board 56 is a printed wiring board on which hardware circuits for realizing the functions of the transmitting section 11, the receiving section 12, and the switching section 2 shown in the block diagram of FIG. there is

The transmission unit 11 is a hardware circuit that outputs, for example, a transmission signal TS, which is an impulse electric signal shown in FIG. As shown in FIG. 4, the level of this electrical signal is an impulse signal that changes from a positive value to a negative value in a short period of time.

The switch unit 2 includes a CMOS switch 17 that selects a transmitting antenna that transmits microwave impulse signals from among the transmitting antennas SA1 to SA8, and a receiving antenna that receives scattered signals reflected by abnormal tissue from among the receiving antennas RA1 to RA8. and a CMOS switch 19 for selecting .

The CMOS switch 17 receives the transmission signal TS from the input terminal. The CMOS switch 17 switches the antenna that outputs the input transmission signal TS to one of the plurality of transmission antennas SA1, SA2, . . . , SA8 according to the control signal CS1.

The receiving section 12 is a hardware circuit that outputs the reception signal RS input from the CMOS switch 19 to the control section 14 . The confocal image generation unit 141 of the control unit 14 generates a confocal image showing the estimated position of the abnormal tissue based on the input reception signal RS.

As shown in the example of FIG. 3, when the transmission antenna SA2 is selected by the control signal CS1, the CMOS switch 17 switches to output the transmission signal TS to the transmission antenna SA2. When the reception antenna RA2 is selected by the control signal CS2, the CMOS switch 19 switches to output the reception signal RS received by the reception antenna RA2 to the reception section 12. FIG. In this case, the microwave impulse signal MW is transmitted and the reflected signal RW is received by a combination of the transmitting antenna SA2 and the receiving antenna RA2.

The control unit 14 controls the CMOS switches 17 and 19 by the control signals CS1 and CS2 to output the transmission signal TS to the CMOS switch 17 while switching the combination of the transmission antennas SA1 to SA8 and the reception antennas RA1 to RA8. Also, the control unit 14 inputs the received signal RS output from the receiving antennas RA1 to RA8 via the CMOS switch 19. FIG. The confocal image generation unit 141 of the control unit 14 generates a confocal image based on the input reception signal RS. The image processing unit 142 of the control unit 14 removes noise from the generated confocal image.

Next, the basic operation of abnormal tissue detection using microwaves in the abnormal tissue detection device 1 will be described. The abnormal tissue detection apparatus 1 according to the present embodiment transmits and receives impulse microwave radio signals, and detects abnormal tissue, ie, breast cancer, based on the results of the transmission and reception.

As shown in FIGS. 5A and 5B, the abnormal tissue detection apparatus 1 sets the rotation angle θ of the rotating portion 9, that is, the antenna portion 3 including the transmitting antenna SA1 and the receiving antenna RA1, to an initial angle θ ₁ (angle1). , the microwave impulse signal MW is radiated from the transmission antenna SA1. A part of the radiated microwave propagates into the living body. In general, abnormal tissue CA such as cancer tissue is known to have a dielectric constant approximately 5 to 10 times higher than that of normal living tissue. Therefore, when abnormal tissue CA is present, microwaves are reflected at the interface between regions having different dielectric constants, that is, at the surface of abnormal tissue CA, and are received by receiving antenna RA1.

Here, if the time from the radiation of the microwave impulse signal MW to the reception of the reflected wave by the receiving antenna RA1 is T ₁ [s], then T ₁ ·c (c: speed of light in the living body) is , is the travel distance of the microwave impulse signal MW.

Therefore, the abnormal tissue CA is located on the ellipse E1, with the transmission antenna SA1 and the reception antenna RA1 at the rotation angle _θ1 as focal points, where the sum of the distances from the transmission antenna SA1 and the reception antenna RA1 is _T1 · _c . It will be.

Subsequently, the drive control unit 34 rotates the rotating unit 9 including the transmitting antenna SA1 and the receiving antenna RA1 to set the rotation angle θ to θ ₂ (angle2). Then, the abnormal tissue detection apparatus 1 radiates a microwave impulse signal MW from the transmitting antenna SA1 and receives it from the receiving antenna RA1. Assuming that T ₂ [s] is the time from when the microwave impulse signal MW is radiated until the receiving antenna RA1 receives the reflected wave, T ₂ ·c is the travel distance of the microwave impulse signal MW.

