Information processing apparatus, information processing method, and storage medium

Abstract

An information processing apparatus, an information processing method, and a storage medium capable of improving the accuracy of determining arteries and veins of a blood vessel are provided. An information processing apparatus that performs processing on an ultrasonic image (U) generated by transmitting an ultrasonic beam (UB) toward a living body (30) from a transducer array (13) and receiving ultrasonic echoes generated in the living body (30) has a blood vessel detection section (71) that detects a blood vessel region (Ra) containing a blood vessel (B) from the ultrasonic image (U), a structure detection section (72) that detects a structure other than the blood vessel (B) from the ultrasonic image (U), and an artery and vein determination section (73) that determines which of an artery and a vein the blood vessel (B) contained in the blood vessel region (Ra) is based on the relative positional relationship between the blood vessel region (Ra) and the structure.

Description

Technical Field

[0001] The present invention relates to an information processing device, an information processing method, and a storage medium. Background Technology

[0002] Ultrasound diagnostic devices have long been known as instruments for obtaining images of the interior of a subject. These devices typically include an ultrasound probe with an array of multiple ultrasonic transducers arranged in a radii array. While in contact with the surface of the subject, the probe transmits an ultrasonic beam from the transducer array into the subject and receives the ultrasonic echoes from the subject via the transducer array. This allows the acquisition of an electrical signal corresponding to the ultrasonic echo. Furthermore, the ultrasound diagnostic device processes the acquired electrical signal to generate an ultrasound image of that specific area of ​​the subject.

[0003] A procedure is known in which a puncture needle is inserted into a blood vessel of the patient while an ultrasound diagnostic device is used to observe the patient's body (so-called echo-guided puncture). In echo-guided puncture, the surgeon typically needs to confirm the ultrasound image to determine the location and shape of the blood vessels contained within it; however, to accurately determine this, a certain level of proficiency is required. Therefore, a method is proposed that automatically detects blood vessels contained in the ultrasound image and alerts the surgeon to the detected blood vessels (for example, see Patent Document 1).

[0004] Previous technical documents

[0005] Patent documents

[0006] Patent Document 1: Japanese Patent Publication No. 2017-524455 Summary of the Invention

[0007] The technical problem to be solved by the invention

[0008] During a puncture, the surgeon needs to accurately determine whether the blood vessel is an artery or a vein based on the ultrasound image. This determination will be referred to below as arteriovenous identification.

[0009] Another approach is to determine the arteries and veins of blood vessels based on ultrasound images and information processing methods such as image analysis. However, when determining the arteries and veins separately, errors can easily occur when arteries and veins have similar shapes.

[0010] The technical objective of this invention is to provide an information processing device, information processing method, and program that can improve the accuracy of arteriovenous identification of blood vessels.

[0011] means for solving technical problems

[0012] The information processing apparatus of the present invention processes an ultrasound image generated by transmitting an ultrasound beam from a transducer array toward a living body and receiving ultrasound echoes generated within the living body. The apparatus includes: a blood vessel detection unit for detecting a blood vessel region containing blood vessels within the ultrasound image; a structure detection unit for detecting structures other than blood vessels within the ultrasound image; and an artery / vein determination unit for determining whether the blood vessel region contains an artery or a vein based on the relative positional relationship between the blood vessel region and the structure.

[0013] Preferably, it also includes an emphasis display unit that displays the vascular region in the ultrasound image displayed on the display device in a way that allows identification of whether the blood vessel contained in the vascular region is an artery or a vein.

[0014] In addition to detecting the vascular region, the vascular detection unit also determines whether the blood vessel contained in the vascular region is an artery or a vein.

[0015] Preferably, it also includes a correction unit that corrects the arterial and vein determination results determined by the blood vessel detection unit based on the arterial and vein determination results based on the arterial and vein determination results of the arterial and vein determination unit.

[0016] The correction unit preferably compares the reliability of the arteriovenous determination performed by the vascular detection unit with the reliability of the arteriovenous determination performed by the arteriovenous determination unit, and selects the determination result with higher reliability.

[0017] It is emphasized that the display unit preferably displays the confidence level of the determination result selected by the correction unit on the display device.

