Information processing device, information processing method, and program

Abstract

The present invention provides an information processing device, an information processing method, and a program each of which can assist with determining whether a blood vessel is an artery or a vein. This information processing device carries out a process for transmitting ultrasonic beams (UB) from an oscillator array (13) into a living body (30) and displaying, on a display device (21), an ultrasonic image (U) generated by receiving ultrasonic echoes produced in the living body (30), said information processing device comprising: a blood vessel group detection unit (72) that detects, in the ultrasonic image (U), a blood vessel group region (Ra) including a blood vessel group, which is a group of three or more blood vessels (B); and a highlighting unit (54) that highlights the blood vessel group region (Ra) in the ultrasonic image (U).

Description

[Technical Field]

[0001] The technology disclosed herein relates to an information processing device, an information processing method, and a program. [Background technology]

[0002] Ultrasound diagnostic devices have been known for obtaining images of the inside of a subject. Ultrasound diagnostic devices generally have an ultrasound probe equipped with a transducer array in which multiple ultrasound transducers are arranged. This ultrasound probe, when placed in contact with the body surface of the subject, transmits ultrasound beams from the transducer array toward the inside of the subject, and receives ultrasound echoes from the subject with the transducer array. Electrical signals corresponding to the ultrasound echoes are thereby acquired. Furthermore, the ultrasound diagnostic device processes the acquired electrical signals to generate an ultrasound image of the corresponding part of the subject.

[0003] Incidentally, a procedure (so-called echo-guided puncture) is known in which a puncture needle is inserted into a blood vessel of a subject while observing the inside of the subject using an ultrasound diagnostic device. In the echo-guided puncture, the surgeon usually needs to check the ultrasound image to grasp the position, shape, etc. of the blood vessels contained in the ultrasound image, but accurately grasping these requires a certain level of skill. Therefore, it has been proposed to automatically detect blood vessels contained in the ultrasound image and present the detected blood vessels to the surgeon (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Special Publication No. 2017-524455 Summary of the Invention [Problem to be solved by the invention]

[0005] When performing puncture, the surgeon must accurately determine whether the blood vessel is an artery or a vein based on the ultrasound image. This determination will be referred to as arteriovenous determination hereinafter.

[0006] It is possible to determine whether a blood vessel is an artery or vein by performing information processing such as image analysis based on ultrasound images. However, if blood vessels are determined to be arteries or veins individually, it is easy to make an error in the determination if the shapes of arteries and veins are similar.

[0007] The technology of the present disclosure aims to provide an information processing device, an information processing method, and a program that can assist in determining whether a blood vessel is an artery or a vein. [Means for solving the problem]

[0008] The information processing device disclosed herein is an information processing device that performs processing to transmit an ultrasonic beam from a transducer array toward a living body and display an ultrasonic image generated by receiving ultrasonic echoes generated within the living body on a display device, and is equipped with a vascular aggregate detection unit that detects a vascular aggregate region containing a vascular aggregate formed by a collection of three or more blood vessels from within the ultrasound image, and a highlighting display unit that highlights the vascular aggregate region within the ultrasound image.

[0009] It is preferable to further include an artery / vein determining unit that determines whether each blood vessel included in the blood vessel assembly region is an artery or a vein based on the feature amount of each blood vessel in the blood vessel assembly region.

[0010] The artery / vein determining unit preferably makes the determination based on at least one of the following characteristic quantities: blood vessel diameter, the amount of displacement of the blood vessel from the center of the blood vessel assembly region, and the circularity of the blood vessel.

[0011] It is preferable to further include a blood vessel unit detection unit that detects a blood vessel unit region containing a blood vessel unit from within the ultrasound image and determines whether the blood vessel contained in the detected blood vessel unit region is an artery or a vein.

[0012] It is preferable to further include a correction unit that corrects the result of the artery / vein determination made by the blood vessel individual detection unit based on the result of the artery / vein determination made by the artery / vein determination unit.

[0013] The correction unit preferably compares the certainty of the artery / vein determination made by the single blood vessel detection unit with the certainty of the artery / vein determination made by the artery / vein determination unit, and selects the determination result with the higher certainty.

[0014] It is preferable that the highlighting unit displays each of the blood vessels included in the blood vessel aggregate region in a manner that allows arteries and veins to be distinguished based on the correction result by the correction unit.

[0015] The highlighting unit preferably displays the degree of certainty of the determination result selected by the correcting unit on the display device.

[0016] It is preferable that the highlighting unit displays a message on the display device to call attention when the confidence level for the determination result selected by the correction unit is lower than a certain value.

[0017] The information processing method disclosed herein is an information processing method that performs processing to transmit an ultrasonic beam from a transducer array toward a living body, receive ultrasonic echoes generated within the living body, and display an ultrasonic image generated by the received ultrasonic beam on a display device. The method detects a vascular aggregate region containing a vascular aggregate consisting of three or more blood vessels from within the ultrasonic image, and highlights the detected vascular aggregate region within the ultrasonic image.

[0018] The program disclosed herein is a program that causes a computer to execute a process of transmitting an ultrasonic beam from a transducer array toward a living body and displaying an ultrasonic image generated by receiving ultrasonic echoes generated within the living body on a display device, and causes the computer to execute a process of detecting a vascular aggregate region containing a vascular aggregate consisting of three or more blood vessels from within the ultrasonic image and highlighting the detected vascular aggregate region within the ultrasonic image. [Effects of the Invention]

[0019] According to the technology of the present disclosure, it is possible to provide an information processing device, an information processing method, and a program that are capable of assisting in determining whether a blood vessel is an artery or a vein. [Brief explanation of the drawings]

