Ultrasonic diagnostic device and recording medium

Abstract

The purpose of the present invention is to provide an ultrasound diagnostic device and a recording medium which are capable of easily evaluating the tissue of a subject. An irregularity analysis unit (22) executes: a feature point extraction process for extracting a plurality of feature points for each of a plurality of B-mode images acquired at different positions in a subject (14); and an evaluation process for generating information indicating irregularity of the tissue in the subject (14) on the basis of the feature points extracted for each of the plurality of B-mode images. The evaluation processing includes: performing distance evaluation processing for each image pair of a reference B-mode image and each of the other plurality of B-mode images among the plurality of B-mode images; a distance evaluation process for obtaining a feature point distance between one feature point and the other feature point in the two B-mode images for each feature point in each of the two B-mode images; and generating information indicating irregularity on the basis of the feature point distance obtained for each feature point in each image pair.

Description

Technical Field

[0001] The present invention relates to an ultrasonic diagnostic apparatus and an ultrasonic diagnostic program, and particularly to the evaluation of tissues in a subject's body. Background Art

[0002] An ultrasonic diagnostic apparatus is used to determine the benign or malignant nature of a tumor. Usually, a doctor manually determines the benign or malignant nature of a tumor, and the doctor refers to ultrasonic images such as B-mode images obtained by the ultrasonic diagnostic apparatus. A technique for diagnosing hemangiomas using a plurality of ultrasonic images acquired in time series is described in Cited Document 1.

[0003] Patent Document 1: Japanese Unexamined Patent Application Publication No. 2020-54815 Summary of the Invention

[0004] Malignant tumors are characterized by a higher degree of irregularity in internal structure than benign tumors. Focusing on this feature, it is considered to use an ultrasonic diagnostic apparatus to determine the benign or malignant nature of a tumor. However, a doctor requires a large amount of experience and a large amount of labor. Therefore, a technique for easily determining the benign or malignant nature of a tumor by an ultrasonic diagnostic apparatus is required.

[0005] An object of the present invention is to provide an ultrasonic diagnostic apparatus and an ultrasonic diagnostic program that can easily evaluate tissues in a subject.

[0006] The ultrasonic diagnostic apparatus according to the present invention is characterized by including an information processing unit that executes: a feature point extraction process for extracting a plurality of feature points from a plurality of ultrasonic images acquired at different positions in a subject; and an evaluation process for generating information indicating the irregularity of tissues in the subject based on the feature points extracted from the plurality of ultrasonic images.

[0007] In one embodiment, the evaluation process includes the following processes: a distance evaluation process is performed for each image pair of a reference ultrasonic image, which is any one of the plurality of ultrasonic images, and each of the other plurality of ultrasonic images. The distance evaluation process calculates a feature point distance between a feature point in one of the two ultrasonic images and a corresponding feature point in the other ultrasonic image for each feature point in the two ultrasonic images; and information indicating the irregularity is generated based on the feature point distances calculated for the feature points in each image pair.

[0008] In one embodiment, the distance evaluation process includes the following processes: a process of calculating an offset of the other ultrasonic image with respect to the reference ultrasonic image, which is one of the image pair, and calculating the feature point distance after correcting the position of the feature point of the other ultrasonic image based on the calculated offset.

[0009] In one embodiment, the information processing unit performs the following processing: for each feature point in the reference ultrasonic image, determine a color based on the feature point distance obtained for each of the image pairs; and generate an image in which colors are added to local regions corresponding to the feature points in the reference ultrasonic image.

[0010] In one embodiment, the evaluation processing includes the following processing: for each of the plurality of ultrasonic images, obtain a feature point shape defined by a plurality of feature points, and perform a feature point shape evaluation processing for obtaining difference information between one feature point shape and another feature point shape for each image pair of a reference ultrasonic image that is any one of the plurality of ultrasonic images and each of the other plurality of ultrasonic images; and generate information indicating the irregularity based on the difference information obtained for each of the image pairs.

[0011] In one embodiment, the feature point shape evaluation processing includes the following processing: including a process of obtaining an offset of another ultrasonic image with respect to the reference ultrasonic image that is one of the image pair, and obtaining the difference information after correcting the position of the feature point shape of the other ultrasonic image based on the obtained offset.

[0012] In one embodiment, the feature point is a point at which a feature amount representing a geometric feature appearing in the ultrasonic image becomes maximum or a point exceeding a specified threshold.

[0013] Furthermore, the ultrasonic diagnostic program according to the present invention is characterized in that a processor is caused to execute the following processing: a feature point extraction processing for extracting a plurality of feature points for each of a plurality of ultrasonic images obtained at different positions in a subject; and an evaluation processing for generating information indicating the irregularity of tissues in the subject based on the feature points extracted for each of the plurality of ultrasonic images.