Therefore, the abnormal tissue CA is located on an ellipse E2 at the rotation angle _θ2 , with the transmission antenna SA1 and the reception antenna RA1 as focal points, and the sum of the distances from the transmission antenna SA1 and the reception antenna RA1 being _T2 · _c . It will be.

Similar processing is performed while sequentially rotating the transmitting antenna SA1 and the receiving antenna RA1, and the intersection points of a plurality of ellipses E ₁ to E _N (N is a natural number; E ₃ and below are not shown), thereby obtaining the abnormal tissue CA. position can be obtained.

Further, the antenna for transmitting the impulse signal MW is switched to the transmitting antenna SA2, the transmitting antenna SA2 emits microwaves, the receiving antenna RA2 receives the microwaves, and similar processing is performed. After that, by sequentially switching the transmitting antennas, microwaves are emitted, reflected waves are received by the receiving antennas, and similar processing is performed, thereby making it possible to identify the position of the abnormal tissue CA more accurately.

In addition, in FIG. 5B of the above example, the two-dimensional case is explained for easy understanding, but the above-mentioned processing is actually performed in the three-dimensional case.

Next, abnormal tissue detection processing in the abnormal tissue detection device 1 according to this embodiment will be described based on the flowchart shown in FIG.

The control unit 14 sets the rotation angle θ of the rotation unit 9 including the antenna unit 3 to the initial rotation angle (θ=0°) (step S11).

The operator of the abnormal tissue detection device 1, such as a doctor or the subject, brings the abnormal tissue detection device 1 into contact with the test site of the subject (step S12). More specifically, the abnormal tissue detection device 1 is brought into contact so that the antenna cover 41 of the abnormal tissue detection device 1 covers the breast of the subject. At this time, as shown in FIG. 7(A), the reference direction of the abnormal tissue detection device 1 is set to the direction of 0° (first direction) with respect to the subject, that is, the direction marker M is positioned at the head of the subject. The abnormal tissue detection device 1 is installed so as to face the direction of the body. In the present embodiment, the first azimuth is the 0° azimuth, but is not particularly limited, and may be any azimuth in which the direction of the output result of abnormal tissue detection for the subject's body can be confirmed.

Also, petroleum jelly may be applied to the inspection site before the abnormal tissue detection device 1 is installed. As a result, it is possible to reduce the change in the dielectric constant in the path of the microwave, improve the adhesion between the antenna cover 41 and the subject's body, and suppress the generation of noise.

Subsequently, the abnormal tissue detection device 1 acquires transmission/reception data of microwaves (step S13). Specifically, the transmission antenna SA1 is caused to transmit the transmission signal TS, and the reception antenna RA1 is caused to receive the reception signal RS. The control unit 14 causes the storage unit 15 to store the reception signal RS received by the reception antenna RA1 as the input signal IS_1.

The control unit 14 rotates the rotating unit 9 by a preset angle by operating the driving unit 33 . In the present embodiment, the preset angle is 6°, and the rotating portion 9 rotates once in 60 rotations. After the rotation of the rotating part 9 is completed, the control part 14 causes the transmission antenna SA1 to transmit the transmission signal TS and the reception antenna RA1 to receive the reception signal RS. The control unit 14 causes the storage unit 15 to store the reception signal RS as the input signal IS_2.

The control unit 14 further rotates the rotating unit 9 to transmit and receive signals between the transmitting antenna SA1 and the receiving antenna RA1. The control unit 14 repeats the above-described rotation operation and signal transmission/reception until the rotation angle θ reaches 360°, and causes the storage unit 15 to store 60 input signals IS_n (n=1 to 60).

When the rotation angle θ reaches 360° and reception of the input signal IS_n is completed, the abnormal tissue detection device 1 sequentially switches between the transmission antenna SA and the reception antenna RA, and acquires transmission/reception data in the same manner as described above. Then, the abnormal tissue detection device 1 ends the measurement in the first direction and generates first confocal image data (step S14).