[0020] [Figure 1] 1 is an external view showing an example of the configuration of an ultrasound diagnostic apparatus according to a first embodiment. [Figure 2] FIG. 1 is a diagram illustrating an example of an ultrasound-guided puncture method. [Figure 3] FIG. 1 is a block diagram showing an example of the configuration of an ultrasound diagnostic apparatus. [Figure 4] FIG. 2 is a block diagram showing an example of the configuration of a receiving circuit. [Figure 5] FIG. 2 is a block diagram showing an example of the configuration of an image generating unit. [Figure 6] FIG. 2 is a block diagram showing an example of the configuration of an image analysis unit. [Figure 7] 10A and 10B are diagrams illustrating an example of a single blood vessel detection process. [Figure 8] FIG. 10 is a diagram illustrating an example of a learning phase for learning a simple blood vessel detection model. [Figure 9] 10A and 10B are diagrams illustrating an example of a blood vessel aggregate detection process. [Figure 10] FIG. 10 is a diagram illustrating an example of a learning phase for learning a blood vessel aggregate detection model. [Figure 11] 10A and 10B are diagrams illustrating an example of an artery / vein determination process. [Figure 12] 10A and 10B are diagrams illustrating an example of an artery / vein determination process in more detail. [Figure 13] FIG. 10 is a diagram illustrating an example of a correction process. [Figure 14] FIG. 10 is a diagram illustrating an example of highlighting processing. [Figure 15] 10 is a flowchart illustrating an example of the operation of the ultrasound diagnostic apparatus. [Figure 16] FIG. 10 is a diagram illustrating a modified example of the artery / vein determination process. [Figure 17] FIG. 10 is a diagram illustrating a modified example of the artery / vein determination process. [Figure 18]FIG. 10 is a diagram showing a modified example of highlighting. [Figure 19] FIG. 10 is a diagram showing an example of displaying a score. [Figure 20] FIG. 10 is a diagram showing an example of displaying a message to alert the surgeon. [Figure 21] 10 is a flowchart illustrating an example of a threshold setting process. [Figure 22] FIG. 10 is a diagram illustrating an example of thresholds set by a threshold setting process. [Figure 23] FIG. 10 is a diagram showing an example in which a blood vessel aggregate region is not displayed. [Figure 24] FIG. 10 is a diagram showing a first modified example of an ultrasonic diagnostic apparatus. [Figure 25] FIG. 10 is a diagram showing a second modified example of the ultrasonic diagnostic device. DETAILED DESCRIPTION OF THE INVENTION

[0021] Hereinafter, embodiments of the technology of the present disclosure will be described with reference to the accompanying drawings. The following description of the components will be based on a representative embodiment of the present invention, but the technology of the present disclosure is not limited to such an embodiment.

[0022] [First embodiment] FIG. 1 shows an example of the configuration of an ultrasound diagnostic device 2 according to the technique of the present disclosure. The ultrasound diagnostic device 2 according to this embodiment is composed of an ultrasound probe 10 and a device main body 20. The ultrasound probe 10 is held by an operator and is brought into contact with the surface of a living body that is the measurement target. The ultrasound probe 10 transmits and receives ultrasound beams to and from the interior of the living body.

[0023] The device main body 20 is, for example, a smartphone or a tablet terminal. When a program such as application software is installed on the device main body 20, the device main body 20 performs imaging of the signal output from the ultrasound probe 10. The ultrasound probe 10 and the device main body 20 communicate with each other wirelessly (i.e., wirelessly) via, for example, WiFi or Bluetooth (registered trademark). The device main body 20 is not limited to a mobile terminal such as a smartphone or a tablet terminal, but may also be a PC (Personal Computer) or the like. The device main body 20 is an example of an "information processing device" according to the technology of the present disclosure.

[0024] The ultrasonic probe 10 has a housing 11. The housing 11 is composed of an array housing portion 11A and a grip portion 11B. The array housing portion 11A houses the transducer array 13 (see FIG. 3). The grip portion 11B is connected to the array housing portion 11A and is held by the surgeon. For the sake of explanation, the direction from the grip portion 11B toward the array housing portion 11A is defined as the +Y direction, the width direction of the ultrasonic probe 10 perpendicular to the Y direction is defined as the X direction, and the direction perpendicular to the X and Y directions (i.e., the thickness direction of the ultrasonic probe 10) is defined as the Z direction.

[0025] An acoustic lens is disposed at the +Y direction end of the array housing portion 11A. A so-called acoustic matching layer (not shown) is disposed on the transducer array 13, and an acoustic lens is disposed on the acoustic matching layer. The multiple transducers included in the transducer array 13 are arranged linearly along the X direction. That is, the ultrasonic probe 10 of this embodiment is a linear type, and transmits an ultrasonic beam UB linearly. Note that the ultrasonic probe 10 may be a convex type in which the transducer array 13 is arranged on a convex curved surface. In this case, the ultrasonic probe 10 transmits the ultrasonic beam UB radially. The ultrasonic probe 10 may also be a sector type.

[0026] Furthermore, linear guide markers M extending along the Y direction are attached to the outer periphery of the array housing portion 11A. The guide markers M are used as a guide when the operator brings the ultrasonic probe 10 into contact with the living body.

[0027] The device main body 20 has a display device 21 for displaying an ultrasound image based on a signal transmitted from the ultrasound probe 10. The display device 21 is, for example, a display device such as an organic electroluminescence (EL) display or a liquid crystal display. A touch panel is incorporated in the display device 21. The surgeon can perform various operations on the device main body 20 using the touch panel.

[0028] FIG. 2 is a diagram illustrating an example of an ultrasound-guided puncture method. As shown in FIG. 2, an ultrasound probe 10 is used when an operator inserts a puncture needle 31 into a blood vessel B in a living body 30 while checking an ultrasound image displayed on the device main body 20. The living body 30 is, for example, a human arm. The ultrasound probe 10 is placed against the surface of the living body 30 so that the width direction (i.e., the X direction) of the ultrasound probe 10 crosses the running direction of the blood vessel B. This procedure is called the short-axis method (or cross-sectional method). A cross section of the blood vessel B is displayed on the ultrasound image. The operator punctures, for example, a vein from one or more blood vessels B displayed on the ultrasound image.

[0029] After detecting blood vessels from the ultrasound image, the device main body 20 determines whether the blood vessels are arteries or veins, and displays the results of the artery and vein determination within the ultrasound image displayed on the display device 21, thereby assisting the surgeon in performing puncture.

[0030] 3 shows an example of the configuration of the ultrasound diagnostic device 2. The ultrasound probe 10 has a transducer array 13, a transmission / reception circuit 14, and a communication unit 15. The transmission / reception circuit 14 includes a transmission circuit 16 and a reception circuit 17. The transmission circuit 16 and the reception circuit 17 are each connected to the transducer array 13. In addition, the transmission / reception circuit 14 inputs and outputs signals to and from a processor 25 of the device main body 20 via the communication unit 15.

[0031] The transducer array 13 has a plurality of transducers (not shown) arranged one-dimensionally or two-dimensionally. Each of these transducers transmits an ultrasonic beam UB in accordance with a drive signal supplied from the transmission circuit 16, and receives ultrasonic echoes from the living body 30. The transducer outputs a signal based on the received ultrasonic echoes. The transducer is configured, for example, by forming electrodes on both ends of a piezoelectric body. The piezoelectric body is made of piezoelectric ceramics such as PZT (Lead Zirconate Titanate), polymer piezoelectric elements such as PVDF (Poly Vinylidene Di Fluoride), and piezoelectric single crystals such as PMN-PT (Lead Magnesium Niobate-Lead Titanate).