[0014] In one embodiment, the evaluation processing includes the following processing: performing a distance evaluation processing for each image pair of a reference ultrasonic image that is any one of the plurality of ultrasonic images and each of the other plurality of ultrasonic images, the distance evaluation processing obtaining a feature point distance between a feature point in one of the two ultrasonic images and a feature point corresponding to the feature point in the other of the two ultrasonic images for each feature point in each of the two ultrasonic images; and generating information indicating the irregularity based on the feature point distances obtained for each feature point in each of the image pairs.

[0015] In one embodiment, the evaluation process includes the following processes: for each of the plurality of ultrasonic images, a feature point contour defined by a plurality of feature points is obtained, and a feature point contour evaluation process for obtaining difference information between one feature point contour and another feature point contour is performed for each image pair of a reference ultrasonic image, which is any one of the plurality of ultrasonic images, and each of the other plurality of ultrasonic images; and information representing the irregularity is generated based on the difference information obtained for each image pair.

[0016] Advantageous Effects of the Invention

[0017] According to the present invention, it is possible to easily evaluate the tissue of a subject. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 FIG. is a diagram showing the configuration of an ultrasonic diagnostic apparatus according to an embodiment of the present invention.

[0019] Figure 2 FIG. is a diagram showing analysis target data.

[0020] Figure 3 FIG. is a diagram showing a process of obtaining a feature point evaluation value.

[0021] Figure 4 FIG. is a flowchart of a process of obtaining a feature point evaluation value.

[0022] Figure 5 FIG. is a diagram showing a color map image.

[0023] Figure 6 FIG. is a diagram showing a process when analysis image data is sequentially generated over time.

[0024] Figure 7 FIG. is an example diagram of a process of obtaining a feature point contour for each B-mode image data.

[0025] Figure 8 FIG. is a flowchart of a process of generating analysis image data based on feature point contours obtained for each of n frames of B-mode image data.

[0026] Figure 9 FIG. is a diagram showing a case where an irregularity index determination process is performed for each of reference B-mode image data and second to n-th frame B-mode image data.

[0027] Figure 10 FIG. is a diagram showing a feature point contour obtained for reference B-mode image data and a feature point contour obtained for another frame of B-mode image data.

[0028] Figure 11This is a diagram showing an example of a process for obtaining an evaluation area of the feature point shape for each B-mode image data.

[0029] Figure 12 This is a diagram for explaining the correction process for feature points.

[0030] Symbol Explanation

[0031] 10 - Transmission unit, 12 - Ultrasonic probe, 14 - Subject, 16 - Reception unit, 18 - B-mode image generation unit, 20 - Image memory, 22 - Irregularity analysis unit, 24 - Analysis image generation unit, 26 - Display processing unit, 28 - Display, 30 - Control unit, 32 - Operation unit, 40-1 - Reference B-mode image, 40-2 - Second frame B-mode image, 40-3 - Third frame B-mode image, 40-n - nth frame B-mode image, 42 - Color map image, 50, 52 - Feature point shapes, 100 - Ultrasonic diagnostic device. Detailed Embodiments

[0032] Referring to the respective drawings, embodiments of the present invention will be described. The same components shown in multiple drawings are denoted by the same reference numerals, and their descriptions are omitted. Figure 1 The structure of the ultrasonic diagnostic device 100 according to the embodiment of the present invention is shown. The ultrasonic diagnostic device 100 includes a transmission unit 10, an ultrasonic probe 12, a reception unit 16, an information processing unit 34, a display 28, a control unit 30, and an operation unit 32. The operation unit 32 may include buttons, levers, keyboards, mice, etc. The operation unit 32 may also be a touch panel provided on the display 28. The control unit 30 performs overall control of the ultrasonic diagnostic device 100 according to the operation of the user on the operation unit 32.

[0033] The ultrasonic probe 12 includes a plurality of vibration elements. The transmission unit 10 outputs a transmission signal to each vibration element. Each vibration element converts the transmission signal into ultrasonic waves and transmits them to the subject 14. The transmission unit 10 adjusts the delay time of the transmission signal output to each vibration element so that the ultrasonic waves emitted from each vibration element reinforce each other in a specific direction. Thus, a transmission beam based on ultrasonic waves is formed in this specific direction. The transmission unit 10 changes the delay time of the transmission signal output to each vibration element to scan the transmission beam over the subject 14.

[0034] A plurality of vibration elements respectively receive the ultrasonic waves reflected by the subject 14, and convert them into electrical signals and output them to the receiving unit 16. The receiving unit 16 combines the electrical signals output from the respective vibration elements in phase to enhance the electrical signals based on the ultrasonic waves received from the direction of the transmission beam with each other to generate a received signal, and outputs the received signal to the information processing unit 34. By combining in phase, a receiving beam is formed in the ultrasonic probe 12. Here, the receiving beam refers to a directivity pattern indicating the direction of arrival of the ultrasonic waves that enhance the received signal. The received signal corresponding to the receiving beam is output from the receiving unit 16 to the information processing unit 34 as a signal for generating B-mode image data. In the following description, the transmission beam and the receiving beam are collectively referred to as the transceiver beam.