Specifically, the confocal image generation unit 141 reads the input signal IS_n related to each transmission/reception antenna from the storage unit 15 . The confocal image generating unit 141 calculates the length of the signal path from the input signal IS_n, and based on the positions of the transmitting antenna SA and the receiving antenna RA corresponding to each signal and the calculated length of the signal path, Calculate the reflection position of the signal. Thereby, the confocal image generator 141 estimates the position of the abnormal tissue and generates the confocal image data _I0 , which is the first confocal image data. The confocal image data _I0 is set as brightness data corresponding to the three-dimensional coordinates of the inspection site so that the brightness of the coordinate where the abnormal tissue is estimated to have a large value.

Subsequently, the control unit 14 sets the rotation angle θ of the rotation unit 9 including the antenna unit 3 to the initial rotation angle (θ=0°).

The operator separates the abnormal tissue detection device 1 from the subject's body and again installs the abnormal tissue detection device 1 so that the antenna cover 41 of the abnormal tissue detection device 1 covers the subject's breast. At this time, the abnormal tissue detection device 1 is brought into contact with the examination site of the subject so that the reference direction of the abnormal tissue detection device 1 is in a second direction different from the first direction (step S15). The second orientation is not particularly limited as long as it is an orientation different from the first orientation with respect to the subject. For example, as shown in FIG. 7A, when the direction in which the orientation marker M faces the subject's head is set as the first orientation, as shown in FIG. The direction in which the azimuth marker M faces the side of the subject may be taken as the second azimuth.

By separating the abnormal tissue detection device 1 from the subject's body and bringing it into contact again in a second orientation different from the first orientation, the gap between the antenna cover 41 and the subject's body, the position, size, etc. of the foreign matter are detected. will be different. Also, the examination site may be cleaned or reapplied with petroleum jelly before the abnormal tissue detecting device 1 is placed in the subject's examination site in the second orientation. As a result, the adhesion between the antenna cover 41 and the subject's body can be enhanced, making it difficult for air gaps, foreign matter, and the like to enter.

Subsequently, similarly to the measurement in the first azimuth, the abnormal tissue detecting device 1 acquires microwave transmission/reception data (step S16). Specifically, the transmission antenna SA1 is caused to transmit the transmission signal TS, and the reception antenna RA1 is caused to receive the reception signal RS. The control unit 14 causes the storage unit 15 to store the reception signal RS received by the reception antenna RA1 as the input signal IS_1.

The control unit 14 rotates the rotating unit 9 by a preset angle by operating the driving unit 33 . After the rotation of the rotating part 9 is completed, the control part 14 causes the transmission antenna SA1 to transmit the transmission signal TS and the reception antenna RA1 to receive the reception signal RS. The control unit 14 causes the storage unit 15 to store the reception signal RS as the input signal IS_2.

The control unit 14 further rotates the rotating unit 9 to transmit and receive signals between the transmitting antenna SA1 and the receiving antenna RA1. The control unit 14 repeats the above-described rotation operation and signal transmission/reception until the rotation angle θ reaches 360°, and causes the storage unit 15 to store 60 input signals IS_n (n=1 to 60).

When the rotation angle θ reaches 360° and reception of the input signal IS_n is completed, the abnormal tissue detection device 1 sequentially switches between the transmission antenna SA and the reception antenna RA, and acquires transmission/reception data in the same manner as described above. Then, the abnormal tissue detection apparatus 1 ends the measurement in the second orientation and generates second confocal image data (step S17).

Specifically, the confocal image generation unit 141 reads the input signal IS_n related to each transmission/reception antenna from the storage unit 15 . The confocal image generating unit 141 calculates the length of the signal path from the input signal IS_n, and based on the positions of the transmitting antenna SA and the receiving antenna RA corresponding to each signal and the calculated length of the signal path, Calculate the reflection position of the signal. Thereby, the confocal image generator 141 estimates the position of the abnormal tissue and generates confocal image data _I90 , which is the second confocal image data. The confocal image data I ₉₀ , like the confocal image data I ₀ , is set as brightness data corresponding to the three-dimensional coordinates of the inspection site so that the coordinates where abnormal tissue is estimated to have large values. .

