Ultrasonic thickness measuring device and ultrasonic thickness measuring method

Abstract

To provide an ultrasonic thickness measurement device which can improve the estimation accuracy of the thickness of a muscle layer and a fat layer.SOLUTION: An ultrasonic thickness measurement device 100 comprises: an ultrasonic probe 60 which has an ultrasonic transmission element and 8 or less ultrasonic elements having reception elements; and a control unit 33 which obtains the thickness of a target body tissue from the tomography image data of the body of a subject obtained on the basis of a reception signal received by each reception element of each ultrasonic element. The control unit obtains the thickness from the tomographic image based on each reception signal received by each ultrasonic element.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to an ultrasonic thickness measuring device and an ultrasonic thickness measuring method for determining the thickness of a target body tissue from an ultrasonic tomographic image of a subject's body. [Background technology]

[0002] 2. Description of the Related Art Conventionally, there is known a measuring device that uses ultrasound to measure the thickness of muscle layers and fat layers in the body (Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 61-220634 Summary of the Invention [Problem to be solved by the invention]

[0004] In Patent Document 1, the thickness of the muscle layer and fat layer is estimated based on the A-mode image of the ultrasound signal, but the accuracy is insufficient. Increasing the number of ultrasound elements is considered as a method to improve the accuracy, but in this case, the amount of information processed by the processing unit that estimates the thickness increases, and the circuit system is required to have high processing power, which causes the device to become larger. In addition, the probe size also becomes larger. [Means for solving the problem]

[0005] In order to solve the above problems, the ultrasonic thickness measuring device of the present invention comprises an ultrasonic probe having eight or less ultrasonic elements each having an ultrasonic transmitting element and a receiving element, and a control unit that determines the thickness of a target body tissue from tomographic image data of the subject's body obtained based on a receiving signal received by each of the receiving elements of each of the ultrasonic elements, and the control unit determines the thickness from the tomographic image data based on the respective receiving signals received by each of the ultrasonic elements.

[0006] In addition, the ultrasonic thickness measurement method of the present invention is an ultrasonic thickness measurement method for determining the thickness of a target body tissue from tomographic image data of a subject's body obtained by receiving reflected waves of ultrasonic waves transmitted to the body from eight or less ultrasonic elements having ultrasonic transmitting elements and receiving elements, and receiving the reflected waves from the ultrasonic elements, characterized in that the thickness is determined from the tomographic image data based on the respective received signals received by the ultrasonic elements. [Brief description of the drawings]

[0007] [Figure 1] 1 is a schematic diagram showing the overall configuration of an ultrasonic thickness measurement device according to a first embodiment of the present invention. [Diagram 2] FIG. 2 is a schematic front view of the ultrasonic probe according to the first embodiment. [Diagram 3] FIG. 2 is a front view of the ultrasonic probe of the first embodiment attached to the body. [Figure 4] FIG. 2 is a diagram showing an A-mode image obtained by one ultrasonic element of the first embodiment and a corresponding B-mode image of one line. [Diagram 5] FIG. 1A shows input image data consisting of an 8-line B-mode image that is input to the control unit of embodiment 1 to determine the target thickness, and FIG. 1B shows output image data with the target thickness estimated and output from the control unit. [Figure 6] FIG. 2A shows input image data consisting of an 8-line B-mode image in the first embodiment, and FIG. 2B shows corresponding training data. [Figure 7] A diagram of input image data consisting of B-mode images obtained with an ultrasound probe having 64 or more ultrasound elements (A), and a diagram of the corresponding training data (B). [Figure 8] 4 is a schematic front view of an ultrasound probe attached to a body at a position different from that shown in FIG. 3. [Figure 9] A B-mode image obtained with an ultrasound probe attached at the position shown in Figure 8. [Figure 10] A systematic diagram of the learning model update. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] The present invention will first be briefly described below. In order to solve the above problem, an ultrasonic thickness measuring device according to a first aspect of the present invention comprises an ultrasonic probe having eight or less ultrasonic elements each having an ultrasonic transmitting element and a receiving element, and a control unit that determines the thickness of a target body tissue from tomographic image data of the subject's body obtained based on a receiving signal received by each of the receiving elements of each of the ultrasonic elements, and the control unit determines the thickness from the tomographic image data based on the respective receiving signals received by each of the ultrasonic elements. Here, "thickness" is not limited to the thickness of muscle layers or fat layers of the body, but also includes thickness for identifying organs having a layered structure, thickness for identifying abnormal areas, and the like.