[0032] The transmission circuit 16 includes, for example, multiple pulse generators. The transmission circuit 16 adjusts the delay amount of the drive signal based on a transmission delay pattern selected in response to a control signal transmitted from the processor 25 of the device main body 20, and supplies the drive signal to the multiple transducers included in the transducer array 13. The transmission circuit 16 adjusts the delay amount of the drive signal so that ultrasonic waves transmitted from the multiple transducers form an ultrasonic beam UB. The drive signal is a pulsed or continuous wave voltage signal. When the drive signal is applied, the transducer transmits pulsed or continuous wave ultrasonic waves by expanding and contracting. The ultrasonic waves transmitted from the multiple transducers are combined to form an ultrasonic beam UB as a composite wave.

[0033] The ultrasonic beam UB transmitted into the living body 30 is reflected by a part such as a blood vessel B in the living body 30, and becomes an ultrasonic echo, which propagates toward the transducer array 13. The ultrasonic echo propagating toward the transducer array 13 in this manner is received by the multiple transducers that make up the transducer array 13. The transducers expand and contract upon receiving the ultrasonic echo, generating an electrical signal. The electrical signal generated by the transducer is output to the receiving circuit 17.

[0034] The receiving circuit 17 generates sound ray signals by processing the electrical signals output from the transducer array 13 in accordance with a control signal transmitted from the processor 25 of the device main body 20. As an example, as shown in Fig. 4, the receiving circuit 17 is configured by connecting an amplifier 41, an A / D (Analog to Digital) converter 42, and a beam former 43 in series.

[0035] The amplifier 41 amplifies signals input from the multiple transducers that make up the transducer array 13 and transmits the amplified signals to the A / D converter 42. The A / D converter 42 converts the signals transmitted from the amplifier 41 into digital reception data and transmits the converted reception data to the beamformer 43. The beamformer 43 adds each reception data converted by the A / D converter 42 with a delay in accordance with the speed of sound or the distribution of sound speeds that is set based on the reception delay pattern selected in response to a control signal transmitted from the processor 25 of the device main body 20. This addition process is called reception focusing process. By this reception focusing process, each reception data converted by the A / D converter 42 is phased and added, and a sound ray signal with a narrowed focus of the ultrasonic echo is acquired.

[0036] The device main body 20 has a display device 21, an input device 22, a communication unit 23, a storage device 24, and a processor 25. The input device 22 is, for example, a touch panel incorporated in the display device 21. When the device main body 20 is a PC or the like, the input device 22 may be a keyboard, a mouse, a trackball, a touchpad, or the like. The communication unit 23 performs wireless communication with the communication unit 15 of the ultrasound probe 10.

[0037] The processor 25 is connected to an input device 22 and a storage device 24. The processor 25 and the storage device 24 are connected so as to enable bidirectional information exchange between them.

[0038] The storage device 24 is a device that stores the program 26 that operates the ultrasound diagnostic device 2, and is, for example, a flash memory, an HDD (Hard Disc Drive), or an SSD (Solid State Drive). When the device main body 20 is a PC or the like, the storage device 24 can be a recording medium such as an FD (Flexible Disc), an MO (Magneto-Optical) disk, a magnetic tape, a CD (Compact Disc), a DVD (Digital Versatile Disc), an SD (Secure Digital) card, or a USB (Universal Serial Bus) memory, or a server.

[0039] The processor 25 is, for example, a CPU (Central Processing Unit). The processor 25 performs processing based on a program 26 in cooperation with a RAM (Random Access Memory) (not shown) and the like, thereby functioning as a main control unit 50, an image generation unit 51, a display control unit 52, an image analysis unit 53, and a highlighting unit 54.

[0040] The processor 25 is not limited to a CPU, but may be configured using an FPGA (Field Programmable Gate Array), a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), a GPU (Graphics Processing Unit), or other ICs (Integrated Circuits), or may be configured using a combination of these.

[0041] The main controller 50 controls each part of the ultrasound diagnostic apparatus 2 based on input operations by the operator via the input device 22. The main controller 50 transmits the above-mentioned control signal to the ultrasound probe 10 via the communication unit 23. The processor 25 receives, via the communication unit 23, an acoustic ray signal generated by the receiving circuit 17 from the ultrasound probe 10.

[0042] The image generation unit 51, under the control of the main control unit 50, acquires sound ray signals input from the ultrasonic probe 10 and generates an ultrasonic image U based on the acquired sound ray signals. As an example, as shown in Fig. 5, the image generation unit 51 is configured by connecting a signal processing unit 61, a DSC (Digital Scan Converter) 62, and an image processing unit 63 in series.

[0043] The signal processing unit 61 performs correction for attenuation due to distance on the sound ray signals generated by the receiving circuit 17 in accordance with the depth of the ultrasonic wave reflection position, and then performs envelope detection processing to generate a B-mode image signal, which is tomographic image information regarding the tissue within the subject.

[0044] The DSC 62 converts the B-mode image signal generated by the signal processing unit 61 into an image signal that conforms to the scanning method of a normal television signal (so-called raster conversion). The image processing unit 63 performs various image processing such as gradation processing on the B-mode image signal input from the DSC 62, and then outputs the B-mode image signal to the display control unit 52 and the image analysis unit 53. Hereinafter, the B-mode image signal that has been subjected to image processing by the image processing unit 63 will be simply referred to as an ultrasound image U.

[0045] The transmission / reception circuit 14 and the image generation unit 51 of the ultrasonic probe 10 are controlled by the main control unit 50 so that the ultrasonic images U are periodically generated at a constant frame rate. The transmission / reception circuit 14 and the image generation unit 51 function as an image acquisition unit that acquires the ultrasonic images U.

[0046] Under the control of the main controller 50, the display controller 52 performs predetermined processing on the ultrasound image U generated by the image generator 51, and causes the display device 21 to display the processed ultrasound image U.

[0047] Under the control of the main controller 50, the image analyzer 53 generates a blood vessel detection information DB by performing image analysis on the ultrasound image U input from the image generator 51, and outputs the generated blood vessel detection information DB to the highlighting display unit 54. The blood vessel detection information DB includes, for example, the detection results of the blood vessel region included in the ultrasound image U and the arterial / venous determination results of the detected blood vessel. The blood vessel region also includes a "single blood vessel region" representing the region of a single blood vessel, and a "blood vessel aggregate region" representing the region of a blood vessel aggregate consisting of three or more blood vessels. A blood vessel aggregate is, for example, a complex of an artery and an accompanying vein. Blood vessel aggregates are mainly present in anatomically stable areas.

[0048] Under the control of the main controller 50, the highlighting unit 54 controls the display controller 52 based on the blood vessel detection information DB input from the image analyzer 53, thereby highlighting the single blood vessel region and the blood vessel aggregate region in the ultrasound image U displayed on the display device 21. Furthermore, based on the arteriovenous determination result, the highlighting unit 54 displays the single blood vessel region in a manner that makes it possible to distinguish whether the blood vessel contained in the single blood vessel region is an artery or a vein.