[0018] The display unit preferably displays a warning message on the display device when the confidence level of the determination result selected by the correction unit is lower than a certain value.

[0019] The information processing method of the present invention processes an ultrasound image generated by transmitting an ultrasound beam from a transducer array toward a living body and receiving the ultrasound echo generated in the living body. In this information processing method, a vascular region containing blood vessels is detected from the ultrasound image, and structures other than blood vessels are detected from the ultrasound image. Based on the relative positional relationship between the vascular region and the structure, it is determined whether the blood vessel contained in the vascular region is an artery or a vein.

[0020] The present invention uses a computer-readable storage medium storing a program that causes the computer to perform processing on an ultrasound image generated by transmitting an ultrasound beam from a transducer array toward a living body and receiving ultrasound echoes generated within the living body. The program causes the computer to perform the following processing: detecting vascular regions containing blood vessels from the ultrasound image; detecting structures other than blood vessels from the ultrasound image; and determining, based on the relative positional relationship between the vascular regions and the structures, whether the blood vessels contained in the vascular regions are arteries or veins.

[0021] Invention Effects

[0022] According to the technology of the present invention, an information processing device, information processing method and program can be provided that can improve the accuracy of arteriovenous identification of blood vessels. Attached Figure Description

[0023] Figure 1 This is an external view showing an example of the structure of the ultrasound diagnostic device according to the first embodiment.

[0024] Figure 2 This is a diagram illustrating an example of echo-guided puncture.

[0025] Figure 3 This is a block diagram illustrating an example of the structure of an ultrasound diagnostic device.

[0026] Figure 4 This is a block diagram illustrating an example of the structure of a receiving circuit.

[0027] Figure 5 This is a block diagram representing an example of the structure of the image generation unit.

[0028] Figure 6 This is a block diagram illustrating an example of the structure of the image analysis unit.

[0029] Figure 7 This is a diagram illustrating an example of vascular detection and processing.

[0030] Figure 8 This is a diagram illustrating one example of the learning phase in learning a blood vessel detection model.

[0031] Figure 9 This is a diagram illustrating an example of structural detection and processing.

[0032] Figure 10 This is a diagram illustrating an example of the learning phase of a learning structure detection model.

[0033] Figure 11 This is a diagram illustrating an example of arteriovenous venous diagnoses and treatment.

[0034] Figure 12 This is a diagram illustrating an example of the corrective processing.

[0035] Figure 13 This is a diagram illustrating an example of emphasizing the processing.

[0036] Figure 14 This is a flowchart illustrating an example of the operation of an ultrasound diagnostic device.

[0037] Figure 15 This is a diagram illustrating an example of blood vessel detection processing for other ultrasound images.

[0038] Figure 16 This is a diagram illustrating an example of structure detection processing for other ultrasound images.

[0039] Figure 17 This diagram illustrates an example of arteriovenous identification processing for other ultrasound images.

[0040] Figure 18 This is a diagram illustrating an example of correction processing for other ultrasound images.

[0041] Figure 19 This is a diagram illustrating an example of emphasis processing applied to other ultrasound images.

[0042] Figure 20 This is a diagram illustrating a variation of the arteriovenous defect identification process.

[0043] Figure 21 This is a diagram illustrating a variation of the arteriovenous defect identification process.

[0044] Figure 22 This is a diagram showing an example of displaying fractions.

[0045] Figure 23 This is an example of a diagram showing a reminder message to the surgeon.

[0046] Figure 24 This is a diagram showing the first modified example of an ultrasonic diagnostic device.

[0047] Figure 25 This is a diagram showing a second variation of an ultrasonic diagnostic device. Detailed Implementation

[0048] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. The description of the structural elements described below is based on representative embodiments of the present invention, but the technology of the present invention is not limited to such embodiments.

[0049] [First Implementation]

[0050] Figure 1This illustrates an example of the structure of the ultrasound diagnostic apparatus 2 according to the present invention. The ultrasound diagnostic apparatus 2 according to this embodiment comprises an ultrasound probe 10 and an apparatus body 20. The ultrasound probe 10 is held by a surgeon and contacts the surface of the living body to be measured. The ultrasound probe 10 transmits and receives ultrasound beams UB relative to the interior of the living body.