[0035] The information processing unit 34 includes a B-mode image generation unit 18, an image memory 20, an irregularity analysis unit 22, an analysis image generation unit 24, and a display processing unit 26. The information processing unit 34 may include a processor that implements the functions of each component by executing a program. In this case, the processor constitutes each component (B-mode image generation unit 18, image memory 20, irregularity analysis unit 22, analysis image generation unit 24, and display processing unit 26) by executing a program.

[0036] The received signal output from the receiving unit 16 is input to the B-mode image generation unit 18. The B-mode image generation unit 18 generates B-mode image data based on the received signals obtained for each scanning direction of the transceiver beam.

[0037] The transmission unit 10, the ultrasonic probe 12, and the receiving unit 16 repeatedly perform scanning of the transceiver beam for the subject 14. The B-mode image generation unit 18 sequentially generates B-mode image data at a predetermined frame rate over time.

[0038] The ultrasonic diagnostic apparatus 100 can operate in any one of the B-mode and the irregularity analysis mode described below.

[0039] In the operation of the B-mode, the display processing unit 26 sequentially generates an image signal representing the B-mode image over time based on the B-mode image data sequentially generated over time, and outputs the image signal to the display 28. The display 28 displays an image based on the B-mode image sequentially generated over time according to the image signal, that is, a real-time B-mode image.

[0040] In the operation of the irregularity analysis mode, the ultrasonic diagnostic apparatus 100 stores B-mode image data sequentially acquired over time, and performs an irregularity analysis process for analyzing the irregularity of the tissue within the subject 14. In the irregularity analysis process, the ultrasonic probe 12 is moved along the surface of the subject 14. The movement of the ultrasonic probe 12 can be performed by the user manually moving the ultrasonic probe 12. Also, a mechanism configured for movement can be used. The movement direction of the ultrasonic probe 12 is a direction intersecting the plane of the scanning transmission / reception beam (i.e., the observation plane for acquiring the B-mode image). The movement direction of the ultrasonic probe 12 can be a direction perpendicular to the observation plane. A series of B-mode image data (hereinafter sometimes referred to as analysis target data) sequentially generated by the B-mode image generation unit 18 over time is stored in the image memory 20. The analysis target data stored in the image memory 20 becomes the object of the irregularity analysis process.

[0041] An area for storing n frames of B-mode image data as analysis target data can be ensured in the image memory 20. In this case, when the number of B-mode images stored in the image memory 20 reaches n and a new B-mode image is generated, the earliest B-mode image data can be deleted from the image memory 20, and then the new B-mode image can be stored in the image memory 20.

[0042] The irregularity analysis unit 22 performs an irregularity analysis process on the analysis target data stored in the image memory 20 to generate irregularity information. The irregularity information is information indicating the irregularity of the tissue from which the analysis target data was acquired. The analysis image generation unit 24 generates analysis image data representing the irregularity of the tissue and outputs it to the display processing unit 26. The display processing unit 26 generates a video signal representing the analysis image based on the analysis image data and outputs the video signal to the display 28. The display 28 displays the analysis image based on the video signal.

[0043] A specific description of the irregularity analysis process will be given. Figure 2 Conceptually shows the analysis target data. The analysis target data is composed of n frames of B-mode image data representing n B-mode images F1 to Fn. The analysis target data is generated when the ultrasonic probe 12 is moved in a direction perpendicular to the observation plane. In Figure 2 the example shown, the observation plane of each B-mode image is parallel to the xy plane, and the analysis target data represents n B-mode images arranged in the z-axis direction perpendicular to the xy plane. In the following description, the B-mode image data representing B-mode images F1 to Fn may sometimes be referred to as the first frame to the nth frame B-mode image data, respectively.

[0044] The irregularity analysis unit 22 sets regions of interest in the B-mode image for each of the n frames of B-mode image data constituting the analysis target data, and obtains the distribution of feature amounts in the regions of interest. Here, the feature amount is a value representing the geometric features appearing in the B-mode image. In the present embodiment, the feature amount represents the degree of the angle of a linear image appearing in the image. In and around the angle of the linear image, the feature amount is larger than the surrounding area. As a method for obtaining the feature amount, there are Harris' corner detection method or the minimum eigenvalue method.

[0045] The irregularity analysis unit 22 extracts feature points for each of the n frames of B-mode image data constituting the analysis target data. Here, the feature point is a point where the feature amount becomes maximum or a point where the feature amount exceeds a specified threshold value. Extracting the feature point means obtaining the position coordinates where the feature point exists and acquiring the feature amount at that feature point.

