Ultrasound diagnostic device and method for controlling the ultrasound diagnostic device

Abstract

To provide a diagnostic apparatus and a control method of the diagnostic apparatus which allow a user to accurately perform an examination on a subject.SOLUTION: A diagnostic apparatus (1) includes: a monitor (33) that displays an image in which an organ of an analyte is imaged; an organ region extraction unit (35) that extracts the organ by analyzing the image; a shape analysis unit (36) that approximates a shape of the organ extracted by the organ region extraction unit (35) using an implicit function; and an organ evaluation unit (37) that evaluates a disease state of the organ on the basis of a shape parameter of an approximation curve represented by the implicit function approximated in the shape analysis unit (36).SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a diagnostic device that assists in the diagnosis of a subject and a method for controlling the diagnostic device. [Background technology]

[0002] Conventionally, so-called ultrasound diagnostic devices and the like have been used to acquire images showing cross sections within a subject, and users such as doctors examine the subject based on the acquired images. In this case, technologies such as those described in Patent Document 1 have been developed to enable more accurate examinations of the subject. Patent Document 1 discloses a method for calculating an index value related to the subject's visceral fat based on the cross-sectional area of ​​the subject's abdomen and the thickness of the subject's subcutaneous fat measured from ultrasound images. The cross-sectional area of ​​the subject's abdomen is approximately measured by approximating the cross-section of the subject's abdomen to an ellipse and calculating the area of ​​the ellipse. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2016-202208 Summary of the Invention [Problem to be solved by the invention]

[0004] Incidentally, in order for a user to accurately examine a subject, it may be necessary to not only evaluate the area and size of a part of the subject, but also to evaluate the shape of the subject's organ, for example, when examining an organ whose shape may be abnormal depending on the condition of the prostate, as disclosed in Patent Document 1. In such cases, the user often qualitatively evaluates the shape of the organ by checking images showing cross sections of the subject, but qualitative evaluation has problems in that it is difficult to evaluate changes in the shape of the organ over time, and accurate evaluation is difficult because there are individual differences in the shape of the organ depending on the body type of the subject, etc.

[0005] The present invention has been made to solve such conventional problems, and has as its object to provide a diagnostic device and a method for controlling a diagnostic device that allow a user to perform accurate tests on a subject. [Means for solving the problem]

[0006] In order to achieve the above object, the diagnostic device according to the present invention is characterized by comprising a monitor that displays an image of a subject's organs, an organ region extraction unit that extracts the organs by analyzing the image, a shape analysis unit that approximates the shape of the organ extracted by the organ region extraction unit using an implicit function, and an organ evaluation unit that evaluates the condition of the organ based on the shape parameters of the approximation curve represented by the implicit function approximated in the shape analysis unit.

[0007] The organ evaluation unit can evaluate the condition of the organ using the shape parameters of the approximation curve as evaluation indices. In this case, the shape analysis unit performs function approximation using a hyperellipse, and the organ evaluation unit can evaluate the condition of the organ using a shape parameter that represents the deflection of the approximation curve as an evaluation index. The shape analysis unit performs function approximation using a hyperellipse, and the organ evaluation unit can also evaluate the condition of the organ using a shape parameter that indicates the balance of the shape of the approximation curve as an evaluation index. The shape analysis unit can also perform function approximation using an ellipse.

[0008] The organ evaluation unit creates a machine learning model using the shape parameters of the approximation curve, and can evaluate the condition of the organ using the machine learning model. In this case, the shape analysis unit can perform function approximation using a hyperellipse. The shape analysis unit can also perform function approximation using an ellipse.

[0009] The organ is the prostate, and the organ evaluation unit can evaluate the condition of the prostate by evaluating the shape of the prostate based on the shape parameters. The organ is the bladder, and the organ evaluation unit can evaluate the condition of the bladder by evaluating the cavity of the bladder based on the shape parameters.

[0010] The image may also be an ultrasound image. In this case, the diagnostic device can include an ultrasound probe and an image generating unit that generates an ultrasound image by transmitting and receiving ultrasound beams to and from the subject using the ultrasound probe. The image may also be any of an X-ray image, a computed tomography image, and a nuclear magnetic resonance image.

[0011] The method for controlling a diagnostic device according to the present invention is characterized in that it displays an image of a subject's organ on a monitor, extracts the organ by analyzing the image, approximates the shape of the extracted organ using an implicit function, and evaluates the condition of the organ based on the shape parameters of the approximation curve represented by the approximated implicit function. [Effects of the Invention]

[0012] According to the present invention, the diagnostic device comprises a monitor that displays an image of the subject's organs, an organ area extraction unit that extracts the organs by analyzing the image, a shape analysis unit that approximates the shape of the organ extracted by the organ area extraction unit using an implicit function, and an organ evaluation unit that evaluates the condition of the organ based on the shape parameters of the approximation curve represented by the implicit function approximated in the shape analysis unit, thereby enabling the user to perform accurate examinations on the subject. [Brief explanation of the drawings]

[0013] [Figure 1] 1 is a block diagram showing the configuration of an ultrasound diagnostic apparatus according to an embodiment of the present invention; [Figure 2] 1 is a block diagram showing a configuration of a transmission / reception circuit according to an embodiment of the present invention; [Figure 3] FIG. 2 is a block diagram showing a configuration of an image generating unit according to the embodiment of the present invention. [Figure 4]FIG. 1 is a diagram schematically illustrating an example of an ultrasound image of a normal prostate. [Figure 5] FIG. 1 is a diagram schematically illustrating an example of an ultrasound image of an enlarged prostate gland. [Figure 6] 10A and 10B are diagrams illustrating examples in which a shape parameter representing the aspect ratio of a super ellipse is changed. [Figure 7] 10A and 10B are diagrams illustrating examples in which shape parameters representing the four angles of a super ellipse are changed. [Figure 8] 10A and 10B are diagrams illustrating examples in which shape parameters representing the deflection of a super ellipse are changed. [Figure 9] 10A and 10B are diagrams illustrating examples in which shape parameters representing the balance of the shape of a super ellipse are changed. [Figure 10] FIG. 10 shows an example of a contour of a circular organ to be approximated and the curve represented by the underlying implicit functions used in the approximation. [Figure 11] 4 is a flowchart showing the operation of the ultrasound diagnostic apparatus according to an embodiment of the present invention. [Figure 12] FIG. 10 is a diagram illustrating an example of a curve represented by an implicit function. DETAILED DESCRIPTION OF THE INVENTION

[0014] Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings. The following description of the components will be given based on a representative embodiment of the present invention, but the present invention is not limited to such an embodiment. In this specification, a numerical range expressed using "to" means a range that includes the numerical values ​​before and after "to" as the lower and upper limits. In this specification, the terms "same" and "identical" include a margin of error generally accepted in the technical field.