Subsequently, the image processing unit 142 performs noise removal processing using the confocal image data _I0 , which is the first confocal image data, and the confocal image data _I90 , which is the second confocal image data. Generate confocal image data I _DC1 for estimating the location of the abnormal tissue. Specifically, by calculating the inner product of the confocal image data I ₀ and the confocal image data I ₉₀ , the coordinates at which abnormal tissue is presumed to exist in common in these confocal image data, that is, any image The data also emphasizes the coordinates with high brightness. In addition, coordinates where abnormal tissue is estimated to exist only in some image data, that is, coordinates where luminance is high only in some image data and are considered to be noise are suppressed.

The noise removal processing according to this embodiment will be described in detail below. Confocal image data I ₀ and confocal image data I ₉₀ are represented as follows, respectively.

In order to match the coordinates of the confocal image data I ₀ and the confocal image data I ₉₀ , confocal image data I ^T ₉₀ is generated by rotating the coordinates of the confocal image data I ₉₀ by −90° (step S18). . The confocal image data I ^T ₉₀ is represented by the following formula.

The image processing unit 142 generates the confocal image data I _DC1 by calculating the inner product of the confocal image data I ₀ and the confocal image data I ^T ₉₀ as shown in the following equation (step S19). . That is, the confocal image data I _DC1 is expressed as a set of products of luminance values of corresponding pixels of the confocal image data I ₀ and the confocal image data I ^T ₉₀ .

Since there is a correlation between the confocal image data I ₀ and the confocal image data I ^T ₉₀ at the coordinates where the abnormal tissue is estimated to exist, it is emphasized by calculating the inner product. Coordinates of random anomalous scattering signals due to air gaps, foreign matter, etc. are suppressed because they are not correlated with each other. As a result, it is possible to suppress the influence of noise such as air gaps and foreign matter existing between the antenna unit 3 that transmits and receives microwaves and the human body, and to detect abnormal tissue with high accuracy.

FIG. 8 is an example of a plan view (xy sectional view) of the confocal image according to the present embodiment, and FIG. 9 is an example of a sectional view (yz sectional view) of the confocal image according to the present embodiment. For example. 8A and 9A are confocal image data I ₀ related to the first confocal image data, and FIGS. 8B and 9B are confocal image data I 0 related to the second confocal image data. Focus image data I ₉₀ , FIGS. 8C and 9C are confocal image data I ^T ₉₀ obtained by rotating FIGS. 8B and 9B by −90°. 8D and 9D are confocal image data _IDC1 obtained by inner product of confocal image data _I0 and confocal image data _I90 . As shown in FIGS. 8 and 9, the area estimated to be abnormal tissue included in the first confocal image data and the second confocal image data by the noise removal processing using the inner product according to the present embodiment is emphasized, and a portion presumed to be noise contained only in either the first confocal image data or the second confocal image data is suppressed.

As described above in detail, according to the abnormal tissue detection method and the abnormal tissue detection device according to the present embodiment, the first confocal image data and the second confocal image data acquired in different orientations with respect to the subject's body and the confocal image data of , the overlapping state of the region presumed to be abnormal tissue is derived, and the position of the abnormal tissue is estimated. Therefore, by installing the abnormal tissue detection apparatus 1 in different orientations at the test site of the subject, the effects of noise such as voids and foreign substances whose state changes can be reduced, and the position of the abnormal tissue can be estimated with high accuracy. can be done.

Although two pieces of confocal image data are used in this embodiment, the present invention is not limited to this, and three or more pieces of confocal image data may be used. In this case, each piece of confocal image data is preferably generated based on transmission/reception data acquired in different azimuths. By acquiring transmission/reception data in different orientations, it is possible to generate confocal image data with different states of voids, foreign matter, and the like. Therefore, by sequentially calculating the inner product based on each confocal image data, noise contained in each confocal image data can be removed, and the coordinates where abnormal tissue is presumed to exist can be emphasized. The position of the abnormal tissue can be estimated with higher accuracy.