[0046] The irregularity analysis unit 22 selects one frame of B-mode image data from the n frames of B-mode image data constituting the analysis target data as the reference B-mode image data. The reference B-mode image data can be, for example, the first acquired B-mode image data (the first frame of B-mode image data) with the smallest z-axis coordinate value. Hereinafter, an embodiment in which the first frame of B-mode image data is used as the reference B-mode image data will be described, but the reference B-mode image data can also be any B-mode image data selected from the second frame to the nth frame.

[0047] The irregularity analysis unit 22 obtains the feature point distances between the feature points corresponding to the feature points represented by the reference B-mode image data and the feature points in each of the other frames of B-mode image data. Here, the feature point corresponding to the feature point represented by the reference B-mode image data is the feature point on the other B-mode image that is closest to the feature point on the reference B-mode image when the reference B-mode image and the other B-mode images are overlapped. And the feature point distance is the distance between the feature point on the reference B-mode image and the feature point on the other B-mode image corresponding to that feature point when the reference B-mode image and the other B-mode images are overlapped. The irregularity analysis unit 22 further obtains the average value of the feature point distances obtained between the reference B-mode image data and each of the other frames of B-mode image data for the multiple feature points represented by the reference B-mode image data as the feature point evaluation value.

[0048] Figure 3Conceptually shows the process of obtaining the evaluation value of feature points. Five feature points C are shown in the reference B-mode image 40-1. Arrows are drawn from each feature point C on the reference B-mode image 40-1 to the corresponding feature points in the second frame B-mode image 40-2. The feature point distance between the feature point C represented by the reference B-mode image data and the feature point on the second frame B-mode image 40-2 corresponding to this feature point C is obtained. For the feature point distance of each feature point C on the reference B-mode image 40-1, the same is also obtained for the third frame B-mode image 40-3 to the nth frame B-mode image 40-n respectively. Then, the average value of the feature point distances obtained for the second frame to the nth frame is obtained as the evaluation value of the feature point C.

[0049] There are k feature points on the reference B-mode image, and the number of the identified feature point is set as j (j = 1 to k). And, the number i of the identified reference B-mode image is set as i = 1, and the number i of the identified other B-mode images is set as i = 2 to n. The feature point distance between the jth feature point on the reference B-mode image and the jth feature point on the ith B-mode image is set as d(j, i), and the evaluation value Dj of the feature point for j = 1 to k is represented by (Equation 1). The symbol ∑ in (Equation 1) represents obtaining the added value for i = 2 to n.

[0050] (Equation 1) Dj = (1 / (n - 1))·∑ d(j, i)

[0051] The irregularity analysis unit 22 can obtain the average value of the evaluation values Dj of the feature points for j = 1 to k as the tissue evaluation value D. The tissue evaluation value D represents the irregularity of the tissue, and the greater the irregularity, the greater the value.

[0052] Figure 4 Shows a flowchart of the process of obtaining the evaluation value of feature points. In this flowchart, the B-mode image represented by the ith frame B-mode image data is simply referred to as the ith B-mode image. The irregularity analysis unit 22 selects the jth feature point Pj (j = 1 to k) among the k feature points on the reference B-mode image (S101). Among them, the initial value of j is 1.

[0053] The irregularity analysis unit 22 searches for the feature point closest to the feature point Pj from all the feature points on the ith (i = 2 to n) B-mode image, obtains the feature point distance d(i, j) (S102), and stores the feature point distance d(i, j) (S103). Among them, the initial value of i is 2.

[0054] The irregularity analysis unit 22 determines whether i is n (S104). When i is not n, i is incremented by 1 (S108), and the process returns to the process of step S102. When i is n, the irregularity analysis unit 22 obtains the average value of the feature point distances d(i, j) for i = 2 to n as the feature point evaluation value Dj (S105).

[0055] The irregularity analysis unit 22 determines whether j is k (S106). When j is not k, j is incremented by 1 (S109), and the process returns to the process of step S101. When j is k, the irregularity analysis unit 22 obtains the average value of the feature point evaluation values Dj for j = 1 to k as the tissue evaluation value D (S107).

[0056] The irregularity analysis unit 22 outputs the irregularity information and the reference B-mode image data to the analysis image generation unit 24. Here, the irregularity information may include the feature point evaluation values Dj (j = 1 to k) obtained for k feature points on the reference B-mode image, the position information of the k feature points, the feature amounts of the k feature points, and the tissue evaluation value D. The analysis image generation unit 24 generates analysis image data representing the analysis image. The analysis image may be an image representing the B-mode image represented by the reference B-mode image data, the positions of the respective feature points, the feature amounts at the respective feature points on the reference B-mode image, the feature point evaluation values at the respective feature points on the reference B-mode image, the tissue evaluation value D, and the like.