[0051] The main body 20 of the device is, for example, a smartphone or tablet. The main body 20 performs tasks such as image processing of the signals output from the ultrasonic probe 10 by installing application software or other programs. The ultrasonic probe 10 and the main body 20 communicate wirelessly with each other, for example, via WiFi or Bluetooth (registered trademark). The main body 20 is not limited to mobile terminals such as smartphones or tablets, and can be a PC (Personal Computer). The main body 20 is an example of an "information processing device" according to the technology of this invention.

[0052] The ultrasonic probe 10 has a housing 11. The housing 11 consists of an array housing 11A and a handle 11B. The array housing 11A houses the transducer array 13 (reference). Figure 3 The handle 11B is connected to the array housing 11A and is held by the surgeon. For the sake of explanation, the direction from the handle 11B toward the array housing 11A is defined as the +Y direction, the width direction of the ultrasonic probe 10 orthogonal to the Y direction is defined as the X direction, and the direction orthogonal to both the X and Y directions (i.e., the thickness direction of the ultrasonic probe 10) is defined as the Z direction.

[0053] An acoustic lens is disposed at the +Y direction end of the array housing 11A. An acoustic matching layer (not shown) is disposed on the transducer array 13, and an acoustic lens is disposed on the acoustic matching layer. The plurality of transducers included in the transducer array 13 are arranged in a straight line along the X direction. That is, the ultrasonic probe 10 of this embodiment is linear, transmitting the ultrasonic beam UB in a linear manner. Alternatively, the ultrasonic probe 10 may be convex, with the transducer array 13 configured as a convex curved surface. In this case, the ultrasonic probe 10 transmits the ultrasonic beam UB radially. Furthermore, the ultrasonic probe 10 may be fan-shaped.

[0054] Furthermore, a linear guide mark M extending in the Y direction is provided on the outer periphery of the array housing 11A. The guide mark M is used as a reference when the operator brings the ultrasonic probe 10 into contact with the living body.

[0055] The main body 20 of the device has a display device 21 for displaying an ultrasonic image based on a signal transmitted from the ultrasonic probe 10. The display device 21 is, for example, an organic EL (Organic Electro-Luminescence) display or a liquid crystal display. A touch panel is mounted on the display device 21. The surgeon can perform various operations on the main body 20 of the device via the touch panel.

[0056] Figure 2 This diagram illustrates an example of echo-guided puncture. (See diagram for example.) Figure 2 As shown, the ultrasound probe 10 is used by the surgeon to insert a puncture needle 31 into a blood vessel B within a living body 30 while simultaneously reviewing the ultrasound image displayed on the device body 20. The living body 30 is, for example, a human wrist. The ultrasound probe 10 is brought into contact with the surface of the living body 30, for example, by cutting across the direction of travel of the blood vessel B in the width direction (i.e., the X direction) of the ultrasound probe 10. This procedure is called the short-axis approach (or cross-sectional approach). The ultrasound image shows a cross-section of the blood vessel B. The surgeon performs punctures in one or more blood vessels B displayed in the ultrasound image, for example, veins.

[0057] In addition to blood vessels, anatomical structures (hereinafter referred to as structures) also exist within the living body 30. These structures include, for example, living tissues such as tendons, bones, nerves, or muscles. The types and characteristics of the structures contained within the living body 30 vary depending on the location of the living body 30 (wrist, leg, abdomen, etc.).

[0058] After detecting blood vessels from ultrasound images, the main body 20 of the device performs arterial and venous identification of the blood vessels and displays the results of the arterial and venous identification in the ultrasound image displayed on the display device 21, thereby assisting the puncture performed by the surgeon.

[0059] Figure 3 This illustrates an example of the structure of the ultrasonic diagnostic apparatus 2. The ultrasonic probe 10 includes a transducer array 13, a transceiver circuit 14, and a communication unit 15. The transceiver circuit 14 includes a transmitting circuit 16 and a receiving circuit 17. The transmitting circuit 16 and the receiving circuit 17 are respectively connected to the transducer array 13. Furthermore, the transceiver circuit 14 performs signal input and output with the processor 25 of the apparatus body 20 via the communication unit 15.