[0009] According to this aspect, the control unit obtains the thickness from the tomographic image data based on the reception signals received by the eight or less ultrasonic elements. The tomographic image data based on the reception signals received by the eight or less ultrasonic elements is tomographic image data in a state in which eight or less lines of tomographic image data are arranged in a depth direction from the skin surface. That is, the amount of information processed by the control unit to obtain the thickness is small because it is an arrangement of each of the 8 lines or less of tomographic image data. As a result, the circuit part constituting the control unit is not required to have high processing power, and the device can be made smaller. In addition, the size of the ultrasonic probe can be reduced.

[0010] The ultrasonic thickness measuring device according to a second aspect of the present invention is characterized in that the ultrasonic probe comprises a base member that is attachable to a body, and eight or less of the ultrasonic elements are arranged on the base member.

[0011] According to this aspect, the number of ultrasonic elements included in the ultrasonic probe is eight or less, so that the size of the ultrasonic probe can be reduced. In addition, the ultrasonic probe can be attached to the body via the base member, so that the ultrasonic probe is easy to use.

[0012] The ultrasonic thickness measuring device according to the third aspect of the present invention is characterized in that in the first or second aspect, the input tomographic image data used by the control unit to calculate the thickness is each B-mode image generated from each A-mode image generated by processing each received signal received by the eight or less ultrasonic elements, and arranged in correspondence with the arrangement of each of the eight or less ultrasonic elements.

[0013] Each B-mode image generated from each A-mode image is an image in which the brightness changes in the depth direction from the skin, but the brightness is constant in the width direction of each line, which is the direction intersecting the depth direction. In other words, a tomographic image in which each B-mode image of 8 lines or less is arranged is an image in which the brightness appears to change in a block shape in the depth direction for each line. According to this aspect, the input tomographic image data used by the control unit to calculate the thickness is tomographic image data in a state in which each B-mode image of 8 lines or less is arranged, and the brightness of each line changes in a block shape in the depth direction. As a result, the amount of information processed by the control unit to calculate the thickness is small because it is the tomographic image data that changes in a block shape. Therefore, the circuit part constituting the control unit is not required to have high processing power, and the device can be made smaller.

[0014] The ultrasonic thickness measuring device according to the fourth aspect of the present invention, in any one of the first to third aspects, is characterized in that the control unit is capable of accessing a learning model having parameters for obtaining attribute information corresponding to the thickness of the target body tissue from input tomographic image data, and using the tomographic image data of the subject's body as input image data, performs a process to obtain attribute information corresponding to the thickness using the learning model, and obtains the thickness from the obtained attribute information.

[0015] According to this aspect, the control unit performs a process of obtaining attribute information corresponding to the thickness using the learning model with the tomographic image data of the subject's body as input image data, and obtains the thickness from the obtained attribute information. This makes it possible to automatically and accurately obtain the thickness of the target body tissue.

[0016] The ultrasonic thickness measuring device according to the fifth aspect of the present invention is characterized in that in the fourth aspect, the learning model is configured such that, as one set of data, the learning model performs learning on a plurality of sets of data, the sets of data being tomographic image data based on respective reception signals received from the target body tissue by eight or less of the ultrasonic elements, and teacher image data corresponding to the tomographic image data and in a state in which the thickness can be calculated, and parameters are set to obtain attribute information corresponding to the thickness of the target body tissue from the input tomographic image data.

[0017] According to this aspect, the parameters of the learning model used to obtain attribute information corresponding to the thickness of the target body tissue from the input tomographic image data are set by learning the tomographic image data in a state in which each B-mode image of 8 lines or less is arranged, similar to the input tomographic image data. Therefore, the parameters of the learning model can be set using the tomographic image data with a small amount of information as described above.

[0018] The ultrasonic thickness measuring device of the sixth aspect of the present invention is characterized in that, in the fourth or fifth aspect, the control unit is connected to the learning model via a wireless or wired communication unit, and the learning model is capable of updating the parameters using new tomographic image data and teacher image data corresponding to the new tomographic image data.