[0057] The analysis image generation unit 24 outputs the analysis image data to the display processing unit 26. The display processing unit 26 generates a video signal representing the analysis image based on the analysis image data, and outputs the video signal to the display 28. The display 28 displays the analysis image according to the video signal. The user can determine whether the tissue of the subject 14 is irregular by referring to the analysis image displayed on the display 28. And, based on this determination, it is possible to determine whether there is a tumor or the like in the subject 14. Furthermore, in the case where it is determined that there is a tumor, the user can determine its malignancy.

[0058] In this way, the irregularity analysis unit 22 performs: a feature point extraction process of extracting a plurality of feature points respectively from a plurality of B-mode images (ultrasonic images) obtained at different positions in the subject 14; and an evaluation process of generating information representing the irregularity of the tissue in the subject 14 based on the feature points respectively extracted from the plurality of B-mode images.

[0059] The evaluation process includes the following processes: performing a distance evaluation process on each image pair of a reference B-mode image, which is any one of a plurality of B-mode images, and each of the other plurality of B-mode images, the distance evaluation process obtaining a feature point distance between a feature point in one of the two B-mode images and a corresponding feature point in the other of the two B-mode images for each feature point in each of the two B-mode images; and generating information representing irregularity based on the feature point distances obtained for each feature point in each image pair.

[0060] The ultrasonic diagnostic program for performing the above-mentioned feature point extraction process and evaluation process can be read into the processor of the irregularity analysis unit 22 that constitutes the information processing unit 34. That is, the ultrasonic diagnostic program causes the processor to perform the above-mentioned feature point extraction process and evaluation process.

[0061] As an example of the analysis image, Figure 5 a color map image 42 is shown. In this example, the positions of the feature points and the feature point evaluation values are shown together with the tissue T by circles centered on the positions of the respective feature points. The radius of the circle represents the feature point evaluation value. Also, the feature point evaluation value can be represented by the color inside the circle. For example, the larger the feature point evaluation value, the color with a longer wavelength (reddish color) is added inside the circle, and the smaller the feature point evaluation value, the color with a shorter wavelength (bluish color) is added inside the circle. Additionally, instead of the display where the radius of the circle is determined by the feature point evaluation value, a display where the radius of the circle is determined by the feature quantity in the reference B-mode image can be performed.

[0062] In this way, the analysis image generation unit 24 determines the color for each feature point in the reference B-mode image (reference ultrasonic image) based on the feature point distances obtained for each image pair, and generates an image in which a color is added inside the circle (local area) corresponding to each feature point in the reference B-mode image.

[0063] The ultrasonic diagnostic program read into the processor that constitutes the analysis image generation unit 24 causes the processor to perform the following process: for each feature point in the reference B-mode image (reference ultrasonic image), determine the color based on the feature point distances obtained for each image pair, and generate an image in which a color is added inside the circle (local area) corresponding to each feature point in the reference B-mode image.

[0064] Figure 6 The process of generating analysis image data sequentially over time is conceptually shown. If the n-th frame B-mode image data is stored in the image memory 20, the irregularity analysis unit 22 performs an irregularity analysis process on the first frame to the n-th frame B-mode image data stored by tracing back to the n - 1-th frame, and generates analysis image data 1.

[0065] If the (n + 1)-th frame of B-mode image data is stored in the image memory 20, the irregularity analysis unit 22 generates analysis image data 2 based on the second to (n + 1)-th frame of B-mode image data stored by tracing back to the (n - 1)-th frame. …… If the (n + L)-th frame of B-mode image data is stored in the image memory 20, the irregularity analysis unit 22 generates analysis image data L based on the (1 + L)-th to (n + L)-th frame of B-mode image data stored by tracing back to the (n - 1)-th frame.

[0066] In this way, every time new B-mode image data is stored in the image memory 20, the irregularity analysis unit 22 sequentially generates new analysis image data based on a series of n frames of B-mode image data (analysis target data) traced back from the B-mode image data to the (n - 1)-th frame.

[0067] Next, the irregularity analysis process based on the feature point shape will be described. The feature point shape refers to a figure formed by connecting the closest feature points among multiple feature points on the B-mode image with lines.

[0068] The irregularity analysis unit 22 performs a process of extracting feature points for each of the n frames of B-mode image data constituting the analysis target data. The irregularity analysis unit 22 obtains feature point shape data representing the feature point shape for each B-mode image data.

[0069] Figure 7 Conceptually shows an example of the process of obtaining the feature point shape for each B-mode image data. First, the irregularity analysis unit 22 searches for the first feature point C1 with the largest feature amount. The irregularity analysis unit 22 searches for the second feature point C2 closest to the first feature point C1 and obtains connection line data representing the line connecting the first feature point C1 and the second feature point C2. Hereinafter, it is assumed that the irregularity analysis unit 22 performs the same process and searches out a total of k feature points. The irregularity analysis unit 22 obtains connection line data representing each line connecting from the first feature point C1 to the k-th feature point Ck in sequence and connecting the k-th feature point Ck and the first feature point C1. The connection line data representing each line connecting from the first feature point C1 to the k-th feature point Ck in sequence and connecting the k-th feature point Ck and the first feature point C1 is the feature point shape data representing the feature point shape.