[0060] The oscillator array 13 has multiple oscillators (not shown) arranged in one-dimensional or two-dimensional order. These oscillators transmit ultrasonic beams UB according to drive signals supplied from the transmitting circuit 16 and receive ultrasonic echoes from the living body 30. The oscillator outputs a signal based on the received ultrasonic echoes. The oscillator is constructed, for example, by forming electrodes at both ends of a piezoelectric material. The piezoelectric material is composed of piezoelectric ceramics represented by PZT (Lead Zirconate Titanate), polymer piezoelectric elements represented by PVDF (Poly Vinylidene Di Fluoride), and piezoelectric single crystals represented by PMN-PT (Lead Magnesium Niobate-Lead Titanate).

[0061] The transmitting circuit 16 includes, for example, multiple pulse generators. Based on a transmission delay mode selected according to a control signal sent from the processor 25 of the device body 20, the transmitting circuit 16 adjusts the delay amount of a drive signal supplied to the multiple transducers included in the transducer array 13. The drive signal, adjusted by the transmitting circuit 16, causes the ultrasonic waves transmitted from the multiple transducers to form an ultrasonic beam UB. The drive signal is a pulsed or continuous wave voltage signal. When the drive signal is applied, the transducers transmit pulsed or continuous wave ultrasonic waves by extending and contracting. By synthesizing the ultrasonic waves transmitted from the multiple transducers, an ultrasonic beam UB is formed as a composite wave.

[0062] The ultrasonic beam UB transmitted into the living body 30 is reflected at sites such as blood vessels B within the living body 30, thus becoming an ultrasonic echo that propagates towards the transducer array 13. The ultrasonic echo propagating towards the transducer array 13 is thus received by multiple transducers constituting the transducer array 13. The transducers generate electrical signals by receiving the ultrasonic echoes and expanding / contracting. The electrical signals generated by the transducers are output to the receiving circuit 17.

[0063] The receiving circuit 17 generates an acoustic signal by processing the electrical signal output from the oscillator array 13 based on the control signal sent from the processor 25 of the device body 20. As an example, such as... Figure 4 As shown, the receiving circuit 17 is configured by connecting the amplification unit 41, the A / D (analog-to-digital) conversion unit 42 and the beam shaper 43 in series.

[0064] The amplification unit 41 amplifies the signals input from the multiple oscillators constituting the oscillator array 13 and sends the amplified signals to the A / D conversion unit 42. The A / D conversion unit 42 converts the signals sent from the amplification unit 41 into digital received data and sends the converted received data to the beam shaper 43. The beam shaper 43, following a sound velocity or speed of sound distribution set based on a reception delay mode selected according to a control signal sent from the processor 25 of the device body 20, applies delays to each received data converted by the A / D conversion unit 42 and adds them together. This addition process is called reception focus processing. Through this reception focus processing, an acoustic signal with the focus of the ultrasonic echo is obtained by adding the received data converted by the A / D conversion unit 42 in phase and reducing the phase of the signal.

[0065] The main body 20 of the device includes 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 mounted on the display device 21. If the main body 20 is a PC or similar device, the input device 22 can be a keyboard, mouse, trackball, touchpad, etc. The communication unit 23 communicates wirelessly with the communication unit 15 of the ultrasonic probe 10.

[0066] An input device 22 and a storage device 24 are connected to the processor 25. Furthermore, the processor 25 and the storage device 24 are connected in a manner that enables bidirectional information exchange.

[0067] Storage device 24 is a device that stores programs 26 and the like that cause the ultrasonic diagnostic device 2 to operate, such as flash memory, HDD (Hard Disc Drive), or SSD (Solid State Drive). When the main body 20 is a PC or the like, the storage device 24 can be a recording medium such as FD (Flexible Disc), MO (Magneto-Optical) disc, magnetic tape, CD (Compact Disc), DVD (Digital Versatile Disc), SD (Secure Digital) card, USB (Universal Serial Bus) memory, or a server.

[0068] The processor 25 is, for example, a CPU (Central Processing Unit). The processor 25 processes data in cooperation with RAM (Random Access Memory) (not shown) according to the program 26, thereby functioning as the main control unit 50, the image generation unit 51, the display control unit 52, the image analysis unit 53, and the emphasis display unit 54.