Further, in the present embodiment, after the first confocal image data is generated based on the transmission/reception data acquired in the first orientation, the transmission/reception data is acquired in the second orientation. can't For example, the first confocal image data may be generated during acquisition of transmission/reception data in the second orientation. As a result, the measurement time can be shortened, and the burden on the subject can be reduced. In addition, since the time required to output the confocal image data _IDC1 can be shortened, it is possible to efficiently detect abnormal tissue.

(Embodiment 2)
In Embodiment 1 above, the inner product of the first confocal image data and the second confocal image data generated based on the received signal RS is calculated. The accuracy of abnormal tissue estimation may be improved by processing the confocal image data and the second confocal image data. In this embodiment, a method of binarizing and using the first confocal image data and the second confocal image data will be described.

In the present embodiment, the method of processing confocal image data is different from that in Embodiment 1, and the configuration and the like of the abnormal tissue detection apparatus 1 are the same as those in Embodiment 1. are omitted.

Noise removal according to the present embodiment will be described below with reference to the flowchart of FIG. As in Embodiment 1, the abnormal tissue detection device 1 is installed at the test site of the subject in the first orientation (step S21), and while the antenna unit 3 is rotated, microwave transmission/reception data is acquired (step S22). ). Then, the confocal image generator 141 generates confocal image data _I0 , which is the first confocal image data, from the acquired transmission/reception data (step S23).

Subsequently, after the abnormal tissue detection device 1 is separated from the subject, the abnormal tissue detection device 1 is installed at the examination site of the subject in the second orientation (step S24), and while the antenna unit 3 is rotated, microwaves are transmitted and received. Data is acquired (step S25). Then, the confocal image generator 141 generates confocal image data _I90 , which is the second confocal image data, from the acquired transmission/reception data (step S26).

The image processing unit 142 performs noise removal processing using the generated confocal image data _I0 and confocal image data _I90 . Specifically, first, the confocal image data _I0 and the confocal image data _I90 are binarized using a predetermined threshold value δ (step S27). Confocal image data B(I ₀ ) obtained by binarizing the confocal image data I ₀ is represented by the following formula.

Since the orientations of the confocal image data I ₀ and the confocal image data I ₉₀ are different, the image processing unit 142 rotates the confocal image data I ₉₀ by −90° to align the orientations of both confocal image data, binarize. Specifically, the confocal image data I ₉₀ is rotated by -90° to generate confocal image data I ^T ₉₀ ={I ₉₀ (y, -x, z)}, and a predetermined Confocal image data B(I ^T ₉₀ ) binarized with a threshold value δ of is generated.

The image processing unit 142 generates confocal image data B(I ₀ ) based on the first confocal image data and confocal image data B(I ^T ₉₀ ) to generate the confocal image data I _DC2 (step S28).

Since there is a correlation between the confocal image data B(I ₀ ) and the confocal image data B(I ^T ₉₀ ) at the coordinates where the abnormal tissue is estimated to exist, it is emphasized by calculating the above inner product. . Coordinates of random anomalous scattering signals due to air gaps, foreign matter, etc. are suppressed because they are not correlated with each other. As a result, it is possible to suppress the influence of noise such as air gaps and foreign matter existing between the antenna unit 3 that transmits and receives microwaves and the human body, and to detect abnormal tissue with high accuracy.

FIG. 11 is an example of a confocal image according to this embodiment. FIG. 11(A) is the confocal image data I ₀ , FIG. 11(B) is the confocal image data I ^T ₉₀ obtained by rotating the confocal image data I ₉₀ by −90°, and FIGS. ) are confocal image data B(I ₀ ) and B( ^IT ₉₀ ) obtained by binarizing FIGS. 11A and 11B, respectively. FIG. 11(E) is the confocal image data I _DC2 obtained by the inner product of FIG. 11(C) and FIG. 11(D). In this example, it can be seen that there is no abnormal tissue and noise due to voids contained in the first confocal image data and the second confocal image data is removed.

As described above in detail, according to the abnormal tissue detection method according to the present embodiment, the first confocal image data and the second confocal image data acquired in different orientations with respect to the subject's body By binarizing and calculating the inner product, the overlapping state of the region of the abnormal tissue is derived, and the position of the abnormal tissue is estimated. Therefore, noise can be effectively removed, and the position of the abnormal tissue can be clearly estimated.