[0015] Embodiment 1 shows the configuration of an ultrasonic diagnostic apparatus 1 according to an embodiment of the present invention. The ultrasonic diagnostic apparatus 1 includes an ultrasonic probe 2 and an apparatus main body 3 connected to the ultrasonic probe 2. The ultrasonic probe 2 includes a transducer array 21, to which a transmission / reception circuit 22 is connected.

[0016] The device main body 3 includes an image generation unit 31 connected to the transmission / reception circuit 22 of the ultrasound probe 2. A display control unit 32 and a monitor 33 are sequentially connected to the image generation unit 31. An image memory 34 is also connected to the image memory 34. An organ region extraction unit 35, a shape analysis unit 36, and an organ evaluation unit 37 are sequentially connected to the image memory 34. The organ region extraction unit 35, the shape analysis unit 36, and the organ evaluation unit 37 are also connected to the display control unit 32. An examination result memory 38 is also connected to the image memory 34, the organ region extraction unit 35, the shape analysis unit 36, and the organ evaluation unit 37.

[0017] A main body control unit 39 is connected to the transmission / reception circuit 22, image generation unit 31, display control unit 32, image memory 34, organ region extraction unit 35, shape analysis unit 36, organ evaluation unit 37, and examination result memory 38. An input device 40 is also connected to the main body control unit 39. The image generating unit 31, the display control unit 32, the organ region extracting unit 35, the shape analyzing unit 36, the organ evaluating unit 37 and the main body control unit 39 constitute a processor 41 for the device main body 3.

[0018] The transducer array 21 of the ultrasonic probe 2 has a plurality of ultrasonic transducers arranged one-dimensionally or two-dimensionally. These ultrasonic transducers transmit ultrasonic waves in accordance with drive signals supplied from the transmission / reception circuit 22, receive ultrasonic echoes from the subject, and output signals based on the ultrasonic echoes. Each ultrasonic transducer is configured by forming electrodes on both ends of a piezoelectric element made of, for example, a piezoelectric ceramic typified by PZT (Lead Zirconate Titanate), a polymer piezoelectric element typified by PVDF (Poly Vinylidene Di Fluoride), or a piezoelectric single crystal typified by PMN-PT (Lead Magnesium Niobate-Lead Titanate).

[0019] The transmission / reception circuit 22, under the control of the main body control unit 39, transmits ultrasonic waves from the transducer array 21 and generates sound ray signals based on reception signals acquired by the transducer array 21. As shown in Fig. 2, the transmission / reception circuit 22 has a pulser 23 connected to the transducer array 21, an amplifier 24, an AD (Analog to Digital) converter 25, and a beamformer 26, which are connected in series from the transducer array 21 in this order.

[0020] The pulser 23 includes, for example, a plurality of pulse generators, and adjusts the delay amount of each drive signal and supplies it to the plurality of ultrasonic transducers of the transducer array 21 so that the ultrasonic waves transmitted from the plurality of ultrasonic transducers form an ultrasonic beam based on a transmission delay pattern selected in response to a control signal from the main body control unit 39. In this way, when a pulsed or continuous wave voltage is applied to the electrodes of the ultrasonic transducers of the transducer array 21, the piezoelectric material expands and contracts, and pulsed or continuous wave ultrasonic waves are generated from each ultrasonic transducer, and an ultrasonic beam is formed from a composite wave of these ultrasonic waves.

[0021] The transmitted ultrasonic beam is reflected by an object such as a part of the subject, and propagates toward the transducer array 21 of the ultrasonic probe 2. The ultrasonic echoes propagating toward the transducer array 21 in this manner are received by the respective ultrasonic transducers that make up the transducer array 21. At this time, each ultrasonic transducer that makes up the transducer array 21 expands and contracts upon receiving the propagating ultrasonic echoes, generating received signals that are electrical signals, and outputting these received signals to the amplifier 24.

[0022] The amplifier 24 amplifies signals input from each ultrasonic transducer constituting the transducer array 21 and transmits the amplified signals to the AD converter 25. The AD converter 25 converts the signals transmitted from the amplifier 24 into digital reception data. The beamformer 26 performs so-called reception focusing processing by delaying and adding each piece of reception data received from the AD converter 25. This reception focusing processing causes the reception data converted by the AD converter 25 to be phased and added, and a sound ray signal in which the focus of the ultrasonic echo is narrowed is acquired.

[0023] As shown in FIG. 3, the image generating unit 31 has a configuration in which a signal processing unit 51, a DSC (Digital Scan Converter) 52, and an image processing unit 53 are connected in series.

[0024] The signal processing unit 51 corrects the sound ray signals received from the transmission / reception circuit 22 for attenuation due to distance in accordance with the depth of the ultrasonic reflection position using the sound velocity value set by the main body control unit 39, and then performs envelope detection processing to generate a B-mode image signal, which is tomographic image information regarding the tissue within the subject.

[0025] The DSC 52 converts (raster converts) the B-mode image signal generated by the signal processing unit 51 into an image signal that conforms to the scanning method of a normal television signal. The image processing unit 53 performs various necessary image processing such as gradation processing on the B-mode image signal input from the DSC 52, and then sends the B-mode image signal to the display control unit 32 and the image memory 34. Hereinafter, the B-mode image signal that has been subjected to image processing by the image processing unit 53 will be referred to as an ultrasound image.

[0026] The main body control unit 39 controls the transmitting / receiving circuit 22 of the ultrasonic probe 2 and each part of the device main body 3 according to a pre-recorded program or the like. Under the control of the main body control unit 39 , the display control unit 32 performs predetermined processing on the ultrasound image etc. generated by the image generation unit 31 and displays it on the monitor 33 .

[0027] The monitor 33 performs various displays under the control of the display control unit 32. The monitor 33 includes a display device such as an LCD (Liquid Crystal Display) or an organic EL display (Organic Electroluminescence Display).

[0028] The input device 40 is used by the user to perform input operations and is configured by devices such as a keyboard, a mouse, a trackball, a touchpad, and a touch panel, which allow the user to perform input operations.