[0069] In addition, the processor 25 is not limited to a CPU, and can be constructed using FPGA (Field Programmable Gate Array), DSP (Digital Signal Processor), ASIC (Application Specific Integrated Circuit), GPU (Graphics Processing Unit), other ICs (Integrated Circuit), or a combination of these.

[0070] The main control unit 50 controls each part of the ultrasound diagnostic device 2 based on the input operation performed by the surgeon via the input device 22. The main control unit 50 sends the control signals to the ultrasound probe 10 via the communication unit 23. The acoustic signal generated from the ultrasound probe 10 by the receiving circuit 17 is input to the processor 25 via the communication unit 23.

[0071] Under the control of the main control unit 50, the image generation unit 51 acquires the acoustic signal input from the ultrasonic probe 10 and generates an ultrasonic image U based on the acquired acoustic signal. For example, as shown... Figure 5 As shown, the image generation unit 51 is configured by connecting the signal processing unit 61, the DSC (Digital Scan Converter) 62 and the image processing unit 63 in series.

[0072] The signal processing unit 61 corrects for attenuation caused by distance in the acoustic signal generated by the receiving circuit 17 based on the depth of the ultrasonic wave reflection position, and then generates tomographic image information, i.e., a B-mode image signal, related to the tissue in the subject body by performing envelope detection processing.

[0073] The DSC 62 converts the B-mode image signal generated by the signal processing unit 61 into an image signal according to the scanning method of a normal television signal (so-called raster conversion). The image processing unit 63 performs various image processing operations, such as grayscale 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 for which image processing has been performed by the image processing unit 63 will be referred to simply as the ultrasonic image U.

[0074] Furthermore, the transceiver circuit 14 and image generation unit 51 of the ultrasonic probe 10 are controlled by the main control unit 50 to periodically generate the ultrasonic image U at a predetermined frame rate. The transceiver circuit 14 and the image generation unit 51 function as image acquisition units for acquiring the ultrasonic image U.

[0075] Under the control of the main control unit 50, the display control unit 52 performs prescribed processing on the ultrasonic image U generated by the image generation unit 51 and displays the processed ultrasonic image U on the display device 21.

[0076] Under the control of the main control unit 50, the image analysis unit 53 generates vascular information DB by performing image analysis on the ultrasound image U input from the image generation unit 51, and outputs the generated vascular information DB to the emphasis display unit 54. The vascular information DB includes, for example, the detection results of vascular regions contained in the ultrasound image U and the arterial and venous determination results of the detected vascular vessels.

[0077] Under the control of the main control unit 50, the emphasis display unit 54 controls the display control unit 52 based on the vascular information DB input from the image analysis unit 53, thereby emphasizing the vascular region within the ultrasound image U displayed on the display device 21. Furthermore, based on the arteriovenous determination result, the emphasis display unit 54 displays in a recognizable manner whether the vascular region contains arteries or veins.

[0078] As an example, such as Figure 6 As shown, the image analysis unit 53 consists of a blood vessel detection unit 71, a structure detection unit 72, an artery and vein determination unit 73, and a correction unit 74. The ultrasound image U generated by the image generation unit 51 is input to the blood vessel detection unit 71 and the structure detection unit 72.

[0079] The vascular detection unit 71 determines vascular regions by individually detecting each blood vessel contained in the ultrasound image U, and performs arteriovenous determination of the blood vessels contained in the vascular regions. The vascular detection unit 71 inputs the detection results of the vascular regions and the arteriovenous determination results of each vascular region as vascular detection information D1 to the correction unit 74 and the arteriovenous determination unit 73. Furthermore, information related to at least the vascular regions detected by the vascular detection unit 71 can be input to the arteriovenous determination unit 73.

[0080] The structure detection unit 72 detects structural regions containing structures such as tendons, bones, nerves, or muscles based on the ultrasound image U, and outputs information representing the detected structural regions as structural detection information D2 to the arteriovenous determination unit 73.

[0081] The arteriovenous determination unit 73 determines the arteries and veins of the vessels within the vessel region based on the anatomical relative position of the vessel region included in the vessel detection information D1 and the structural region included in the structural detection information D2. In other words, the arteriovenous determination unit 73 uses the structural region as a marker and determines the arteries and veins based on the relative position of the vessels relative to the marker. The arteriovenous determination unit 73 outputs information representing the arteriovenous determination result as arteriovenous determination information D3 to the correction unit 74.