[0049] 6, the image analysis unit 53 is composed of a single blood vessel detection unit 71, a blood vessel aggregate detection unit 72, an artery / vein determination unit 73, and a correction unit 74. The single blood vessel detection unit 71 and the blood vessel aggregate detection unit 72 receive an ultrasound image U generated by the image generation unit 51 as input.

[0050] The single blood vessel detection unit 71 identifies single blood vessel regions by individually detecting each blood vessel included in the ultrasound image U, and performs arterial / venous determination of the blood vessels included in the single blood vessel regions. The single blood vessel detection unit 71 outputs information including the detection results of the single blood vessel regions and the arterial / venous determination results for each single blood vessel region to the correction unit 74 as single blood vessel detection information D1.

[0051] The vascular aggregate detection unit 72 detects a vascular aggregate region where three or more blood vessels are aggregated based on the ultrasound image U, and outputs information representing the detected vascular aggregate region to the artery / vein determination unit 73 as vascular aggregate detection information D2.

[0052] The artery / vein determining unit 73 performs artery / vein determination for each blood vessel in the blood vessel aggregate region included in the blood vessel aggregate detection information D2, and outputs information indicating the result of the artery / vein determination to the correcting unit 74 as artery / vein determination information D3.

[0053] The correction unit 74 corrects the artery / vein determination results included in the individual blood vessel detection information D1 based on the artery / vein determination information D3. The correction unit 74 outputs information including the corrected individual blood vessel detection information D1, blood vessel aggregate detection information D2, and artery / vein determination information D3 to the highlighting unit 54 as the above-mentioned blood vessel detection information DB.

[0054] FIG. 7 shows an example of single blood vessel detection processing by the single blood vessel detection unit 71. The single blood vessel detection unit 71 uses a known algorithm to detect a single blood vessel region Rs containing a single blood vessel from within the ultrasound image U, and to determine whether it is an artery or a vein. In FIG. 7, symbols B1 to B5 represent blood vessels. Hereinafter, when it is not necessary to distinguish between blood vessels, they will be simply referred to as blood vessel B. The single blood vessel region Rs indicated by the dashed line represents a region containing blood vessel B determined to be an artery. The single blood vessel region Rs indicated by the solid line represents a region containing blood vessel B determined to be a vein.

[0055] Each single blood vessel region Rs is assigned a "label" that indicates the artery / vein determination result and a "score" that indicates the certainty (i.e., likelihood) of the artery / vein determination result. The label indicates whether the blood vessel B contained in the single blood vessel region Rs is an "artery" or a "vein." The score is a value within the range of 0 to 1, with the closer to 1 the higher the certainty. The single blood vessel region Rs to which the label and score have been assigned corresponds to the single blood vessel detection information D1 described above.

[0056] In this embodiment, the blood vessel unit detection unit 71 performs blood vessel unit detection processing using a blood vessel unit detection model 71A (see FIG. 8), which is a trained model generated by machine learning. The blood vessel unit detection model 71A is, for example, an object detection algorithm using deep learning. As the blood vessel unit detection model 71A, for example, an object detection model configured by R-CNN (Regional CNN), which is a type of convolutional neural network (CNN), can be used.

[0057] The single blood vessel detection model 71A detects an area containing a single blood vessel as an object from within the ultrasound image U, and determines a label for the detected area. The single blood vessel detection model 71A then outputs information representing the detected single blood vessel area Rs together with the label and score.

[0058] FIG. 8 is a diagram illustrating an example of a learning phase in which the individual blood vessel detection model 71A is trained by machine learning. The individual blood vessel detection model 71A learns using training data TD1. The training data TD1 includes a plurality of training images P to which a correct answer label L is attached. The training images P included in the training data TD1 are sample images of individual blood vessels (arteries and veins). The training data TD1 includes various training images P with different blood vessel shapes, sizes, etc.

[0059] In the learning phase, a teacher image P is input to the blood vessel unit detection model 71A. The blood vessel unit detection model 71A outputs a judgment result A for the teacher image P. A loss calculation is performed using a loss function based on this judgment result A and the correct label L. Then, various coefficients of the blood vessel unit detection model 71A are updated according to the result of the loss calculation, and the blood vessel unit detection model 71A is updated according to the update setting.

[0060] In the learning phase, a series of processes is repeatedly performed, including input of the teacher image P to the single blood vessel detection model 71A, output of the judgment result A from the single blood vessel detection model 71A, loss calculation, update setting, and update of the single blood vessel detection model 71A. This series of processes is terminated when the detection accuracy reaches a predetermined set level. The single blood vessel detection model 71A whose detection accuracy has thus reached the set level is stored in the storage device 24, and then used by the single blood vessel detection unit 71 in the single blood vessel detection process, which is the operation phase.

[0061] 9 shows an example of the vascular aggregate detection process by the vascular aggregate detection unit 72. The vascular aggregate detection unit 72 performs a process of detecting a vascular aggregate region Ra including a vascular aggregate from within the ultrasound image U using a known algorithm.

[0062] In this embodiment, the blood vessel aggregate detection unit 72 performs blood vessel aggregate detection processing using a blood vessel aggregate detection model 72A (see FIG. 10), which is a trained model generated by machine learning. The blood vessel aggregate detection model 72A is, for example, an object detection algorithm using deep learning. For example, an object detection model configured by R-CNN, which is a type of CNN, can be used as the blood vessel aggregate detection model 72A.

[0063] The blood vessel aggregate detection unit 72 detects a blood vessel aggregate region Ra containing a blood vessel aggregate as an object from within the ultrasound image U. Information representing the blood vessel aggregate region Ra corresponds to the above-mentioned blood vessel aggregate detection information D2.

[0064] FIG. 10 is a diagram illustrating an example of a learning phase in which the vascular aggregate detection model 72A is trained by machine learning. The vascular aggregate detection model 72A learns using training data TD2. The training data TD2 includes multiple training images P to which a correct answer label L is attached. The training images P included in the training data TD2 are sample images of vascular aggregates. The training data TD2 includes various training images P that differ in the number of blood vessels constituting a vascular aggregate, the shape and size of the blood vessels, the arrangement (positional relationship) of the multiple blood vessels, etc.

[0065] In the learning phase, a teacher image P is input to the blood vessel aggregate detection model 72A. The blood vessel aggregate detection model 72A outputs a judgment result A for the teacher image P. A loss calculation is performed using a loss function based on this judgment result A and the correct label L. Then, various coefficients of the blood vessel aggregate detection model 72A are updated according to the result of the loss calculation, and the blood vessel aggregate detection model 72A is updated according to the update setting.