[0070] Figure 8 Shows a flowchart of the process of generating analysis image data based on the feature point shapes obtained for each of the n frames of B-mode image data. The irregularity analysis unit 22 assigns 1 as an evaluation index to each predetermined evaluation position inside the feature point shape obtained for the reference B-mode image data, and assigns 0 as an evaluation index to each evaluation position outside the feature point shape (S201).

[0071] Further, for each of the other frames, the irregularity analysis unit 22 establishes a correspondence relationship between -1 and each evaluation position inside the feature point outline obtained for the B-mode image data as an evaluation index, and sets 0 as the evaluation index for each evaluation position outside the feature point outline (S202).

[0072] The irregularity analysis unit 22 adds the evaluation index assigned to each evaluation position on the reference B-mode image and the evaluation index assigned to each evaluation position on another B-mode image for each evaluation position, and obtains an irregularity index for each evaluation position. The evaluation positions with a positive irregularity index are in the region inside the feature point outline on the reference B-mode image and outside the feature point outline on another B-mode image. The evaluation positions with a negative irregularity index are in the region outside the feature point outline on the reference B-mode image and inside the feature point outline on another B-mode image. The evaluation positions with an irregularity index of 0 are in the region inside the feature point outline on the reference B-mode image and inside the feature point outline on another B-mode image. The irregularity analysis unit 22 performs this kind of irregularity index determination process on each of the other n - 1 B-mode image data that are not the reference B-mode image data (S203).

[0073] Figure 9 conceptually shows the case where the irregularity index determination process is performed on each of the first-frame B-mode image data (i.e., the reference B-mode image data) and the second-frame to n-frame B-mode image data.

[0074] Figure 10 conceptually shows the feature point outline 50 obtained for the reference B-mode image data and the feature point outline 52 obtained for another B-mode image data. At the evaluation positions with a positive irregularity index, in another B-mode image, the feature point outline is smaller than the feature point outline in the reference B-mode image. And at the evaluation positions with a negative irregularity index, in another B-mode image, the feature point outline is larger than the feature point outline in the reference B-mode image. At the evaluation positions with an irregularity index of 0, in another B-mode image, the feature point outline neither expands nor contracts with respect to the feature point outline in the reference B-mode image.

[0075] The irregularity index at each evaluation position can be said to be difference information indicating the difference between the feature point outline 50 obtained for the reference B-mode image data and the feature point outline 52 obtained for another B-mode image data.

[0076] Return to Figure 8 The irregularity analysis unit 22 separately adds up the irregularity indices obtained for the n - 1 frames for each evaluation position and sums them up, and obtains an irregularity evaluation value for each evaluation position (S204).

[0077] The irregularity analysis unit 22 outputs irregularity information including the irregularity evaluation values obtained for each evaluation position and the reference B-mode image data to the analysis image generation unit 24. The analysis image generation unit 24 generates analysis image data representing the analysis image (S205). The analysis image is an image in which colors corresponding to the irregularity evaluation values are added to each evaluation position and the periphery of each evaluation position on the B-mode image represented by the reference B-mode image data. Here, the periphery of the evaluation position refers to, for example, a region within a circle determined with a predetermined radius centered on the evaluation position. Also, the periphery of the evaluation position may be a polygonal region including the evaluation position. It is possible that the larger the absolute value of the irregularity evaluation value, the longer the wavelength of the added color (reddish color), and the smaller the absolute value of the irregularity evaluation value, the shorter the wavelength of the added color (bluish color).

[0078] The irregularity analysis unit 22 can obtain a tissue evaluation value D representing the irregularity of the living tissue from the analysis target data through the process shown below. The irregularity analysis unit 22 performs a correlation operation on the feature point outlines obtained for the reference B-mode image data and the feature point outlines obtained for another frame of B-mode image data to obtain a correlation value. The correlation value represents the degree of approximation of the two feature point outlines. The larger the correlation value, the more similar the two feature point outlines. The irregularity analysis unit 22, for example, obtains the reciprocal of the average value of the correlation values obtained for the second frame to the nth frame as the tissue evaluation value D. In addition to the reciprocal of the average value of the correlation values, the tissue evaluation value D can also be obtained according to other relationships in which the smaller the average value of the correlation values, the larger it is. The more different the feature point outline obtained for the reference B-mode image data is from the feature point outlines obtained for each other frame of B-mode image data, the smaller the average value of the correlation values, and the larger the tissue evaluation value D.