[0029] The image memory 34 stores the ultrasound images generated by the image generation unit 31 under the control of the main body control unit 39, and transmits the stored ultrasound images to the organ region extraction unit 35. As the image memory 34, for example, a recording medium such as a flash memory, an HDD (Hard Disk Drive), an SSD (Solid State Drive), an FD (Flexible Disk), an MO disk (Magneto-Optical disk), an MT (Magnetic Tape), a RAM (Random Access Memory), a CD (Compact Disc), a DVD (Digital Versatile Disc), an SD card (Secure Digital card), or a USB memory (Universal Serial Bus memory) can be used.

[0030] Here, some organs in a subject, such as the prostate, change shape depending on the condition of the subject. For example, as shown in Figures 4 and 5, in ultrasound image U1, an enlarged prostate T2 has a shape that is relatively close to a circle compared to a normal prostate T1, and is often deformed so as to protrude toward the bladder R1. When examining an organ whose shape changes depending on the condition of the subject, a user such as a doctor usually checks ultrasound image U and qualitatively evaluates the shape of the organ shown in ultrasound image U. However, qualitative evaluation has problems in that it is difficult to evaluate changes in the shape of the organ over time, and accurate evaluation is difficult because the shape of the organ varies from person to person depending on the body type of the subject, etc.

[0031] As will be described in detail below, in order to allow a user to perform an accurate examination, the ultrasound diagnostic device 1 according to the embodiment of the present invention analyzes an ultrasound image U to extract organs contained in the ultrasound image U, approximates the extracted organs using a so-called implicit function, and quantitatively evaluates the condition of the organ based on the shape parameters of the approximation curve approximated by the implicit function.

[0032] The organ region extraction unit 35 extracts the organs of the subject by analyzing the ultrasound image acquired from the image memory 34. In this case, the organ region extraction unit 35 can extract the organs of the subject included in the ultrasound image by applying, for example, a method using simple pattern matching, a machine learning method described in Csurka et al.: Visual Categorization with Bags of Keypoints, Proc. of ECCV Workshop on Statistical Learning in Computer Vision, pp. 59-74 (2004), or a general image recognition method using deep learning or a so-called convolutional neural network (CNN) described in Krizhevsk et al.: ImageNet Classification with Deep Convolutional Neural Networks, Advances in Neural Information Processing Systems 25, pp. 1106-1114 (2012).

[0033] Furthermore, the organ region extraction unit 35 can, for example, highlight the extracted organ region on the ultrasound image and display it on the monitor 33. At this time, the organ region extraction unit 35 can, for example, color the organ region on the ultrasound image and display it, display the outline of the organ region on the ultrasound image, or color the outline of the organ region on the ultrasound image and display it, etc.

[0034] The shape analysis unit 36 ​​uses a so-called implicit function to approximate the shape of the organ extracted by the organ region extraction unit 35. For example, when the x-axis is set along the horizontal direction of the ultrasound image U and the y-axis is set along the direction perpendicular to the horizontal direction, the shape analysis unit 36 ​​can use a hyperelliptic function shown in the following equation (1) as the implicit function.

number

[0035] In the function of the superellipse in equation (1), a position parameter representing the placement position or rotation position of the superellipse, a scale parameter representing the size of the superellipse, and a shape parameter representing the shape of the superellipse can be set. For example, by replacing the variables x and y in the function of equation (1) with the variables x1 and y1 shown in equation (2), it is possible to set parameters representing the amount of displacement of the translation of the superellipse, i.e., position parameters L1 and L2 representing the placement position of the superellipse. L1 represents the amount of displacement in the direction along the x-axis of the superellipse, and L2 represents the amount of displacement in the direction along the y-axis of the superellipse.

number

[0036] Furthermore, by replacing the variables x and y with the variables x2 and y2 shown in equation (3), a parameter representing the amount of rotation of the superellipse, i.e., a position parameter A representing the rotation position of the superellipse, can be set. Note that A satisfies 0≦A<2π (π is the constant of the circumference of a circle).

number

[0037] Furthermore, W1 in equation (1) is a scale parameter that represents the scale of the hyperellipse in the direction along the x-axis, and W2 is a scale parameter that represents the scale of the hyperellipse in the direction along the y-axis. In addition, the ratio W1 / W2 or W2 / W1 of the scale parameters W1 and W2 can be set as a shape parameter that represents the aspect ratio of the superellipse. For example, Figure 6 shows curve F1 (circle) when W1:W2=1:1 in equation (1), curve F2 when W1:W2=1:2, and curve F3 when W1:W2=2:1. In this way, the aspect ratio of the superellipse changes depending on the ratio W1 / W2 or W2 / W1 of W1 and W2.

[0038] Also, E in formula (1) is a shape parameter representing the four angles of the super-ellipse. Here, the four angles refer to the degree to which the super-ellipse has a shape close to a rectangle, and the larger E is, the closer the super-ellipse approaches a rectangle. Fig. 7 shows examples of the curve F4 (ellipse) when E = 1 in formula (1), the curve F5 when E > 1, and the curve F6 when E < 1. The curve F4 is an ellipse, the curve F5 has a shape closer to a rectangle than an ellipse, and the curve F6 has a shape closer to a rhombus than an ellipse.

[0039] Also, by substituting the variables x and y with the variables x3 and y3 shown in formula (4), the shape parameter B representing the deflection of the super-ellipse can be set. Note that B satisfies -1 < B < 1 and B ≠ 0, that is, -1 < B < 0 and 0 < B < 1. Also, D in formula (4) is a variable represented by formula (5).

Number

Number

[0040] The deflection of the super-ellipse refers to the deformation such that both ends of the super-ellipse in the direction along the x-axis are pulled downward, that is, in the negative direction, along the y-axis. For example, Fig. 8 shows the curve F7 (ellipse) when no deflection deformation is performed, the curve F8 when the shape parameter B has a certain value, and the curve F9 when the shape parameter B has a larger value. Thus, the larger the shape parameter B, the greater the deflection of the super-ellipse, and the smaller the shape parameter B, the smaller the deflection of the super-ellipse.

[0041] Also, by substituting the variable x with the variable x4 shown in formula (6), the shape parameter T representing the balance of the shape of the super-ellipse can be set. Note that T satisfies -1 < T < 1.