[0019] According to this aspect, the learning model can update the parameters by new tomographic image data and teacher image data corresponding to the new tomographic image data, thereby making it possible to accurately obtain attribute information corresponding to the thickness of a target body tissue from input tomographic image data.

[0020] An ultrasonic thickness measurement method according to a seventh aspect of the present invention is an ultrasonic thickness measurement method for determining the thickness of a target body tissue from tomographic image data of a subject's body obtained by receiving reflected waves of ultrasonic waves transmitted to the body by eight or less ultrasonic elements having ultrasonic transmitting elements and receiving elements, characterized in that the thickness is determined from the tomographic image data based on each received signal received by the eight or less ultrasonic elements. According to this aspect, it is possible to obtain the same effects as those of the first aspect.

[0021] An ultrasonic thickness measuring device according to an eighth aspect of the present invention is characterized in that, in the seventh aspect, a learning model having parameters for obtaining attribute information corresponding to the thickness of the target body tissue from input tomographic image data is accessed, and using the tomographic image data of the subject's body as an input image, a process is performed using the learning model to obtain attribute information corresponding to the thickness, and the thickness is obtained from the obtained attribute information. According to this aspect, it is possible to obtain the same effects as those of the fourth aspect.

[0022] [Embodiment 1] Hereinafter, an ultrasonic thickness measurement device according to a first embodiment of the present invention will be described in detail with reference to FIGS. 1, the ultrasonic thickness measurement device 100 includes an ultrasonic probe 60 having ultrasonic elements that transmit and receive ultrasonic waves, and a control unit 33 that controls the functions of the ultrasonic probe 60 and performs signal processing. The control unit 33 is accommodated in a housing unit 10 (FIG. 3).

[0023] <Ultrasonic probe> As shown in Fig. 1 and Fig. 2, in this embodiment, the ultrasonic probe 60 includes eight ultrasonic elements 1, 2, 3, 4, 5, 6, 7, and 8. In Fig. 1, the illustration of six ultrasonic elements 2, 3, 4, 5, 6, and 7 other than the ultrasonic element 1 and the ultrasonic element 8 is omitted. As shown in Fig. 1, the ultrasonic element 1 has an ultrasonic transmission element 11 and a reception element 12, and the ultrasonic element 8 has an ultrasonic transmission element 81 and a reception element 82. As shown in Fig. 2, the other six ultrasonic elements 2, ultrasonic element 3, ultrasonic element 4, ultrasonic element 5, ultrasonic element 6, and ultrasonic element 7 similarly have ultrasonic transmission elements 21, 31, 41, 51, 61, and 71 and reception elements 22, 32, 42, 52, 62, and 72, respectively. Here, "having an ultrasonic transmitting element and a receiving element" is a functional description, and structurally, one ultrasonic element has the function of a "transmitting element" and the function of a "receiving element." In Fig. 2, reference numeral 9 denotes wiring.

[0024] As shown in FIG. 2, the ultrasonic probe 60 includes a flat base member 14 (FIG. 2) that can be attached to the body, and eight ultrasonic elements 1, 2, 3, 4, 5, 6, 7, and 8 that are arranged on the base member 14 in a row at equal intervals. As shown in Fig. 3, in this embodiment, the housing unit 10 that houses the ultrasonic probe 60 and the control unit 33 of the ultrasonic thickness measuring device 100 is attached to the belt 13. That is, the ultrasonic probe 60 and the control unit 33 are configured to be wrapped around the abdomen of the body and fixed by the belt 13. Specifically, as shown in Fig. 3, the belt 13 is configured to be attached to a target part of the body so that the thickness of a target muscle layer or the like can be measured.

[0025] <Control Unit> The control unit 33 obtains the thickness of the target body tissue from the tomographic image data (FIG. 5) of the subject's body obtained based on the received signals sequentially received by each receiving element 12, 22, 32, 42, 52, 62, 72, 82 of each ultrasonic element 1, 2, 3, 4, 5, 6, 7, 8. That is, the control unit 33 obtains the thickness from data of a tomographic image 30 (FIGS. 4 and 5) described later based on the reception signals sequentially received by the eight ultrasonic elements 1, 2, 3, 4, 5, 6, 7, and 8.