[0079] The irregularity analysis unit 22 outputs irregularity information including the tissue evaluation value D to the analysis image generation unit 24. The analysis image generation unit 24 generates analysis image data representing the tissue evaluation value D and outputs it to the display processing unit 26. The display processing unit 26 generates a video signal representing the analysis image from the analysis image data and outputs the video signal to the display 28. The display 28 displays the analysis image according to the video signal.

[0080] In addition, the irregularity analysis unit 22 can also obtain a tissue evaluation value D representing the irregularity of the tissue through another process shown below. The irregularity analysis unit 22 respectively obtains the feature point outlines for the n frames of B-mode image data constituting the analysis target data, and obtains the evaluation area for the feature point outlines obtained for each B-mode image data.

[0081] Figure 11Conceptually shows an example of a process for obtaining an evaluation area of a feature point shape for B-mode image data. The irregularity analysis unit 22 extracts feature points C from each B-mode image data. The irregularity analysis unit 22 obtains the coordinates of the representative point R of the region where the extracted multiple feature points C exist. The x-coordinate value of the representative point R is, for example, a value obtained by adding the x-coordinate value (x-coordinate maximum value) of the feature point with the largest x-coordinate value among the multiple feature points and the x-coordinate value (x-coordinate minimum value) of the feature point with the smallest x-coordinate value among the multiple feature points and dividing the sum by 2. The y-coordinate value of the representative point R is, for example, a value obtained by adding the y-coordinate value (y-coordinate maximum value) of the feature point with the largest y-coordinate value among the multiple feature points and the y-coordinate value (y-coordinate minimum value) of the feature point with the smallest y-coordinate value among the multiple feature points and dividing the sum by 2. The irregularity analysis unit 22 searches for the feature point Cm with the largest distance from the representative point R, and obtains the area of the circle CM with the distance between the representative point R and the searched feature point Cm as the radius as the evaluation area.

[0082] The irregularity analysis unit 22 obtains the absolute value of the value obtained by subtracting the evaluation area obtained for the reference B-mode image data from the evaluation area obtained for another frame of B-mode image data as the area difference value. The area difference value represents the degree of difference between the two feature point shapes. The greater the area difference, the more different the two feature point shapes. The irregularity analysis unit 22, for example, obtains the average value of the area difference values obtained for each of the second frame to the nth frame as the tissue evaluation value D. The more different the evaluation area obtained for the reference B-mode image data is from the evaluation areas obtained for the other frames of B-mode image data, the greater the area difference value, and the greater the tissue evaluation value D.

[0083] Thus, the evaluation process performed by the irregularity analysis unit 22 includes: a first process of obtaining the feature point shapes defined by multiple feature points for multiple B-mode images (ultrasound images) respectively, and performing a feature point shape evaluation process of obtaining the difference information between one feature point shape and another for each image pair of the reference B-mode image, which is any one of the multiple B-mode images, and the other multiple B-mode images; and a second process of generating information representing irregularity based on the difference information obtained for each image pair. Here, the difference information includes the above-mentioned irregularity index, correlation value, area difference value, etc.

[0084] The ultrasound diagnosis program read into the processor constituting the irregularity analysis unit 22 causes the processor to execute the above-mentioned first process and second process.

[0085] Regarding each of the above-described embodiments, it has been described that the positions of the tissues appearing in the B-mode images of each frame are not shifted. In the diagnosis in the irregularity analysis mode, in another B-mode image with respect to the first B-mode image, the positions of the tissues appearing in the B-mode image may be shifted due to hand tremors or the like. Therefore, in order to compensate for the shift generated in the positions of the tissues, correction processing may be performed on the positions of the feature points of the second frame to the nth frame.

[0086] The irregularity analysis unit 22 performs a correlation operation while changing the position of another B-mode image with respect to the reference B-mode image, obtains the position of another B-mode image at which the correlation value becomes a maximum value, and obtains the position shift amount of the tissue appearing in another B-mode image with respect to the tissue appearing in the reference B-mode image as a position shift vector.

[0087] That is, the irregularity analysis unit 22 separately obtains position shift vectors for the second frame to the nth frame. The irregularity analysis unit 22 subtracts the position shift vectors separately obtained for the second frame to the nth frame from the position coordinates of the feature points separately obtained for the second frame to the nth frame. Thereby, the irregularity analysis unit 22 obtains new position coordinates of the feature points separately obtained for the second frame to the nth frame, and corrects the position coordinates of each feature point.

[0088] Figure 12 (a) shows a reference B-mode image obtained on the xy plane. Tissue T1 appears in the reference B-mode image, and a region X1 where the feature points extracted from the reference B-mode image exist is shown. Figure 12 (b) shows the displaced tissue T2 and the feature point Cx extracted from the tissue T2 in the second-frame B-mode image. Also shown is a position shift vector V indicating the shift of the tissue T2 with respect to the tissue T1.

[0089] Figure 12 (c) shows the case where the position coordinates of the feature point Cx are corrected by subtracting the position shift vector V from the position coordinates of the feature point Cx set for the tissue T2.