Number

[0042] The balance of the shape of a superellipse refers to the balance between the width of the upper part and the width of the lower part of the superellipse, i.e., the width of the part on the positive side of the y-axis and the width of the part on the negative side of the y-axis. For example, FIG. 9 shows a curve F10 (ellipse) when the shape parameter T=0, a curve F11 when the shape parameter T has a constant positive value, and a curve F12 when the shape parameter T has an even larger positive value. As shown, as the shape parameter T increases in the positive direction, the width of the upper part of the superellipse narrows and the width of the lower part widens. Note that, although not shown, as the shape parameter T increases in the negative direction, the width of the upper part of the superellipse widens and the width of the lower part narrows.

[0043] Furthermore, when approximating the shape of the organ extracted by the organ region extraction unit 35 using an implicit function such as a superellipse function, the shape analysis unit 36 ​​can perform approximation using an implicit function by applying the so-called steepest descent method or the so-called least squares method, as described in "ZHANG, Xiaoming; ROSIN, Paul L. Superellipse fitting to partial data. Pattern Recognition, 2003, 36.3: 743-752." to the basic implicit function shown in equation (1).

[0044] In the following, to briefly explain approximation using an implicit function, we will introduce an example in which the steepest descent method is applied to the basic implicit function representing a circle shown in Equation (7), and a circular organ is approximated using the implicit function of Equation (7).

number

[0045] The approximation procedure begins by setting the objective function f(a, b, r) shown in equation (8) using the relationship in equation (7). Here, x i and y iare the observed values, i.e., the coordinates of the points to be set on the contour of the organ to be approximated. Furthermore, n represents the number of observed values, i.e., the number of points to be set on the contour of the organ to be approximated. The objective function f(a, b, r) represents the error between the n observed values ​​and a, b, and r, and the objective of the steepest descent method is to determine the values ​​of a, b, and r so as to minimize the objective function f(a, b, r). By substituting the values ​​of a, b, and r determined in this way into the implicit function of equation (7), the implicit function that approximates the organ can be obtained.

number

[0046] Next, using a, b, and r as variables, the partial derivative of the objective function f(a, b, r) with respect to a, the partial derivative of the objective function f(a, b, r) with respect to b, and the partial derivative of the objective function f(a, b, r) with respect to r are calculated, and each value is multiplied by a learning rate G, which is a constant value less than 1, and −1 to calculate the change amounts H1, H2, and H3 of a, b, and r required to reduce the objective function f(a, b, r), as shown in equations (9) to (11). Here, initial values ​​are set for a, b, and r, and specific values ​​of the change amounts H1, H2, and H3 are calculated by substituting these initial values ​​into equations (9) to (11). Furthermore, specific values ​​of H1, H2, and H3 are added to the initial values ​​of a, b, and r to newly determine specific values ​​of a, b, and r that will further reduce the objective function f(a, b, r).

number

number

number

[0047] By repeating this process, the values ​​of a, b, and r that minimize the objective function f(a, b, r) are determined.

[0048] As a specific example, consider approximating a circle C2 having four observations, i.e., a circle C2 passing through four points P1 to P4, using a circle C1 expressed as (a, b, r) = (0, 0, 1) in the basic implicit function of equation (7), as shown in Fig. 10. The coordinates of point P1 are (0, -1), the coordinates of point P2 are (0, 3), the coordinates of point P3 are (-2, 1), and the coordinates of point P4 are (2, 1).

[0049] First, the initial value f(0,0,1) of the objective function f(a,b,r) is determined to be 96 based on the four observed values. If the learning rate G is set to, for example, 0.01, the deviations H1 to H3 are 0, 1.28, and 0.64, as shown in equations (12) to (14). These are added to the initial values ​​of a, b, and r to obtain new values ​​of a, b, and r: 0, 1.28, and 1.64. In this case, the value f(0,1.28,1.68) of the objective function f(a,b,r) is 10.223, which is smaller than the initial value f(0,0,1). In this way, by repeatedly determining the values ​​of a, b, and r that minimize the objective function f(a,b,r), the values ​​of a, b, and r that minimize the objective function f(a,b,r) are determined, and an approximation curve for a circular organ is obtained.

number

number

number

[0050] When the hyperellipse function shown in Equation (1) is used as the basic implicit function for function approximation, the steepest descent method can be used with the following variables: the parameters of the hyperellipse, such as scale parameters W1 and W2 representing the scale of the hyperellipse, shape parameter E representing the angle of the hyperellipse, shape parameter B representing the curvature of the hyperellipse, and shape parameter T representing the balance of the hyperellipse; and position parameters L1 and L2 representing the position of the hyperellipse and position parameter A representing the rotation position of the hyperellipse. In this case, the shape parameters, position parameters, and scale parameters that minimize the objective function are determined, and an implicit function that approximates the organ is determined. The shape parameter W1 / W2 or W2 / W1 representing the aspect ratio of the hyperellipse is determined by optimizing the scale parameters W1 and W2.

[0051] Furthermore, in order to allow the user to understand how much the finally obtained approximate curve has deformed from the curve represented by the basic implicit function used in the function approximation, the shape analysis unit 36 ​​can display the values ​​of the shape parameters in the approximation function representing the finally obtained approximate curve on the monitor 33. Furthermore, in order to show how much the finally obtained approximate curve has deformed from the curve represented by the basic implicit function used in the function approximation, the shape analysis unit 36 ​​can also display the finally obtained approximate curve and the curve represented by the basic implicit function on the monitor 33.

[0052] The organ evaluation unit 37 evaluates the condition of an organ based on the shape parameters of the approximation function represented by the implicit function approximated by the shape analysis unit 36. The organ evaluation unit 37 can evaluate the condition of an organ by evaluating the shape of the organ using, as evaluation indexes, shape parameters of the approximation function, such as the shape parameter W1 / W2 or W2 / W1 representing the aspect ratio of the approximation curve, the shape parameter E representing the angle of the approximation curve, the shape parameter B representing the deflection of the approximation function, and the shape parameter representing the balance of the approximation curve.

[0053] For example, as shown in FIG. 5, it is known that the prostate gland T2, when enlarged, takes on a shape close to a circle. The organ evaluation unit 37 can evaluate whether or not the subject is suspected of having benign prostatic hyperplasia by, for example, evaluating the shape of the subject's prostate gland based on shape parameters. In this case, the organ evaluation unit 37 can evaluate, for example, whether or not the subject is suspected of having benign prostatic hyperplasia if the shape parameter represented by the ratio W1 / SW or W2 / W1 of scale parameters W1 and W2 representing the scale of the approximate curve is close to 1, i.e., if the value is within a certain range centered around 1, the organ evaluation unit 37 can evaluate whether or not the subject is likely to have benign prostatic hyperplasia. Furthermore, the organ evaluation unit 37 can evaluate, for example, whether or not the subject is suspected of having benign prostatic hyperplasia if the shape parameter E representing the angle of the approximate curve is close to 1, i.e., if the value is within a certain range centered around 1, the organ evaluation unit 37 can evaluate whether or not the subject is likely to have ....