[0026] 1, the control unit 33 includes a drive pulse generating circuit 15, a transmitting circuit 16, a signal processing circuit 17, a receiving circuit 18, a multiplexer 19, a microcomputer 20, and a communication unit 23. The control unit 33 is configured to be able to access the tablet 25 via the communication unit 23. The tablet 25 is equipped with a learning model 27, which will be described later. When transmitting ultrasonic waves, a pattern of a predetermined drive frequency and wave number is generated in the drive pulse generating circuit 15, a transmission waveform of a predetermined drive voltage is output in the transmission circuit 16, and ultrasonic waves are transmitted from each transmitting element 11, 21, 31, 41, 51, 61, 71, 81 of each ultrasonic element 1, 2, 3, 4, 5, 6, 7, 8. When receiving ultrasonic waves, the reception signals by the reception elements 12, 22, 32, 42, 52, 62, 72, 82 of the ultrasonic elements 1, 2, 3, 4, 5, 6, 7, 8 are amplified by the reception circuit 18 and subjected to envelope processing and LOG compression processing by the signal processing circuit 17 to generate an A-mode image 24 which is a tomographic image 30 of each of the eight ultrasonic elements 1, 2, 3, 4, 5, 6, 7, 8. Fig. 4 shows the A-mode image 24 generated from the reception signal of one ultrasonic element. The control of each operation of each circuit 15, 16, 17, 18 is executed via a microcomputer 20. Furthermore, the microcomputer 20 is configured to sequentially switch the transmission and reception operations of each of the eight ultrasonic elements 1, 2, 3, 4, 5, 6, 7, 8 using a multiplexer 19. That is, the eight ultrasonic elements 1, 2, 3, 4, 5, 6, 7, 8 are configured to sequentially receive each reception signal.

[0027] In this embodiment, under the control of the microcomputer 20, data of eight A-mode images 24 of the eight ultrasonic elements 1, 2, 3, 4, 5, 6, 7, and 8 is sent to a tablet 25 equipped with a learning model 27 via a communication unit 23. The communication unit 23 is configured with a wireless LAN circuit in this embodiment, but may be configured with a wired connection. The tablet 25 is equipped with a GPU, and generates data of eight B-mode images 26 from data of the eight sent A-mode images 24. In FIG. 4, one B-mode image 26 corresponding to one A-mode image 24 is shown.

[0028] The data of the input tomographic image 30 used by the control unit 33 to calculate the thickness is obtained by arranging each B-mode image 26 generated from each A-mode image 24 by processing the received signals sequentially received by each of the eight ultrasonic elements 1, 2, 3, 4, 5, 6, 7, 8 in accordance with the arrangement of each of the eight ultrasonic elements 1, 2, 3, 4, 5, 6, 7, 8. 5A shows eight B-mode images 26 arranged in accordance with the arrangement of the eight ultrasonic elements 1, 2, 3, 4, 5, 6, 7, and 8. That is, this is eight lines of B-mode image 26 data in a state in which eight lines of tomographic image data are arranged along the depth direction (downward in the figure) from the skin surface 36. The data of these eight lines of B-mode image 26 is input image data used to calculate the thickness.

[0029] As described above, the control unit 33 can access the learning model 27 having parameters for obtaining attribute information corresponding to the thickness of the target body tissue from the input data of the tomographic image 30. The control unit 33 is configured to use, as input image data, data of an 8-line B-mode image 26 (FIG. 5A), which is data of a tomographic image 30 of the subject's body, to perform processing to obtain attribute information corresponding to the thickness using a learning model 27 mounted on the tablet 25, and to obtain the thickness from the obtained attribute information. The obtained thickness is displayed as output image data on the monitor 28 of the tablet 25, as described later (FIG. 5B). The learning model 27 may be installed in a computer such as a notebook computer instead of the tablet 25.