[0066] In the learning phase, a series of processes is repeatedly performed: input of the teacher image P to the blood vessel aggregate detection model 72A, output of the determination result A from the blood vessel aggregate detection model 72A, loss calculation, update setting, and update of the blood vessel aggregate detection model 72A. This series of processes is terminated when the detection accuracy reaches a predetermined set level. The blood vessel aggregate detection model 72A whose detection accuracy has thus reached the set level is stored in the storage device 24 and then used by the blood vessel aggregate detection unit 72 in the blood vessel aggregate detection process, which is the operation phase.

[0067] 11 shows an example of artery / vein determination processing by the artery / vein determination unit 73. Based on the blood vessel aggregate detection information D2, the artery / vein determination unit 73 performs artery / vein determination for each blood vessel B included in the blood vessel aggregate region Ra, and generates artery / vein determination information D3 by obtaining a label and a score for each blood vessel B. In the artery / vein determination, the artery / vein determination unit 73 uses the feature amounts of anatomical blood vessels in the blood vessel aggregate region Ra. The artery / vein determination unit 73 obtains a score for each of "artery" and "vein" as labels for the blood vessel B, and selects the label with the larger score.

[0068] FIG. 12 is a diagram illustrating an example of the artery / vein determination process in more detail. First, the artery / vein determination unit 73 calculates a feature amount for each blood vessel B included in the blood vessel assembly region Ra. The feature amount includes the diameter D of each blood vessel B in the blood vessel assembly region Ra (hereinafter referred to as the blood vessel diameter), the displacement K of the blood vessel B from the center C of the blood vessel assembly region Ra, and the circularity of each blood vessel B. The displacement K is, for example, the distance from the center C of the blood vessel assembly region Ra to the center of each blood vessel B. The artery / vein determination unit 73 determines whether each blood vessel B is an artery or a vein using at least one or more feature amounts from the blood vessel diameter D, the displacement K, and the circularity. The artery / vein determination unit 73 performs the artery / vein determination using a machine learning algorithm such as AdaBoost or SVM (Support Vector Machine).

[0069] Arteries are often located at the center of a blood vessel assembly, so the displacement K is small. Arteries have a large blood vessel diameter and a high degree of circularity due to their high internal pressure. Conversely, veins accompany arteries, so the displacement K is large. Veins have a small blood vessel diameter and a low degree of circularity due to their low internal pressure. The artery / vein determination unit 73 performs artery / vein determination for each blood vessel B based on these feature amounts, calculates a score, and determines a label. For example, the artery / vein determination unit 73 calculates an artery score and a vein score for each blood vessel B, and selects the label with the larger score. In this way, by performing determination using anatomical feature amounts of a blood vessel assembly, artery / vein determination can be performed with high accuracy.

[0070] FIG. 13 shows an example of the correction process performed by the correction unit 74. For the corresponding blood vessel B, the correction unit 74 compares the score included in the artery / vein determination information D3 with the score included in the blood vessel individual detection information D1 and selects the label with the higher score. For example, in the example shown in FIG. 13, for blood vessel B3, the score (0.90) included in the artery / vein determination information D3 is higher than the score (0.75) included in the blood vessel individual detection information D1. Therefore, the correction unit 74 selects the label (artery) included in the artery / vein determination information D3 as the label for blood vessel B3, instead of the label (vein) included in the blood vessel individual detection information D1. Similarly, for blood vessels B4 and B5, the correction unit 74 selects the label with the higher score. That is, in the example shown in FIG. 13, of the labels for blood vessels B1 to B5 included in the blood vessel individual detection information D1, only the label for blood vessel B3 is corrected. In this manner, the labels included in the blood vessel individual detection information D1 are corrected.

[0071] The correction unit 74 outputs information including the single blood vessel detection information D1 after correcting the labels, the blood vessel aggregate detection information D2, and the artery / vein determination information D3 as the above-mentioned blood vessel detection information DB to the highlighting unit 54. The blood vessel detection information DB includes position information of the single blood vessel region Rs and the blood vessel aggregate region Ra in the ultrasound image U, and the label and score for the single blood vessel region Rs.

[0072] FIG. 14 shows an example of highlighting processing by the highlighting unit 54. The highlighting unit 54 displays a single blood vessel region Rs and a blood vessel aggregate region Ra using rectangular frames based on the blood vessel detection information DB within the ultrasound image U displayed on the display device 21 of the device main body 20. Furthermore, the highlighting unit 54 displays the single blood vessel region Rs based on the artery / vein determination result so that it is possible to distinguish whether the blood vessels contained within the single blood vessel region Rs are arteries or veins. In the example shown in FIG. 14, a single blood vessel region Rs containing a vein is indicated by a solid line, and a single blood vessel region Rs containing an artery is indicated by a dashed line. Furthermore, the blood vessel aggregate region Ra is indicated by a two-dot chain line. The highlighting unit 54 may display the single blood vessel region Rs and the blood vessel aggregate region Ra in a distinguishable manner not limited to line type, but also by line thickness, line color, line brightness, or the like.

[0073] Next, an example of the operation of the ultrasound diagnostic apparatus 2 will be described using the flowchart shown in Fig. 15. First, the main controller 50 determines whether or not a start operation has been performed by the surgeon using the input device 22 or the like (step S10). If the main controller 50 determines that a start operation has been performed (step S10: YES), it operates the transmission / reception circuit 14 and the image generator 51 of the ultrasound probe 10 to generate an ultrasound image U (step S11). The generated ultrasound image U is displayed on the display device 21 by the display controller 52.

[0074] At this time, the surgeon brings the ultrasonic probe 10 into contact with the surface of the living body 30, as shown in Fig. 2. An ultrasonic beam UB is transmitted from the transducer array 13 into the living body 30 in accordance with a drive signal input from the transmission circuit 16. Ultrasonic echoes from within the living body 30 are received by the transducer array 13, and a received signal is output to the reception circuit 17. The received signal received by the reception circuit 17 becomes a sound ray signal via an amplifier 41, an A / D converter 42, and a beam former 43. This sound ray signal is output to the device main body 20 via the communication unit 15.

[0075] The device main body 20 receives the sound ray signals output from the ultrasonic probe 10 via the communication unit 23. The sound ray signals received by the device main body 20 are input to the image generation unit 51. In the image generation unit 51, the sound ray signals are subjected to envelope detection processing in the signal processing unit 61 to become B-mode image signals, which are then output to the display control unit 52 as an ultrasound image U after passing through the DSC 62 and image processing unit 63. The ultrasound image U is also output to the image analysis unit 53.

[0076] In the image analysis unit 53, the blood vessel unit detection unit 71 performs the above-mentioned blood vessel unit detection process (see FIG. 7) (step S12). The blood vessel unit detection information D1 generated by this blood vessel unit detection process is output to the correction unit 74.