In the present embodiment, the first confocal image data and the second confocal image data are binarized, but the present invention is not limited to this. may be used. As a result, the noise removal effect can be enhanced while reducing the computational load due to the binarization process.

In the present embodiment, two confocal image data are used, but the present invention is not limited to this, and three or more confocal image data can be binarized to derive the overlapping state of the abnormal tissue region. . In this case, each piece of confocal image data is preferably generated based on transmission/reception data acquired in different azimuths. By acquiring transmission/reception data in different orientations, it is possible to generate confocal image data with different states of voids, foreign matter, and the like. Therefore, by sequentially calculating the inner product based on each confocal image data, noise contained in each confocal image data can be removed, and the coordinates where abnormal tissue is presumed to exist can be emphasized. The position of the abnormal tissue can be estimated with higher accuracy.

(Embodiment 3)
In Embodiment 1 above, the accuracy of abnormal tissue estimation is improved by calculating the inner product of the first confocal image data and the second confocal image data generated based on the received signal RS. . However, the coordinates of the first confocal image data and the second confocal image data may be shifted due to the difference in the installation position of the abnormal tissue detection device 1 . In this embodiment, a method for correcting the relative coordinates between the first confocal image data and the second confocal image data and improving the detection accuracy of abnormal tissue will be described.

In the present embodiment, the method of processing confocal image data is different from that in Embodiment 1, and the configuration and the like of the abnormal tissue detection apparatus 1 are the same as those in Embodiment 1. are omitted.

Abnormal tissue detection processing according to the present embodiment will be described below with reference to the flowchart of FIG. As in Embodiment 1, the abnormal tissue detection device 1 is installed at the test site of the subject in the first orientation (step S31), and while rotating the antenna unit 3, microwave transmission/reception data is acquired (step S32). ). Then, the confocal image generator 141 generates confocal image data _I0 , which is first confocal image data, from the acquired transmission/reception data (step S33).

Subsequently, after the abnormal tissue detection device 1 is separated from the subject, the abnormal tissue detection device 1 is placed at the examination site of the subject in the second orientation (step S34), and while the antenna unit 3 is rotated, microwaves are transmitted and received. Data is acquired (step S35). Then, the confocal image generator 141 generates confocal image data _I90 , which is the second confocal image data, from the acquired transmission/reception data (step S36).

Subsequently, the image processing unit 142 performs noise removal processing using the confocal image data _I0 and the confocal image data _I90 . Specifically, first, the image processing unit 142 calculates a correlation coefficient indicating the degree of similarity between the confocal image data _I0 and the confocal image data _I90 (step S37). The correlation coefficient r is as shown in the following formula.

The image processing unit 142 calculates the correlation coefficient r(i, j, k) while changing i, j, k. Then, the image processing unit 142 selects the position coordinates (i _max , j _max , k _max ) that maximize the correlation coefficient r(i, j, k). The image processing unit 142 uses (i _max , j _max , k _max ) to displace the first confocal image data according to the following formula (step S38).

Further, as in the first embodiment, the image processing unit 142 rotates the confocal image data I ₉₀ , which is the second confocal image data, by −90° to generate the confocal image data I ^T _{90. FIG} . Then, the image processing unit 142 calculates confocal image data T(I ₀ ) based on the first confocal image data and confocal image data I ^T based on the second confocal image data, as shown in the following equations. An inner product of ₉₀ is calculated to generate confocal image data _IDC3 (step S39).

FIG. 13 is an example of a confocal image according to this embodiment. FIG. 13A shows the confocal image data T(I ₀ ) obtained by displacing the confocal image data I ₀ , and FIG. 13B shows the confocal image data I obtained by rotating the confocal image data I ₉₀ by −90°. It's _a ^T90 . FIG. 13(C) is the confocal image data I _DC3 obtained by the inner product of FIG. 13(A) and FIG. 13(B). By the noise removal processing using the inner product according to the present embodiment, the portion presumed to be abnormal tissue included in the first confocal image data and the second confocal image data is emphasized, and the first confocal image It can be seen that the portion presumed to be noise contained only in either the data or the second confocal image data is suppressed.