[0030] Here, "attribute information corresponding to thickness" refers to, for example, information on the position of the boundary between a fat layer and a muscle layer, which is obtained by utilizing the fact that transmitted ultrasound is largely reflected at the boundary between the fat layer and the muscle layer, and is hardly reflected in the muscle layer or the fat layer. If one surface of a muscle layer is in contact with a fat layer and the other surface is in contact with another fat layer, the positions of "one boundary" corresponding to one surface of the muscle layer and "the other boundary" corresponding to the other surface of the muscle layer can be obtained by ultrasound. The distance between the "one boundary" and the "other boundary" is the thickness of the muscle layer that is being sought. Alternatively, when determining the thickness of subcutaneous fat, the position of the boundary between the subcutaneous fat and the muscle layer beneath it can be determined using ultrasound as "attribute information," and the distance between the determined boundary and the skin surface becomes the thickness of the subcutaneous fat. 5B shows output image data displayed on the monitor 28 of the tablet 25. In this embodiment, image data is displayed in which the muscle layer is outlined and the muscle layer thickness 34 and subcutaneous fat thickness 35 can be seen.

[0031] <Learning model parameters> In this embodiment, the learning model 27 is constructed as follows. As shown in Fig. 6, eight lines of B-mode images 26 (Fig. 6(A)) which are data of a tomographic image 30 based on the respective reception signals sequentially received from the body tissue of a specified part by eight ultrasonic elements 1, 2, 3, 4, 5, 6, 7, 8, and teacher image data 50 (Fig. 6(B)) in a state in which the thickness corresponding to the eight lines of B-mode images 26 is obtained are treated as one set of data, and multiple sets, for example, 100 sets or more, are prepared. Then, machine learning is performed on the multiple sets of data prepared, and parameters are set for obtaining attribute information corresponding to the thickness of the target body tissue from the input image data. One example of a method for generating a learning model is the convolution processing method of U-Net, which is aimed at region segmentation of medical images. Specifically, the 8-line B-mode image 26 in FIG. 6(A) is subjected to normal convolution processing in a machine learning neural network and further to pooling processing. The output image data from this processing is compared with the teacher image data 50 in FIG. 6(B), and the parameters are set by performing calculations so as to reduce the error, that is, by using the cross-entropy error function and gradient descent method. This setting is performed for 100 or more sets to optimize the parameters. In this embodiment, as shown in Fig. 1, the tablet 25 is connected to the learning model generation center 37 via a communication line. The parameters are set in the learning model generation center 37, and each time the estimation accuracy is improved, the result is transmitted to the tablet 25 and the learning model 27 is updated.

[0032] The learning model 27 may be constructed as follows. FIG. 7 shows (A) input image data consisting of a B-mode tomographic image 30 obtained from 64 or more received signals received by all ultrasonic elements using an ultrasonic probe having 64 or more ultrasonic elements, and (B) the corresponding teacher image data 50. From the data of the tomographic image 30 in Fig. 7(A) and the teacher image data 50 in Fig. 7(B), 8 lines of tomographic image data and teacher data are cut out at the sensor arrangement interval, and 100 or more sets of such data may be prepared, and machine learning similar to that described above may be performed to set parameters for obtaining attribute information corresponding to the thickness of the target body tissue from the input image data. B-mode images using 64 or more ultrasonic elements have the advantage that muscle layers can be easily distinguished and teacher data can be easily created.

[0033] In addition, in this embodiment, the control unit 33 is connected to the learning model 27 via the wireless communication unit 23, and the learning model 27 can update parameters using data of a new tomographic image 30 corresponding to (A) in Figure 6 and teacher image data 50 corresponding to (B) in Figure 6 that corresponds to the data of the new tomographic image 30. That is, first, new tomographic image 30 data for tissue at a specific part of the body is obtained by ultrasound, and the thickness of the target body tissue is calculated. In addition, the new tomographic image 30 data is sent from the tablet 25 to the learning model generation center 37, and the parameters of the learning model 27 are updated from a set of teacher image data 50 corresponding to the image. If the estimation accuracy is improved, the result is sent to the tablet 25, and the parameters of the learning model 27 are updated.