[0090] The above-described distance evaluation process performed by the irregularity analysis unit 22 may include a first correction process, which includes a process of obtaining the shift of the other with respect to the reference B-mode image (reference ultrasonic image) that is one of the image pair, and obtaining the feature point distance after correcting the position of the feature point of the other according to the obtained shift. Also, the above-described feature point shape evaluation process may include a second correction process, which includes a process of obtaining the shift of the other with respect to the reference ultrasonic image that is one of the image pair, and obtaining the difference information after correcting the position of the feature point shape of the other according to the obtained shift.

[0091] The ultrasonic diagnosis program read into the processor that constitutes the irregularity analysis unit 22 enables the processor to execute the above-described first correction process and second correction process.

[0092] In this way, by correcting the positions of the feature points, when the ultrasonic probe 12 is moved in the irregularity analysis mode, even when there is an offset from the ideal linear movement trajectory, it is possible to obtain the positions of appropriate feature points.

Claims

1. An ultrasonic diagnostic apparatus, characterized in that, comprises an information processing unit that performs: feature point extraction processing for respectively extracting a plurality of feature points from a plurality of ultrasonic images obtained at different positions in a subject; and evaluation processing for generating information representing the irregularity of tissues in the subject based on the feature points respectively extracted from the plurality of ultrasonic images.

2. The ultrasonic diagnostic apparatus according to claim 1, wherein the evaluation processing includes the following processing: performing distance evaluation processing on each image pair of a reference ultrasonic image as any one of the plurality of ultrasonic images and each of the other plurality of ultrasonic images, the distance evaluation processing obtaining a feature point distance between a feature point in one of the two ultrasonic images and a corresponding feature point in the other of the two ultrasonic images for each feature point in each of the two ultrasonic images; and generating information representing the irregularity based on the feature point distances obtained for each feature point in each of the image pairs.

3. The ultrasonic diagnostic apparatus according to claim 2, wherein the distance evaluation processing includes the following processing: including a process of obtaining an offset of the other with respect to the reference ultrasonic image as one of the image pairs, and obtaining the feature point distance after correcting the position of the feature point of the other based on the obtained offset.

4. The ultrasonic diagnostic apparatus according to claim 2 or 3, wherein the information processing unit performs the following processing: determining a color for each feature point in the reference ultrasonic image based on the feature point distances obtained for each of the image pairs; and generating an image with colors added to local regions corresponding to each feature point in the reference ultrasonic image.

5. The ultrasonic diagnostic apparatus according to claim 1, wherein the evaluation processing includes the following processing: respectively obtaining feature point outlines defined by a plurality of feature points for the plurality of ultrasonic images, and performing feature point outline evaluation processing for obtaining difference information between the feature point outline of one and the feature point outline of the other on each image pair of a reference ultrasonic image as any one of the plurality of ultrasonic images and each of the other plurality of ultrasonic images; and generating information representing the irregularity based on the difference information obtained for each of the image pairs.

6. The ultrasonic diagnostic apparatus according to claim 5, wherein the feature point outline evaluation processing includes the following processing: including a process of obtaining an offset of the other with respect to the reference ultrasonic image as one of the image pairs, and obtaining the difference information after correcting the position of the feature point outline of the other based on the obtained offset.

7. The ultrasonic diagnostic apparatus according to claim 1, wherein the feature point is a point at which a feature quantity representing a geometric feature appearing in the ultrasonic image becomes maximum or exceeds a specified threshold.

8. A processor-readable recording medium storing an ultrasonic diagnosis program, characterized in that, The ultrasonic diagnostic program causes a processor to perform the following processing: Feature point extraction process for extracting a plurality of feature points from a plurality of ultrasonic images obtained at different positions in a subject; and Evaluation process for generating information representing the irregularity of tissues in the subject based on the feature points extracted from the plurality of ultrasonic images respectively.

9. The recording medium according to claim 8, wherein the evaluation process includes the following processes: Performing a distance evaluation process on each image pair of a reference ultrasonic image that is any one of the plurality of ultrasonic images and each of the other plurality of ultrasonic images, the distance evaluation process calculating a feature point distance between a feature point in one of the two ultrasonic images and a corresponding feature point in the other of the two ultrasonic images for each feature point in each of the two ultrasonic images; and Generating information representing the irregularity based on the feature point distances calculated for each feature point in each of the image pairs.

10. The recording medium according to claim 8, wherein the evaluation process includes the following processes: Respectively obtaining a feature point contour defined by a plurality of feature points for the plurality of ultrasonic images, and performing a feature point contour evaluation process for obtaining difference information between the feature point contour of one and the feature point contour of the other on each image pair of a reference ultrasonic image that is any one of the plurality of ultrasonic images and each of the other plurality of ultrasonic images; and Generating information representing the irregularity based on the difference information obtained for each of the image pairs.

Images