[0054] Furthermore, the organ evaluation unit 37 can evaluate that there is a suspicion of the onset of benign prostatic hyperplasia when the shape parameter B representing the deflection of the approximate curve is smaller than a certain value, and can also evaluate that there is a low possibility of the onset of benign prostatic hyperplasia in other cases. Furthermore, the organ evaluation unit 37 can evaluate that there is a suspicion of the onset of benign prostatic hyperplasia when the shape parameter T representing the balance of the approximate curve has a value within a certain range close to 0, and can also evaluate that there is a low possibility of the onset of benign prostatic hyperplasia in other cases.

[0055] Furthermore, as shown in FIG. 5, the enlarged prostate T2 protrudes toward the bladder R1, and the bladder R1 appears to have a deep, concave shape in the ultrasound image U. In such cases, it is generally known that cystitis is likely to occur. When the organ region extraction unit 35 detects the bladder R1 as the organ to be examined, the shape analysis unit 36 ​​can evaluate whether or not cystitis is suspected by evaluating the concave shape of the bladder R1. In this case, for example, when a shape parameter B representing the deflection in the approximation curve of the bladder R1 is greater than a certain value, the shape analysis unit 36 ​​can evaluate that cystitis is suspected as a bladder condition, and in other cases, can evaluate that the risk of cystitis is low.

[0056] Furthermore, the organ evaluation unit 37 can also indirectly evaluate the condition of an organ by inputting the shape parameters into a machine learning model such as a so-called regression analysis model or a so-called classification tree model. In this case, the organ evaluation unit 37 can output the evaluation result of the organ by inputting the shape parameters of an approximate curve represented by an implicit function approximated by the shape analysis unit 36 ​​into a machine learning model that has previously learned the relationship between the condition of the organ and the shape parameters.

[0057] Under the control of the main body control unit 39, the examination result memory 38 stores, as examination results, the ultrasound image U from which the organ has been extracted by the organ region extraction unit 35, the results of organ extraction by the organ region extraction unit 35, information on the implicit function approximated by the shape analysis unit 36, and the evaluation results on the condition of the organ by the organ evaluation unit 37, all in association with one another. The information on the implicit function approximated by the shape analysis unit 36 ​​includes shape parameters of the implicit function, etc. The examination result memory 38 can be, for example, a recording medium such as a flash memory, HDD, SSD, FD, MO disk, MT, RAM, CD, DVD, SD card, or USB memory.

[0058] The processor 41 having the image generation unit 31, display control unit 32, organ area extraction unit 35, shape analysis unit 36, organ evaluation unit 37 and main body control unit 39 is composed of a CPU (Central Processing Unit) and a control program for causing the CPU to perform various processes, but may also be composed using an FPGA (Field Programmable Gate Array), DSP (Digital Signal Processor), ASIC (Application Specific Integrated Circuit), GPU (Graphics Processing Unit), or other ICs (Integrated Circuits), or a combination of these.

[0059] In addition, the image generation unit 31, display control unit 32, organ area extraction unit 35, shape analysis unit 36, organ evaluation unit 37 and main body control unit 39 of the processor 41 can be partially or entirely integrated into a single CPU or the like.

[0060] Next, the basic operation of the ultrasound diagnostic apparatus 1 according to this embodiment will be described with reference to the flowchart of FIG.

[0061] First, in step S1, a user of the ultrasound diagnostic apparatus 1 brings the ultrasound probe 2 into contact with the body surface of the subject, and in this state, an ultrasound image U is acquired. When the ultrasound image U1 is acquired, the transmission / reception circuit 22 performs so-called reception focus processing under the control of the main body control unit 39 to generate sound ray signals. The sound ray signals generated by the transmission / reception circuit 22 are sent to the image generation unit 31. The image generation unit 31 generates an ultrasound image U using the sound ray signals sent from the transmission / reception circuit 22. The ultrasound image U acquired in this manner is sent to the display control unit 32 and stored in the image memory 34.

[0062] In step S2, the ultrasound image U acquired in step S1 and sent to the display control unit 32 is subjected to predetermined processing and then displayed on the monitor 33. This allows the user to check the acquired ultrasound image U.

[0063] In step S3, the organ region extraction unit 35, under the control of the main body control unit 39, reads out the ultrasound image U acquired in step S1 from the image memory 34 and extracts the organs of the subject by analyzing this ultrasound image U. In this case, the organ region extraction unit 35 can extract the organs of the subject included in the ultrasound image U by applying, for example, a method using simple pattern matching, a machine learning method described in Csurka et al.: Visual Categorization with Bags of Keypoints, Proc. of ECCV Workshop on Statistical Learning in Computer Vision, pp. 59-74 (2004), or a general image recognition method using deep learning or CNN described in Krizhevsk et al.: ImageNet Classification with Deep Convolutional Neural Networks, Advances in Neural Information Processing Systems 25, pp. 1106-1114 (2012).

[0064] Furthermore, the organ region extraction unit 35 can highlight the extracted organ region on the ultrasound image U and display it on the monitor 33. At this time, the organ region extraction unit 35 can, for example, color the organ region on the ultrasound image U, display the outline of the organ region on the ultrasound image U, or color the outline of the organ region on the ultrasound image U. Furthermore, the organ extraction results in step S3 can be saved in the examination result memory 38.

[0065] In step S4, the shape analysis unit 36 ​​approximates the shape of the organ extracted in step S3 using an implicit function. For example, the shape analysis unit 36 ​​may set the x-axis along the horizontal direction of the ultrasound image U, set the y-axis along the direction perpendicular to the horizontal direction, and use the hyperellipse function shown in Equation (1) as the implicit function. In this case, the shape analysis unit 36 ​​can determine the implicit function approximating the organ by using a steepest descent method with the following shape parameters as variables: scale parameters W1 and W2 representing the scale of the hyperellipse; shape parameter W1 / W2 or W2 / W1 representing the aspect ratio of the hyperellipse; shape parameter E representing the angle of the hyperellipse; shape parameter B representing the deflection of the hyperellipse; and shape parameter T representing the balance of the hyperellipse; and position parameters such as position parameters L1 and L2 representing the position of the hyperellipse and position parameter A representing the rotation position of the hyperellipse.