As described above in detail, according to the abnormal tissue detection method according to the present embodiment, the first confocal image data and the second confocal image data acquired in different orientations with respect to the subject's body By calculating the inner product after correcting the deviation of the coordinates from , the overlapping state of the region presumed to be the abnormal tissue is derived, and the position of the abnormal tissue is estimated. Therefore, it is possible to more accurately suppress noise and estimate the position of the abnormal tissue.

In the present embodiment, the first confocal image data is displaced to correct the coordinate shift between the first confocal image data and the second confocal image data, but the present invention is not limited to this. The second confocal image data may be displaced.

Further, in the present embodiment, the inner product is calculated by displacing the coordinates of the first confocal image data and the second confocal image data generated based on the received signal RS. can't For example, after displacing the coordinates of the first confocal image data and the second confocal image data, binarization is performed in the same manner as in Embodiment 2, and the binarized confocal image data is used to obtain the inner product may be calculated. As a result, it is possible to efficiently remove unnecessary noise, suppress problems such as the abnormal tissue being estimated to be small, and accurately detect the abnormal tissue.

FIG. 14 is an example of a confocal image when binarization is performed after displacement. FIG. 14A shows the confocal image data T(I ₀ ) obtained by displacing the confocal image data I ₀ , and FIG. 14B shows the confocal image data I obtained by rotating the confocal image data I ₉₀ by −90°. It's _a ^T90 . FIGS. 14C and 14D are confocal image data B(T(I ₀ )) and B( ^IT ₉₀ ) obtained by binarizing FIGS. 14A and 14B, respectively. FIG. 14(E) is the confocal image data I _DC4 obtained by the inner product of FIG. 14(C) and FIG. 14(D). By the noise removal processing using the inner product according to the present embodiment, the portion presumed to be abnormal tissue included in the first confocal image data and the second confocal image data is more clearly emphasized, and the first It can be seen that the portion presumed to be noise contained only in either the confocal image data or the second confocal image data is suppressed.

Although two pieces of confocal image data are used in this embodiment, the present invention is not limited to this, and three or more pieces of confocal image data may be used. In this case, each piece of confocal image data is preferably generated based on transmission/reception data acquired in different azimuths. By acquiring transmission/reception data in different orientations, it is possible to generate confocal image data with different states of voids, foreign matter, and the like. Therefore, by sequentially calculating the inner product based on each confocal image data, noise contained in each confocal image data can be removed, and the coordinates where abnormal tissue is presumed to exist can be emphasized. The position of the abnormal tissue can be estimated with higher accuracy.

Further, in this case, one confocal image data out of a plurality of confocal image data is used as a reference, and the correlation coefficient of the other confocal image data with respect to the confocal image data selected as the reference is calculated. The other confocal image data may be displaced based on the correlation coefficient.

In each of the above-described embodiments, the abnormal tissue detection device 1 includes the confocal image generator 141 and the image processor 142, but the configuration is not limited to this. For example, an image processing device independent from the abnormal tissue detection device 1 may include the confocal image generator 141 and the image processor 142 . In this case, the image processing device may acquire data related to the received signal RS from the abnormal tissue detection device 1 via a network or the like, and perform processing such as generation of confocal images and noise removal. As a result, the image processing can be performed off-line separately from the data acquisition processing, so that the inspection results can be analyzed more efficiently.

Also, the abnormal tissue detection method related to the noise removal processing of each of the above embodiments can be implemented using a normal computer system. For example, by distributing a computer program for executing abnormal tissue detection according to each of the above embodiments via a network such as the Internet and installing the computer program on a computer, the computer device can be used for the above abnormal tissue detection. can function as an abnormal tissue detection device that performs

(Numerical example)
A numerical example using a phantom instead of a subject will be described below. In this example, transmission/reception data was obtained using a phantom imitating a breast and a metal target imitating an abnormal tissue (breast cancer).