[0077] Furthermore, steps S13 and S14 are performed in parallel with step S12. In step S13, the blood vessel aggregate detection unit 72 performs the above-mentioned blood vessel aggregate detection process (see FIG. 9). The blood vessel aggregate detection information D2 generated by this blood vessel aggregate detection process is output to the artery / vein determination unit 73. In step S14, the artery / vein determination unit 73 performs the above-mentioned artery / vein determination process (see FIGS. 11 and 12). The artery / vein determination information D3 generated by this artery / vein determination process is output to the correction unit 74.

[0078] Furthermore, in the image analysis unit 53, the correction unit 74 performs the above-mentioned correction process (see FIG. 13) (step S15). In this correction process, the label for the blood vessel unit region Rs included in the blood vessel unit detection information D1 is corrected based on the artery / vein determination information D3. As a result of this correction process, the blood vessel detection information DB is output to the highlighting unit 54.

[0079] Then, the highlighting unit 54 performs the above-mentioned highlighting process (see FIG. 14) (step S16). This highlighting process highlights the single blood vessel region Rs and the blood vessel aggregate region Ra in the ultrasound image U displayed on the display device 21. The single blood vessel region Rs is also displayed so that it is possible to distinguish whether the blood vessel contained therein is an artery or a vein. By performing highlighting in this way, the surgeon can accurately grasp the blood vessel aggregate present in the ultrasound image U, and can also accurately grasp whether the single blood vessel is an artery or a vein.

[0080] Next, the main controller 50 determines whether or not a termination operation has been performed by the surgeon using the input device 22 or the like (step S17). If the main controller 50 determines that a termination operation has not been performed (step S17: NO), the main controller 50 returns the process to step S11. As a result, a new ultrasound image U is generated. On the other hand, if the main controller 50 determines that a termination operation has been performed (step S17: YES), the main controller 50 terminates the operation of the ultrasound diagnostic apparatus 2.

[0081] Conventionally, blood vessels are detected by individual blood vessel detection processing, and then artery / vein determination is performed for each detected blood vessel. This method often results in errors in determining whether a blood vessel is an artery or vein in a blood vessel cluster, and the determination of artery / vein changes from frame to frame. When a surgeon attempts to perform puncture based on the results of this artery / vein determination, they may end up puncturing the wrong blood vessel.

[0082] In contrast, the technology disclosed herein detects a blood vessel assembly region Ra from an ultrasound image U and highlights the detected blood vessel assembly region Ra within the ultrasound image U, thereby assisting the surgeon in determining whether a blood vessel is an artery or a vein. Furthermore, the technology disclosed herein performs artery / vein determination based on the feature amounts of the blood vessels within the blood vessel assembly region Ra, enabling accurate artery / vein determination even for blood vessels that constitute a blood vessel assembly. This allows the surgeon to accurately identify the blood vessel (e.g., vein) to be punctured.

[0083] [Variations] Various modifications of the ultrasonic diagnostic apparatus 2 according to the first embodiment will be described below.

[0084] In the first embodiment, in the artery / vein determination process (see FIGS. 11 and 12), the artery / vein determination unit 73 calculates a score for each of "artery" and "vein" as labels for blood vessel B, and selects the label with the larger score. Alternatively, as shown in FIG. 16, for example, a threshold value for the score may be set, and a label with a score equal to or greater than the threshold may be selected.

[0085] Also, as shown in FIG. 17, for example, scores may be calculated for each of "artery" and "vein" as labels for blood vessel B, and both scores may be below the threshold. In this case, the artery / vein determination unit 73 may determine that it is difficult to determine the label (i.e., artery / vein determination) and may stop the artery / vein determination. When it is difficult to determine the artery / vein in this way, the highlighting unit 54 may display the single blood vessel region Rs in the ultrasound image U without distinguishing between an "artery" and a "vein." In this case, the highlighting unit 54 may simply display the single blood vessel region Rs as a "blood vessel," as shown in FIG. 18, for example.

[0086] 19, the highlighting unit 54 may display the score for the label selected by the correction unit 74 (i.e., the confidence level for the determination result selected by the correction unit 74) in association with the single blood vessel region Rs. This allows the surgeon to understand the confidence level of the artery / vein determination for each blood vessel.

[0087] 20, the highlighting unit 54 may display a message urging the operator to be careful when the score for the label selected by the correction unit 74 (i.e., the confidence level for the determination result selected by the correction unit 74) is lower than a certain value. This allows the operator to reliably understand that the confidence level for the arteriovenous determination is low and that caution is required when performing puncture.

[0088] Furthermore, the arteriovenous determination unit 73 may change the criteria for arteriovenous determination for each vascular aggregate region Ra. This is because, for example, if the pattern of the vascular aggregate within the vascular aggregate region Ra is an anatomically typical pattern, the determination result is likely to be correct even if the score of the arteriovenous determination is low. The vascular aggregate pattern is information expressed by the relative positions of multiple blood vessels in the vascular aggregate, the number of blood vessels, the size of each blood vessel, etc.

[0089] Before performing artery / vein determination on the vascular aggregate region Ra, the artery / vein determination unit 73 performs, for example, a threshold setting process shown in FIG. 21. First, the artery / vein determination unit 73 analyzes the pattern of the vascular aggregate in the vascular aggregate region Ra (step S20) and determines whether the vascular aggregate has a typical pattern (step S21). If the artery / vein determination unit 73 determines that the vascular aggregate has a typical pattern (step S21: YES), it sets the threshold for artery / vein determination to a "first threshold" (step S22). On the other hand, if the artery / vein determination unit 73 determines that the vascular aggregate does not have a typical pattern (step S21: NO), it sets the threshold for artery / vein determination to a "second threshold" (step S23). Here, the first threshold is a value smaller than the second threshold.

[0090] As shown in Figure 22, when the pattern of the vascular aggregate in the vascular aggregate region Ra is typical, the threshold value serving as the criterion for arteriovenous determination is set low. On the other hand, when the pattern of the vascular aggregate is atypical, the threshold value serving as the criterion for arteriovenous determination is set high. In this way, when the pattern of the vascular aggregate is atypical, the criterion for arteriovenous determination is set high, thereby making a more reliable determination. Note that the criterion for arteriovenous determination may be changed by changing the algorithm for arteriovenous determination, not limited to the threshold value for the score.

[0091] Furthermore, as shown in FIG. 23, the highlighting unit 54 may hide the blood vessel aggregate region Ra after the artery / vein determining unit 73 has determined whether the blood vessel is an artery or a vein.