[0066] The value of the learning rate G used in the steepest descent method may be preset as an initial setting, or may be input by the user via the input device 40. When the learning rate G is input by the user, for example, the learning rate G may be input before the start of an examination on the subject, or a message prompting the user to input the learning rate G may be displayed on the monitor 33 when step S4 is started, and the user may input the learning rate G after confirming this. Additionally, information including the shape parameters of the implicit functions approximated in step S4 can be stored in the inspection result memory 38.

[0067] Finally, in step S5, the organ evaluation unit 37 evaluates the condition of the organ based on the shape parameters of the approximation curve represented by the implicit function approximated in step S4. The organ evaluation unit 37 can evaluate the condition of the organ by using, as evaluation indices, shape parameters of the approximation function, such as the shape parameter W1 / W2 or W2 / W1 representing the aspect ratio of the approximation curve, the shape parameter E representing the angle of the approximation curve, the shape parameter B representing the deflection of the approximation function, and the shape parameter representing the balance of the approximation curve.

[0068] It is known that the prostate gland T2, when enlarged, takes on a shape close to a circle, as shown in Fig. 5. Therefore, the organ evaluation unit 37 can evaluate that there is a suspicion of the onset of benign prostatic hyperplasia when, for example, the shape parameter W1 / W2 or W2 / W1 representing the aspect ratio of the approximate curve has a value within a certain range centered around 1, when the shape parameter E representing the angle of the approximate curve has a value within a certain range centered around 1, when the shape parameter B representing the deflection of the approximate curve is smaller than a certain value, or when the shape parameter T representing the balance of the approximate curve has a value within a certain range close to 0, and can evaluate that there is a low possibility of the onset of benign prostatic hyperplasia in other cases.

[0069] 5, the enlarged prostate T2 protrudes toward the bladder R1, and the bladder R1 appears to have a deep, concave shape in the ultrasound image U. In such cases, it is generally known that cystitis is likely to occur. When the bladder R1 is detected as the organ to be examined in step S3, the shape analysis unit 36 ​​can evaluate the concave shape of the bladder R1. In this case, for example, when a shape parameter B representing the deflection in the approximation curve of the bladder R1 is greater than a certain value, the shape analysis unit 36 ​​can evaluate that cystitis is suspected as a bladder condition, and in other cases, can evaluate that the risk of cystitis is low.

[0070] Furthermore, the organ evaluation unit 37 can also evaluate the condition of an organ by indirectly evaluating shape parameters using a machine learning model such as a so-called regression analysis model or a so-called classification tree model. In this case, the organ evaluation unit 37 can output an evaluation result for the organ by inputting shape parameters of an approximate curve represented by an implicit function approximated by the shape analysis unit 36 ​​into a machine learning model that has previously learned the relationship between the condition of the organ and the shape parameters.

[0071] The evaluation results of the subject's organs obtained in this manner are displayed on the monitor 33 via the display control unit 32. The evaluation results of the subject's organs are also stored in the examination result memory 38. The evaluation results stored in the examination result memory 38 can be viewed by the user after the examination, for example.

[0072] When the processing of step S5 is completed, the operation of the ultrasonic diagnostic apparatus 1 according to the flowchart of FIG. 11 ends.

[0073] When examining an organ whose shape changes depending on the condition, such as the prostate, a doctor or other user usually checks the ultrasound image U and qualitatively evaluates the shape of the organ shown in the ultrasound image U. However, qualitative evaluation has the problems that it is difficult to evaluate changes in the shape of the organ over time, and accurate evaluation is difficult because the shape of the organ varies from person to person depending on the body type of the subject, etc.

[0074] According to the ultrasound diagnostic device 1 of the embodiment of the present invention, an ultrasound image U is analyzed to extract organs included in the ultrasound image U, the extracted organs are approximated by an implicit function, and the pathology of the organs is quantitatively evaluated based on the shape parameters of the approximated curve approximated by the implicit function. Therefore, even if there are individual differences in the shape of the organ due to the body type of the subject, the user can accurately evaluate the shape of the organ by checking the obtained evaluation results and perform an accurate examination. Furthermore, when periodically performing examinations on the same subject using the ultrasound diagnostic device 1, the user can easily evaluate changes in the shape of the organ over time by checking past examination results stored in the examination result memory 38, and perform the examination more accurately.

[0075] It has been explained that one frame of ultrasound image U is generated in step S1, and then that ultrasound image U is analyzed in step S3 to extract organs. However, in step S1, multiple frames of ultrasound image U may be generated within a set time of several seconds, and one frame of ultrasound image U to be used for analysis may be selected from the multiple frames of ultrasound image U.

[0076] For example, when an examination of a subject is started based on a user's instruction via the input device 40, multiple frames of ultrasound image U are generated within a predetermined time of several seconds from the start of the examination and stored in the image memory 34. The series of ultrasound images U stored in the image memory 34 in this manner are displayed on the monitor 33 in a list display or a so-called scroll display. The user can check the series of ultrasound images U displayed on the monitor 33 and select the ultrasound image U that most clearly shows the organ being examined as the ultrasound image U to be used in the processing from step S2 onwards. The selected ultrasound image U is displayed on the monitor 33 in step S2 and analyzed by the organ region extraction unit 35 in step S3. In this way, the user selects one frame of ultrasound image U to be used in the processing from step S2 onwards, thereby making it possible to process an ultrasound image U that more clearly shows the organ, thereby improving the accuracy of the evaluation result of the organ's condition finally obtained by the organ evaluation unit 37.

[0077] The organ region extraction unit 35 can also automatically extract ultrasound images U to be used to extract organs by analyzing a series of ultrasound images U stored in the image memory 34. In this case, the organ region extraction unit 35 can, for example, calculate the sharpness of the edges of organs appearing in each of the series of ultrasound images U, and select the ultrasound image U of one frame with the highest calculated sharpness as the ultrasound image U to be used for subsequent processing. Here, the organ region extraction unit 35 can, for example, detect the edges of the organs in the ultrasound images U, and calculate the sharpness based on the contrast of pixels surrounding the detected edges. By automatically selecting an ultrasound image U in this way, it is possible to process an ultrasound image U that more clearly shows the organ, thereby improving the accuracy of the evaluation result of the organ's condition finally obtained by the organ evaluation unit 37.