[0034] <Explanation of the Function of the First Embodiment> Next, a step of determining the thickness of the target body tissue by the ultrasonic thickness measuring device 100 of the first embodiment is performed. First, the ultrasonic probe 60 and the control unit 33 of the ultrasonic thickness measuring device 100 of this embodiment are wrapped around the abdomen of the subject and fixed with the belt 13 as shown in Fig. 3. Then, ultrasonic waves are transmitted and received sequentially by the eight ultrasonic elements 1, 2, 3, 4, 5, 6, 7, and 8, and eight A-mode images 24 are generated based on the eight received signals. Furthermore, eight B-mode images 26 are generated from the eight A-mode images 24 by the tablet 25, that is, data of eight lines of the B-mode image 26 (Fig. 5(A)) is generated. Next, the data of the 8-line B-mode image 26 (FIG. 5(A)) is used as input image data, and a learning model 27 is used to perform a process of determining attribute information corresponding to the thickness. The thickness is then determined from the determined attribute information (FIG. 5(B)), and this is displayed on a monitor 28. The thickness information displayed on the monitor 28 is not limited to the form shown in FIG. 5B, but may be in the form of, for example, numerical values ​​or a comparable graph.

[0035] <Explanation of Effects of First Embodiment> (1) According to this embodiment, the control unit 33 obtains the thickness of the target body tissue from a tomographic image based on each of the reception signals sequentially received by the eight ultrasonic elements 1, 2, 3, 4, 5, 6, 7, and 8. The tomographic image data based on each of the reception signals sequentially received by the eight ultrasonic elements 1, 2, 3, 4, 5, 6, 7, and 8 is tomographic image data ((A) in FIG. 5) in which data of each of the tomographic images 30 of eight lines along the depth direction from the skin surface 36 is arranged. That is, the amount of information processed by the control unit 33 to obtain the thickness is an arrangement of data of each of the eight lines of the tomographic images 30 (FIG. 5(A)), and therefore the amount of information is small. As a result, the circuit part constituting the control unit 33 is not required to have high processing power, and the device 100 can be made compact. Also, the size of the ultrasonic probe 60 can be reduced.

[0036] (2) In the present embodiment, the ultrasonic probe 60 includes eight ultrasonic elements, so that the size of the ultrasonic probe 60 can be made smaller than that of the conventional ultrasonic probe 60. In addition, the ultrasonic probe 60 can be attached to the body via the base member 14, so that the ultrasonic probe 60 is easy to use. (3) As shown in Fig. 4, the B-mode image 26 generated from the A-mode image 24 is an image in which the brightness changes on one line in the depth direction from the skin surface 36, but the brightness is constant in the width direction of the line, which is the direction intersecting the depth direction. In other words, the brightness changes in a block shape in the depth direction on the one line. As shown in Fig. 5(A), a tomographic image 30 in which eight lines of B-mode images are arranged is an image in which the brightness appears to change in a block shape in the depth direction for each line. According to this embodiment, the data of the tomographic image 30 inputted and used by the control unit 33 to calculate the thickness is tomographic image data (FIG. 5A) in which eight lines of B-mode images 26 are arranged, and the brightness of each line changes in a block shape in the depth direction. As a result, the amount of information processed by the control unit 33 to calculate the thickness is small because it is the data of the tomographic image 30 that changes in a block shape. Therefore, the circuit part constituting the control unit 33 is not required to have high processing power, and the device 100 can be made smaller.

[0037] (4) In this embodiment, the control unit 33 uses data of the tomographic image 30 of the subject's body as input image data, performs processing to obtain attribute information corresponding to the thickness using the learning model 27, and obtains the thickness from the obtained attribute information. This makes it possible to automatically obtain the thickness of the target body tissue with high accuracy. (5) In this embodiment, the parameters of the learning model 27 used to obtain attribute information corresponding to the thickness of the target body tissue from the input data of the tomographic image 30 are set by learning the tomographic image data ((A) in FIG. 6) in which each of eight lines of B-mode images is arranged, similar to the input data of the tomographic image 30. Therefore, the parameters of the learning model 27 can be set using the data of the tomographic image 30 with a small amount of information as described above. (6) In addition, in this embodiment, the learning model 27 can update the parameters using data of a new tomographic image 30 (FIG. 6A) and teacher image data (FIG. 6B) corresponding to the new tomographic image 30 data. This allows attribute information corresponding to the thickness of a target body tissue to be obtained with high accuracy from the data of the input tomographic image 30.