[0092] In the first embodiment, the individual blood vessel detection unit 71 and the blood vessel aggregate detection unit 72 are configured using separate object detection models, but the individual blood vessel detection unit 71 and the blood vessel aggregate detection unit 72 can also be configured using a single object detection model. In this case, the object detection model can be trained using training data including training images of individual blood vessels and training images of blood vessel aggregates. The individual blood vessel detection unit 71, the blood vessel aggregate detection unit 72, and the artery / vein determination unit 73 can also be configured using a single object detection model. Furthermore, the individual blood vessel detection unit 71, the blood vessel aggregate detection unit 72, the artery / vein determination unit 73, and the correction unit 74 can also be configured using a single object detection model.

[0093] In addition, in the first embodiment, the individual blood vessel detection unit 71 and the blood vessel aggregate detection unit 72 are configured using an object detection model consisting of CNN, but the object detection model is not limited to CNN and may be a segmentation or other general detection model.

[0094] Furthermore, the object detection model constituting the individual blood vessel detection unit 71 and the blood vessel aggregate detection unit 72 may be configured with a classifier that classifies objects based on image features, such as AdaBoost or SVM. In this case, the teacher image is converted into a feature vector, and then the classifier is trained based on the feature vector. The distance between blood vessels (such as the distance between the centers of blood vessels or the distance between the outer peripheries of blood vessels) can be used as a feature of a blood vessel aggregate. The classifier recognizes a blood vessel aggregate when the distance between blood vessels is equal to or less than a certain value and the number of blood vessels is three or more.

[0095] Furthermore, the individual blood vessel detection unit 71 and the blood vessel aggregate detection unit 72 are not limited to object detection models based on machine learning, and may perform object detection using template matching. In this case, the individual blood vessel detection unit 71 pre-stores typical pattern data of individual blood vessels as templates, and calculates the similarity to the pattern data while searching the ultrasound image U with the templates. The individual blood vessel detection unit 71 then identifies, as an individual blood vessel region Rs, a location where the similarity is greater than or equal to a certain level and is at its maximum. The blood vessel aggregate detection unit 72 pre-stores typical pattern data of blood vessel aggregates as templates, and calculates the similarity to the pattern data while searching the ultrasound image U with the templates. The blood vessel aggregate detection unit 72 then identifies, as a blood vessel aggregate region Ra, a location where the similarity is greater than or equal to a certain level and is at its maximum. The template may be a portion of an actual ultrasound image, or may be a schematic image of a blood vessel or a blood vessel aggregate.

[0096] In addition to simple template matching, similarity calculations can also use, for example, the 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 the general image recognition method using deep learning described in Krizhevsk et al.: ImageNet Classification with Deep Convolutional Neural Networks, Advances in Neural Information Processing Systems 25, pp. 1106-1114 (2012).

[0097] In the first embodiment, the ultrasonic probe 10 and the device main body 20 are connected by wireless communication, but instead, the ultrasonic probe 10 and the device main body 20 may be connected by wire.

[0098] Furthermore, in the first embodiment, the image generation unit 51 that generates the ultrasound image U based on the sound ray signal is provided in the device main body 20, but instead, the image generation unit 51 may be provided in the ultrasound probe 10. In this case, the ultrasound probe 10 generates the ultrasound image U and outputs it to the device main body 20. The processor 25 of the device main body 20 performs image analysis and the like based on the ultrasound image U input from the ultrasound probe 10.

[0099] In addition, in the first embodiment, the display device 21, the input device 22, and the ultrasonic probe 10 are directly connected to the processor 25, but the display device 21, the input device 22, the ultrasonic probe 10, and the processor 25 may be indirectly connected via a network.

[0100] 24 shows an example of an ultrasound diagnostic device 2A in which a display device 21, an input device 22, and an ultrasound probe 10A are connected to an ultrasound diagnostic device main body 20A via a network NW. The ultrasound diagnostic device 2A is the ultrasound diagnostic device 20 according to the first embodiment except for the display device 21 and the input device 22, and is configured with the transmission / reception circuit 14, a storage device 24, and a processor 25. The ultrasound probe 10A is the ultrasound probe 10 according to the first embodiment except for the transmission / reception circuit 14.

[0101] In this way, in the ultrasound diagnostic device 2A, the display device 21, the input device 22, and the ultrasound probe 10A are connected to the device main body 20A via the network NW, so that the device main body 20A can be used as a so-called remote server. This allows, for example, the surgeon to have the display device 21, the input device 22, and the ultrasound probe 10A at hand, improving convenience. Furthermore, by configuring the display device 21 and the input device 22 as a mobile terminal such as a smartphone or a tablet terminal, convenience is further improved.

[0102] As another example, in an ultrasound diagnostic device 2B shown in Fig. 25, a display device 21 and an input device 22 are mounted on a device main body 20B, and an ultrasound probe 10A is connected to the device main body 20B via a network NW. In this case, the device main body 20B may be configured as a remote server. Alternatively, the device main body 20B may be configured as a mobile terminal such as a smartphone or a tablet terminal.

[0103] In the first embodiment, the following various processors can be used as the hardware structure of processing units that perform various processes, such as the main control unit 50, image generation unit 51, display control unit 52, image analysis unit 53, and highlighting unit 54. As described above, the various processors include a CPU, which is a general-purpose processor that executes software (program 26) and functions as various processing units, as well as a programmable logic device (PLD), which is a processor whose circuit configuration can be changed after manufacture, such as an FPGA, and a dedicated electrical circuit, such as an ASIC, which is a processor with a circuit configuration designed specifically for performing specific processes.

[0104] A single processing unit may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs and / or a combination of a CPU and an FPGA). Also, multiple processing units may be configured with a single processor.

[0105] Examples of configuring multiple processing units with a single processor include, first, a form in which one processor is configured with a combination of one or more CPUs and software, and this processor functions as multiple processing units, as typified by client and server computers. Second, a form in which a processor is used to realize the functions of an entire system including multiple processing units on a single IC chip, as typified by a system on chip (SoC). In this way, various processing units are configured using one or more of the above-mentioned various processors as a hardware structure.

[0106] Furthermore, more specifically, the hardware structure of these various processors can be an electric circuit that combines circuit elements such as semiconductor elements.

[0107] From the above description, the techniques described in Supplementary Items 1 to 9 below can be understood.