[0078] Furthermore, the result of organ extraction in step S3 can be modified by the user via, for example, the input device 40. For example, the organ extracted in step S3 is superimposed on the ultrasound image U and displayed in an emphasized manner on the monitor 33, and at this time, modifications to the organ region by the user via the input device 40 are accepted. When the user specifies a region representing the organ on the ultrasound image U via the input device 40, the organ region extracted by the organ region extraction unit 35 is replaced with the organ region specified by the user. In this case, in step S4 following step S3, the organ region specified by the user is approximated by an implicit function. This makes it possible to accurately evaluate the shape of the organ even if the organ region extraction unit 35 fails to correctly extract the organ for some reason, such as when the ultrasound image U is not sufficiently clear.

[0079] Furthermore, although the transmitting / receiving circuit 22 has been described as being provided in the ultrasonic probe 2, it may also be provided in the device main body 3 instead of in the ultrasonic probe 2. Furthermore, although the image generating unit 31 has been described as being provided in the device main body 3, it may be provided in the ultrasound probe 2 instead of in the device main body 3. Furthermore, although the ultrasonic probe 2 and the device main body 3 are shown to be connected by wire, they may be connected wirelessly.

[0080] Furthermore, in the image generation unit 31, the DSC 52 is connected to the signal processing unit 51, and the image processing unit 53 is connected to the DSC 52. However, the image processing unit 53 may be connected to the signal processing unit 51, and the DSC 52 may be connected to the image processing unit 53. In this case, the image processing unit 53 performs predetermined processing such as gradation processing on the ultrasound image U generated by the signal processing unit 51, and then the ultrasound image U is raster converted by the DSC 52. In this way, even when the signal processing unit 51, image processing unit 53, and DSC 52 are connected in this order, the ultrasound image U is generated in the image generation unit 31 in the same way as when the signal processing unit 51, DSC 52, and image processing unit 53 are connected in this order.

[0081] Furthermore, although the types of implicit functions used by the shape analysis unit 36 ​​when performing function approximation of an organ are exemplified as a hyperellipse, an ellipse, and a circle, the implicit functions used are not particularly limited to these. For example, the function shown in the following equation (15) can also be used as the implicit function. For example, in equation (15), when M=1.1, N=1.0, K=0.4, and Q=1.2, the curve F13 shown in FIG. 12 is obtained. Equation (15) can be used, for example, as the implicit function when approximating the shape of the prostate. When equation (15) is used in the steepest descent method, for example, M, N, K, Q, and M / N are determined so as to minimize the objective function using M, N, K, Q, and M / N as variables.

number

[0082] The shape analysis unit 36 ​​can also store in advance a plurality of basic implicit functions used in organ approximation, such as the hyperellipse function of equation (1), the circle function of equation (7), and the function of equation (15), and set the basic implicit function to be used in organ approximation according to the type of organ extracted by the organ region extraction unit 35. In this case, the shape analysis unit 36 ​​can set the basic implicit function based on a user instruction via the input device 40. The shape analysis unit 36 ​​can also store in advance the relationship between the types of multiple organs and the basic implicit functions corresponding to each organ, and automatically set the basic implicit function corresponding to the type of organ extracted by the organ region extraction unit 35 based on the stored relationship.

[0083] In this way, by setting a basic implicit function according to the type of organ, the organ can be more accurately approximated using the implicit function, and the accuracy of the evaluation by the organ evaluation unit 37 can be improved.

[0084] Furthermore, the shape analysis unit 36 ​​can also divide the organ extracted by the organ region extraction unit 35 into a plurality of regions and approximate the organ using a basic implicit function corresponding to each of the divided regions. By combining a plurality of implicit functions in this way, the shape analysis unit 36 ​​can more accurately approximate the organ using implicit functions even when the organ has a complex shape, for example.

[0085] Furthermore, the shape analysis unit 36 ​​can recognize the shape of the organ extracted by the organ region extraction unit 35, and set an approximation method using an implicit function depending on the recognized shape of the organ. For example, if the shape of the organ extracted by the organ region extraction unit 35 is an ellipse or other shape whose shape parameters can be uniquely determined using the least squares method, the shape analysis unit 36 ​​uses the least squares method, and if not, the shape analysis unit 36 ​​uses the steepest descent method. This allows an approximation method suited to the shape of the organ to be used, thereby more accurately approximating the organ and improving the accuracy of evaluation by the organ evaluation unit 37.

[0086] Generally, in the steepest descent method, the objective function f(a, b, r) may not converge to a constant value depending on the value of the learning rate G. Therefore, when the objective function f(a, b, r) becomes larger than a predetermined value, the shape analysis unit 36 ​​can notify the user of this by displaying a message on the monitor 33, for example, and urge the user to change the value of the learning rate G. The user can confirm the notification from the shape analysis unit 36 ​​and appropriately change the value of the learning rate G, thereby determining a, b, and r that minimize the objective function f(a, b, r).

[0087] Furthermore, although it has been explained that processing such as organ extraction is performed by the organ region extraction unit 35 on the ultrasound image U generated by the image generation unit 31, processing can also be performed on ultrasound images U that have been stored in advance in the image memory 34, such as ultrasound images U stored in the image memory 34 in a past examination, rather than on ultrasound images U generated in the examination currently being performed.

[0088] Furthermore, although not shown, the ultrasound diagnostic device 1 includes an external device connection circuit that is connected to an external device by wired or wireless connection, and can store an ultrasound image U input from the external device via the external device connection circuit in the image memory 34. In this case, for example, the ultrasound image U input from the external device can be subjected to processing such as organ extraction by the organ region extraction unit 35. Examples of external devices for inputting the ultrasound image U include an external ultrasound diagnostic device, an external server device that stores the ultrasound image U, or an external storage medium that stores the ultrasound image U.

[0089] Furthermore, the test results stored in the test result memory 38 can be output to an external device (not shown), such as a workstation, via an external device connection circuit (not shown), allowing the user to check the test results using the external device, for example, after the test is completed.