FIG. 15 is an example of a confocal image when a target is present. FIG. 15A is the confocal image data I ₀ , and FIG. 15B is the confocal image data I ₉₀ . FIG. 15C shows the confocal image data I _DC1 generated by the inner product calculation according to the first embodiment, and FIG. 15D shows the confocal image data I DC1 generated by the inner product calculation according to the third embodiment. _DC3 . As shown in FIGS. 15C and 15D, in both cases of the confocal image data I _DC1 according to the first embodiment and the confocal image data I _DC3 according to the third embodiment, the original confocal image It can be seen that the portions presumed to be abnormal tissue are emphasized in comparison with the data I ₀ and I ₉₀ .

FIG. 16 is an example of a confocal image when there is no target. FIG. 16A is the confocal image data I ₀ , and FIG. 16B is the confocal image data I ₉₀ . FIG. 16C shows the confocal image data I _DC1 generated by the inner product calculation according to the first embodiment, and FIG. 16D shows the confocal image data I DC1 generated by the inner product calculation according to the third embodiment. _DC3 . As shown in FIGS. 16C and 16D, in both cases of the confocal image data I _DC1 according to the first embodiment and the confocal image data I _DC3 according to the third embodiment, the original confocal image It can be seen that the portion presumed to be noise is suppressed compared to the data I ₀ and I ₉₀ .

INDUSTRIAL APPLICABILITY The present invention is suitable for detecting abnormal tissue in the body of a subject. In particular, it is suitable for abnormal tissue detection, in which the abnormal tissue detection device is brought into contact with the subject and the position of the abnormal tissue is estimated using reflected microwaves.

1 Abnormal Tissue Detection Device 2 Switch Unit 3 Antenna Unit 9 Rotating Unit 11 Transmitting Unit 12 Receiving Unit 14 Control Unit 141 Confocal Image Generating Unit 142 Image Processing Unit 15 Storage Unit 17, 19 CMOS switch, 31 fixed base, 33 drive unit, 34 drive control unit, 37 antenna base, 38 antenna array, 39 antenna mounting unit, 41 antenna cover, 42 sleeve, 51 handle, 52 lid, 55 control board, 56 high frequency board, 57 housing, A1, A2, A3, A4 antenna array, CA abnormal tissue, CS1, CS2 control signal, M bearing marker, MW impulse signal, RW reflected signal, RA receiving antenna, SA transmitting antenna, TS transmitting signal, RS receiving signal

Claims (6)

An antenna unit having a transmitting antenna and a receiving antenna is brought into contact with a test site of a subject in a plurality of orientations, and based on reflected signals received by the receiving antenna, a plurality of confocal image data are generated,
estimating the position of the abnormal tissue based on the overlapping state of the regions of the abnormal tissue estimated in the plurality of confocal image data;
An abnormal tissue detection method characterized by: wherein the overlap state is derived by computing an inner product of the plurality of confocal image data;
The abnormal tissue detection method according to claim 1, characterized in that: The overlapping state is derived by binarizing the plurality of confocal image data and calculating an inner product,
The abnormal tissue detection method according to claim 1, characterized in that: calculating a correlation coefficient indicating the degree of similarity of the plurality of confocal image data;
Deriving the overlapping state by displacing at least one of the plurality of confocal image data so that the correlation coefficient is maximized and calculating an inner product of the plurality of confocal image data.
The abnormal tissue detection method according to claim 1, characterized in that: An antenna unit having a transmitting antenna for transmitting a microwave impulse signal, a receiving antenna for receiving the impulse signal, and an antenna cover for contacting a subject;
a drive unit that rotates the antenna unit;
a confocal image generating unit that generates confocal image data based on the reflected signal of the impulse signal received by the receiving antenna;
An image processing unit that removes noise components from the confocal image generated by the confocal image generation unit,
The image processing unit
Based on the overlapping state of the region of the abnormal tissue estimated by a plurality of confocal image data generated based on the received data acquired by bringing the antenna unit into contact with the subject in a plurality of orientations, the abnormal tissue region is determined. to estimate the position,
An abnormal tissue detection device characterized by: the computer,
Confocal image generation for generating a plurality of confocal image data based on received signals received by the receiving antenna by bringing an antenna unit having a transmitting antenna and a receiving antenna into contact with a test site of a subject in a plurality of orientations. part,
an image processing unit that estimates the position of the abnormal tissue based on the overlapping state of the regions of the abnormal tissue estimated by the plurality of confocal image data;
A program that acts as a

Images