[0108] [Additional note 1] An information processing device that performs processing to transmit an ultrasonic beam from a transducer array toward a living body, and to display an ultrasonic image generated by receiving an ultrasonic echo generated in the living body on a display device, a processor; The processor: A blood vessel aggregate region including a blood vessel aggregate in which three or more blood vessels are aggregated is detected from the ultrasound image, and the detected blood vessel aggregate region is highlighted in the ultrasound image. Information processing device. [Additional note 2] The processor: Determining whether each blood vessel included in the blood vessel assembly region is an artery or a vein based on the feature amount of each blood vessel within the blood vessel assembly region. Item 1. An information processing device according to item 1. [Additional note 3] The processor: The determination is made based on at least one or more feature amounts selected from the blood vessel diameter, the displacement amount of the blood vessel from the center of the blood vessel assembly region, and the circularity of the blood vessel. Item 2. An information processing device according to claim 2. [Additional note 4] The processor: A single blood vessel region containing a single blood vessel is detected from the ultrasound image, and it is determined whether the blood vessel contained in the detected single blood vessel region is an artery or a vein. Item 2 or 3. An information processing device according to item 2 or 3. [Additional note 5] The processor: The result of the artery / vein determination for the blood vessel in the single blood vessel region is corrected based on the result of the artery / vein determination for each blood vessel in the blood vessel assembly region. Item 4. An information processing device according to claim 4. [Additional note 6] The processor: The certainty of the artery / vein determination for the blood vessel in the single blood vessel region is compared with the certainty of the artery / vein determination for each blood vessel in the blood vessel assembly region, and the determination result with the higher certainty is selected. Item 5. An information processing device according to item 5. [Additional note 7] The processor: Based on the correction result, each of the blood vessels included in the blood vessel aggregate region is displayed so that arteries and veins can be distinguished. Item 6. An information processing device according to claim 6. [Additional note 8] The processor: The confidence level for the selected determination result is displayed on the display device. Item 7. An information processing device according to item 7. [Additional note 9] The processor: If the confidence level for the selected determination result is lower than a certain value, a message calling for attention is displayed on the display device. An information processing device according to claim 8.

[0109] The technology of the present disclosure can be appropriately combined with the various embodiments and / or various modified examples described above. Furthermore, it is needless to say that it is not limited to the above embodiments and various configurations can be adopted as long as they do not deviate from the gist of the present disclosure.

[0110] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[0111] In this specification, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B.

[0112] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference. [Explanation of symbols]

[0113] 2, 2A, 2B Ultrasound diagnostic equipment 4. Receiving circuit 10,10A Ultrasonic Probe 11. Housing 11A Array housing 11B Grip section 13 Transducer array 14 Transmitting and receiving circuit 15 Communications Department 16 Transmitting circuit 17 Receiving circuit 20, 20A, 20B Device body 21 Display device 22 Input Devices 23 Communications Department 24 Storage device 25 processors 26 Programs 30 Living organisms 31 Puncture needle 41 Amplification section 42 A / D conversion section 43 Beamformer 50 Main control unit 51 Image generation unit 52 Display control unit 53 Image analysis unit 54 Highlight section 61 Signal processing section 62 DSC 63 Image processing section 71 Single blood vessel detection unit 71A Single blood vessel detection model 72 Blood vessel aggregate detection unit 72A Vascular aggregate detection model 73 Arteriovenous determination section 74 Correction section A Judgment result B,B1~B5 Blood vessels C center D Vessel diameter K displacement D1 Single blood vessel detection information D2 Vascular aggregate detection information D3 Arteriovenous determination information DB Blood vessel detection information L Correct label M Guide Marker NW Network P Teacher image Ra blood vessel aggregate area Rs Single blood vessel region TD1,TD2 training data U Ultrasound image UB ultrasonic beam

Claims

1. An information processing device that performs processing to transmit an ultrasonic beam from a transducer array toward a living body, and to display an ultrasonic image generated by receiving an ultrasonic echo generated in the living body on a display device, a vascular aggregate detection unit that detects a vascular aggregate region including a vascular aggregate formed by an aggregation of three or more blood vessels from the ultrasound image; a highlighting unit that highlights the blood vessel aggregate region in the ultrasound image; an artery / vein determination unit that determines whether each blood vessel included in the blood vessel assembly region is an artery or a vein based on a feature amount of each blood vessel in the blood vessel assembly region; Equipped with The feature amount includes at least one of a blood vessel diameter, a displacement amount of the blood vessel from the center of the blood vessel assembly region, and a circularity of the blood vessel. Information processing device.

2. The displacement is the distance from the center of the blood vessel assembly region to the center of the blood vessel. The information processing device according to claim 1 .

3. a blood vessel unit detection unit that detects a blood vessel unit region including a blood vessel unit from the ultrasound image and determines whether the blood vessel included in the detected blood vessel unit region is an artery or a vein; The information processing device according to claim 1 or 2, further comprising:

4. a correction unit that corrects the result of the artery / vein determination made by the blood vessel individual detection unit based on the result of the artery / vein determination made by the artery / vein determination unit; The information processing device according to claim 3 , further comprising:

5. The correction unit compares the certainty of the artery / vein determination made by the blood vessel individual detection unit with the certainty of the artery / vein determination made by the artery / vein determination unit, and selects the determination result with the higher certainty. The information processing device according to claim 4 .

6. The highlighting unit displays each of the blood vessels included in the blood vessel aggregate region in a manner that allows arteries and veins to be distinguished based on the correction result by the correction unit. The information processing device according to claim 5 .

7. The highlighting unit displays, on the display device, a degree of certainty for the determination result selected by the correction unit. The information processing device according to claim 6 .

8. The highlighting unit displays a message on the display device to call attention when the confidence level of the determination result selected by the correction unit is lower than a certain value. The information processing device according to claim 7 .

9. An information processing method in which a computer performs a process of transmitting an ultrasound beam from a transducer array toward a living body, and displaying an ultrasound image generated by receiving ultrasound echoes generated within the living body on a display device, the method comprising: The computer a process of detecting a vascular aggregate region including a vascular aggregate formed by an aggregation of three or more blood vessels from the ultrasound image, and highlighting the detected vascular aggregate region in the ultrasound image; a process of determining whether each blood vessel included in the blood vessel assembly region is an artery or a vein based on a feature amount of each blood vessel in the blood vessel assembly region; and The feature amount includes at least one of a blood vessel diameter, a displacement amount of the blood vessel from the center of the blood vessel assembly region, and a circularity of the blood vessel. Information processing methods.

10. A program that causes a computer to execute a process of transmitting an ultrasonic beam from a transducer array toward a living body and displaying an ultrasonic image generated by receiving an ultrasonic echo generated in the living body on a display device, a process of detecting a vascular aggregate region including a vascular aggregate formed by an aggregation of three or more blood vessels from the ultrasound image, and highlighting the detected vascular aggregate region in the ultrasound image; a process of determining whether each blood vessel included in the blood vessel assembly region is an artery or a vein based on a feature amount of each blood vessel in the blood vessel assembly region; on the computer, The feature amount includes at least one of a blood vessel diameter, a displacement amount of the blood vessel from the center of the blood vessel assembly region, and a circularity of the blood vessel. program.

Images