[0090] Furthermore, although the present invention has been described as being applied to an ultrasound diagnostic device 1 equipped with an ultrasound probe 2, the present invention can also be applied to an imaging diagnostic device not equipped with an ultrasound probe 2, for example, an imaging diagnostic device configured with an apparatus main body 3 excluding an image generating unit 31. In this case, an organ is extracted based on an ultrasound image U stored in advance in an image memory 34 by an organ region extraction unit 35, the shape of the extracted organ is approximated by an implicit function by a shape analysis unit 36, and the organ condition is evaluated based on the shape parameters of the approximated curve by an organ evaluation unit 37. By checking the evaluation results of the organ condition obtained by the imaging diagnostic device, a user can accurately evaluate the shape of the organ and perform an accurate examination.

[0091] Furthermore, although it has been described that an organ is extracted from an ultrasound image U, the organ is approximated using an implicit function, and the condition of the organ is evaluated based on the shape parameters of the approximation curve represented by the approximated implicit function, the same processing can be performed on any medical image, such as a simple X-ray image, a CT (Computed Tomography) image, or an MRI (Magnetic Resonance Imaging) image, without being limited to the ultrasound image U. Therefore, the present invention can be applied to various diagnostic devices that perform diagnosis on any medical image, such as an X-ray image diagnostic device, a CT image diagnostic device, or an MRI image diagnostic device, in addition to the ultrasound diagnostic device 1. [Explanation of symbols]

[0092] 1 Ultrasound diagnostic device, 2 Ultrasound probe, 3 Device main body, 21 Transducer array, 22 Transmitting / receiving circuit, 23 Pulser, 24 Amplifier, 25 AD converter, 26 Beamformer, 31 Image generator, 32 Display controller, 33 Monitor, 34 Image memory, 35 Organ region extraction unit, 36 Shape analyzer, 37 Organ evaluation unit, 38 Examination result memory, 39 Main body controller, 40 Input device, 41 Processor, 51 Signal processor, 52 DSC, 53 Image processor, C1, C2 Circles, F1 to F13 Curves, P1 to P4 Points, R1 Bladder, T1, T2 Prostate, U Ultrasound image.

Claims

1. a monitor that displays an image of the subject's organs; an organ region extraction unit that extracts the organ by analyzing the image; a shape analysis unit that stores a plurality of basic implicit functions corresponding to a plurality of types of organs, selects a basic implicit function corresponding to the organ extracted by the organ region extraction unit from the plurality of basic implicit functions, and calculates an approximate implicit function that approximates the shape of the organ based on the basic implicit function; an organ evaluation unit that evaluates the condition of the organ based on shape parameters of the approximate curve represented by the approximate implicit function calculated by the shape analysis unit; Equipped with A diagnostic device in which each of the plurality of basic implicit functions has a plurality of variables, and a curve represented by each of the plurality of basic implicit functions can be deformed by changing at least one of the plurality of variables.

2. The diagnostic device according to claim 1 , wherein the organ evaluation unit evaluates the condition of the organ using the shape parameters of the approximation curve as evaluation indices.

3. the shape analysis unit uses a superellipse as the basic implicit function, The diagnostic device according to claim 2 , wherein the organ evaluation unit evaluates the condition of the organ using a shape parameter representing the deflection of the approximation curve as an evaluation index.

4. the shape analysis unit uses a superellipse as the basic implicit function, The diagnostic device according to claim 2 , wherein the organ evaluation unit evaluates the condition of the organ using a shape parameter that represents a balance of the shape of the approximation curve as an evaluation index.

5. The diagnostic device according to claim 2 , wherein the shape analysis unit uses an ellipse as the basic implicit function.

6. The diagnostic device according to claim 1 , wherein the organ evaluation unit creates a machine learning model using the shape parameters of the approximation curve, and evaluates the condition of the organ using the machine learning model.

7. The diagnostic device according to claim 6 , wherein the shape analysis unit uses a superellipse as the basic implicit function.

8. The diagnostic device according to claim 6 , wherein the shape analysis unit uses an ellipse as the basic implicit function.

9. the organ is the prostate; 9. The diagnostic device according to claim 1, wherein the organ evaluation unit evaluates the condition of the prostate by evaluating the shape of the prostate based on the shape parameters.

10. the organ is the bladder, The diagnostic device according to claim 3 , wherein the organ evaluation unit evaluates a condition of the bladder by evaluating a depression of the bladder based on the shape parameters.

11. The diagnostic device according to any one of claims 1 to 10, wherein the image is an ultrasound image.

12. The diagnostic apparatus according to claim 11, comprising: an ultrasonic probe; and an image generating unit that generates the ultrasonic image by transmitting and receiving ultrasonic beams to and from the subject using the ultrasonic probe.

13. 11. The diagnostic device according to claim 1, wherein the image is any one of an X-ray image, a computed tomography image, and a nuclear magnetic resonance image.

14. Displaying the image of the subject's organs on a monitor; extracting the organ by analyzing the image; storing a plurality of basic implicit functions corresponding to a plurality of organ types; selecting a basic implicit function corresponding to the extracted organ from the plurality of basic implicit functions; calculating an approximate implicit function that approximates the shape of the organ based on the basic implicit function; approximating the extracted shape of the organ using an implicit function; evaluating the pathology of the organ based on the shape parameters of the approximate curve represented by the calculated approximate implicit function; Each of the plurality of basic implicit functions has a plurality of variables, and the curve represented by each of the plurality of basic implicit functions can be deformed by changing at least one of the plurality of variables. A method for controlling a diagnostic device.

15. A monitor that displays an image of the subject's organs; an organ region extraction unit that extracts the organ by analyzing the image; a shape analysis unit that stores a plurality of basic implicit functions, divides the organ extracted by the organ region extraction unit into a plurality of regions, selects a basic implicit function corresponding to each of the plurality of regions from the plurality of basic implicit functions, and calculates a plurality of approximation implicit functions that approximate the organ extracted by the organ region extraction unit based on the basic implicit function corresponding to each of the plurality of regions; an organ evaluation unit that evaluates the condition of the organ based on shape parameters of an approximate curve represented by the plurality of approximate implicit functions calculated by the shape analysis unit; A diagnostic device comprising:

16. Displaying an image of the subject's organs on a monitor; extracting the organ by analyzing the image; Memorize multiple basic implicit functions, Dividing the extracted organ into a plurality of regions; selecting a basic implicit function corresponding to each of the plurality of regions from the plurality of basic implicit functions; calculating a plurality of approximate implicit functions that approximate the extracted organ based on the basic implicit functions corresponding to each of the plurality of regions; Evaluating the condition of the organ based on shape parameters of the approximate curve represented by the calculated plurality of approximate implicit functions A method for controlling a diagnostic device.

Images