[0038] [Embodiment 2] Next, an ultrasonic thickness measurement device 100 according to a second embodiment of the present invention will be described with reference to FIGS. The ultrasonic thickness measuring device 100 of this embodiment is for determining visceral fat thickness, and as shown in FIG. 8, the ultrasonic probe 60 and the control unit 33 are attached to the ulnar groove of the body when used. In FIG. 8, the belt 13 is not shown. FIG. 9 is a tomographic image 30 of the ulnar groove of the body. The tomographic image 30 in FIG. 9 is a B-mode image obtained from 64 or more received signals received by all ultrasonic elements using an ultrasonic probe having 64 or more ultrasonic elements. The ulnar groove of the body does not have a rectus abdominis muscle 39, and an image for subcutaneous fat thickness 35 and an image for visceral fat thickness 38 can be obtained. Therefore, by measuring the epigastric area using the ultrasonic thickness measuring device 100 of this embodiment, it is possible to create a learning model 27 having parameters for calculating visceral fat thickness in the same manner as in embodiment 1, and the visceral fat thickness of the subject can be calculated using the learning model 27 in the same manner as above.

[0039] [Embodiment 3] Next, a third embodiment of the present invention will be described with reference to FIG. FIG. 10 shows four ultrasonic thickness measuring devices being used individually by user A, user B, user C, and user D, and connected wirelessly or via wires to a learning model generation center 37. User A and User D are both individual users and use the ultrasonic thickness measuring device 100 of this embodiment. User B is a training gym and uses the ultrasonic thickness measuring device 200 to visually check abdominal muscles and fat thickness using a portable echo probe 43 and a laptop computer 44 having a learning model. User C is a health and medical institution and uses an ultrasonic thickness measuring device 300 consisting of an ultrasonic diagnostic device connected to a normal ultrasonic probe 45.

[0040] Data of tomographic images 30 acquired by each of users A, B, C, and D is sent to a learning model generation center 37. The learning model generation center 37 identifies the body tissues, i.e., muscle layers and fat layers, that are the subject of the collected data of each tomographic image 30, creates teacher image data 50 (FIG. 6(B) and FIG. 7(B)), and generates a learning model 27 and parameters for the identified body tissues. This is performed for the data of each transmitted tomographic image 30 to improve the accuracy of the learning model 27 and parameters. It can be said that the more data of the tomographic images 30 that are collected, the more accurate the learning model 27 becomes. Every time the parameters of the learning model 27 are updated and the accuracy is improved, the latest parameters are sent from the learning model generation center 37 to the tablet 25 or notebook computer on which the learning model 27 is installed, and the learning model 27 is updated.

[0041] Other Embodiments The ultrasonic thickness gage 100 according to the embodiment of the present invention is based on the configuration described above, but it is of course possible to modify or omit some of the configuration without departing from the gist of the present invention. (1) In the above embodiment, the ultrasonic probe 60 is described as having eight ultrasonic elements 1, 2, 3, 4, 5, 6, 7, and 8, but the number is not limited to eight. Four ultrasonic elements, or less than eight, may be used to arrange four lines of tomographic image data. The number of ultrasonic elements may be more than eight, such as 16 or 32, as long as reception is performed by eight or less ultrasonic elements among them and the number of received signals is eight or less. (2) In the above embodiment, the ultrasonic probe 60 and the control unit 33 are configured to be wrapped around the abdomen of the body and fixed with the belt 13, but it is needless to say that this is not limited to this. The ultrasonic probe 60 may be a commonly used ultrasonic echo probe. (3) In the above embodiment, a structure in which the control unit 33 and the learning model 27 are separate has been described. However, a configuration in which the learning model 27 and the communication unit 23 are provided within the control unit 33 may also be used. [Explanation of symbols]

[0042] 1, 2, 3, 4, 5, 6, 7, 8 ultrasonic element, 9 wiring, 10 housing part, 11 transmitting element, 12 receiving element, 13 belt, 14 base member, 15 drive pulse generating circuit, 16 transmission circuit, 17 signal processing circuit, 18 receiving circuit, 19 multiplexer, 20 microcontroller, 21 transmitting element, 22 receiving element, 23 communication unit, 24 A-mode image, 25 Tablet, 26 B-mode image, 27 Learning model, 28 Monitor, 30 tomographic image, 31 transmitting element, 32 receiving element, 33 control unit, 34 muscle layer thickness, 35 Subcutaneous fat thickness, 36 Skin surface, 37 Learning model generation center, 38 Visceral fat thickness, 39 Rectus abdominis muscle, 41 Transmitting element, 42 Receiving element, 43 Portable echo probe, 44 Laptop, 45 normal ultrasound probe, 50 teacher image data, 51 transmitting element, 52 receiving element, 60 ultrasonic probe, 61 transmitting element, 62 receiving element, 71 transmitting element, 72 receiving element, 81 transmitting element, 82 receiving element, 100 Ultrasonic thickness measuring device

Claims

1. An ultrasonic probe having eight or less ultrasonic elements each having an ultrasonic transmission element and a receiving element; a control unit for determining a thickness of a target body tissue from tomographic image data of the subject's body obtained based on a reception signal received by each of the reception elements of each of the ultrasonic elements; The control unit is data of each B-mode image generated from each A-mode image generated by processing each reception signal sequentially received by each of the ultrasonic elements is arranged in accordance with the arrangement of each of the ultrasonic elements; The thickness is obtained from the B-mode data arranged in the arrangement. An ultrasonic thickness measuring device characterized by:

2. 2. The ultrasonic thickness measurement device according to claim 1, The ultrasonic probe includes: A base member that can be attached to the body; Eight or less of the ultrasonic elements are disposed on the base member. An ultrasonic thickness measuring device characterized by:

3. 3. The ultrasonic thickness measuring device according to claim 2, The eight or less ultrasonic elements are arranged at equal intervals in a row on the base member, The control unit is data of each B-mode image generated from each A-mode image generated by processing each reception signal sequentially received by each of the ultrasonic elements is arranged in accordance with the arrangement of each of the ultrasonic elements; The thickness is obtained from the B-mode data arranged in the arrangement. An ultrasonic thickness measuring device characterized by:

4. 4. The ultrasonic thickness measuring device according to claim 1, The control unit is A learning model having parameters for obtaining attribute information corresponding to the thickness of the target body tissue from input tomographic image data is accessible; using the tomographic image data of the subject's body as input image data, performing a process of obtaining attribute information corresponding to the thickness using the learning model, and obtaining the thickness from the obtained attribute information; An ultrasonic thickness measuring device characterized by:

5. 5. The ultrasonic thickness measuring device according to claim 4, The learning model is tomographic image data based on respective reception signals received from the target body tissue by eight or less of the ultrasonic elements; Teacher image data corresponding to the tomographic image data and in a state in which the thickness can be obtained; a set of data is used as one set of data, and learning is performed on a plurality of sets of data to set parameters for obtaining attribute information corresponding to the thickness of the target body tissue from the input tomographic image data. An ultrasonic thickness measuring device characterized by:

6. 6. The ultrasonic thickness measuring device according to claim 4, The control unit is connected to the learning model via a wireless or wired communication unit, The learning model is capable of updating the parameters using new tomographic image data and teacher image data corresponding to the new tomographic image data. An ultrasonic thickness measuring device characterized by:

7. An ultrasonic thickness measurement method for determining a thickness of a target body tissue from tomographic image data of a subject's body obtained by receiving reflected waves of ultrasonic waves transmitted to a body from eight or less ultrasonic elements having ultrasonic transmitting elements and receiving elements, the method comprising: data of each B-mode image generated from each A-mode image generated by processing each reception signal sequentially received by each of the ultrasonic elements is arranged in correspondence with the arrangement of the ultrasonic elements, and the thickness is obtained from the data of the B-mode arranged in the arrangement; The ultrasonic thickness measurement method according to the present invention is characterized in that

8. The ultrasonic thickness measurement method according to claim 7, accessing a learning model having parameters for obtaining attribute information corresponding to the thickness of the target body tissue from the input tomographic image data; using the tomographic image data of the subject's body as an input image, performing a process of obtaining attribute information corresponding to the thickness using the learning model, and obtaining the thickness from the obtained attribute information; The ultrasonic thickness measurement method according to the present invention is characterized in that

Images