[0082] The correction unit 74 corrects the arterial and vein determination results contained in the vascular detection information D1 based on the arterial and vein determination information D3. The correction unit 74 outputs the corrected vascular detection information D1 as the aforementioned vascular information DB to the emphasis display unit 54.

[0083] Figure 7 This illustrates an example of blood vessel detection processing based on the blood vessel detection unit 71. The blood vessel detection unit 71 performs processing to detect a blood vessel region Ra containing a blood vessel B within an ultrasound image U using a known algorithm, and determines whether the blood vessel B within the blood vessel region Ra is an artery or vein. If the blood vessel region Ra contains a blood vessel B determined to be an artery, it is represented by a dashed line. If the blood vessel region Ra contains a blood vessel B determined to be a vein, it is represented by a solid line. Figure 7 The vascular region Ra shown includes blood vessel B, which is identified as an artery. Furthermore, in the case where the ultrasound image U contains multiple blood vessels B, the vascular detection unit 71 detects the vascular region Ra for each blood vessel B.

[0084] The vascular region Ra is assigned a "label" to indicate the arteriovenous determination result and a "score" to indicate the confidence (i.e., certainty) of the determination result. The label indicates whether the vessel B contained in the vascular region Ra is an "artery" or a "vein". The score is a value between 0 and 1, with the closer to 1, the higher the confidence. The vascular region Ra assigned a label and score corresponds to the vascular detection information D1 mentioned above.

[0085] exist Figure 7 In the ultrasound image U shown, in addition to blood vessel B, the extensor carpi radialis longus (hereinafter referred to as ECRL) and extensor carpi radialis brevis (hereinafter referred to as ECRB) are also visible in the human wrist joint. ECRL and ECRB are examples of structures.

[0086] In this embodiment, the blood vessel detection unit 71 uses a learning-completed model generated through machine learning, namely the blood vessel detection model 71A (see reference). Figure 8 The blood vessel detection model 71A performs blood vessel detection processing. For example, it uses a deep learning object detection algorithm. As the blood vessel detection model 71A, for example, it can use an object detection model composed of R-CNN (Regional CNN), which is a type of convolutional neural network (CNN).

[0087] The blood vessel detection model 71A detects regions containing individual blood vessels as objects within the ultrasound image U and determines labels for the detected regions. Then, the blood vessel detection model 71A outputs information representing the detected blood vessel region Ra along with the labels and scores.

[0088] Figure 8 This diagram illustrates an example of the learning phase of a blood vessel detection model 71A using machine learning. The blood vessel detection model 71A learns using teacher data TD1. Teacher data TD1 contains multiple teacher images P assigned the correct label L. The teacher images P in teacher data TD1 are sample images of individual blood vessels (arteries and veins). Teacher data TD1 contains various teacher images P with different blood vessel shapes, sizes, etc.

[0089] During the learning phase, a teacher image P is input into the blood vessel detection model 71A. The blood vessel detection model 71A outputs a judgment result A for the teacher image P. Based on this judgment result A and the correct label L, a loss calculation using a loss function is performed. Then, based on the result of the loss calculation, the various coefficients of the blood vessel detection model 71A are updated, and the blood vessel detection model 71A is updated according to the updated settings.

[0090] During the learning phase, a series of processes are repeatedly performed, including inputting the teacher image P to the blood vessel detection model 71A, outputting the judgment result A from the blood vessel detection model 71A, loss calculation, setting update, and updating the blood vessel detection model 71A. This series of processes is repeated until the detection accuracy reaches a predetermined set level. After the blood vessel detection model 71A with the detection accuracy reaching the set level is stored in the storage device 24, it is used by the blood vessel detection unit 71 in the application phase, i.e., blood vessel detection processing.

[0091] Figure 9 This is an example of structure detection processing based on the structure detection unit 72. The structure detection unit 72 performs processing to detect the structure region Rb containing the structure from the ultrasonic image U using a known algorithm.

[0092] In this embodiment, the vascular structure detection unit 72 uses a learned end model generated through machine learning, namely the vascular structure detection model 72A (reference). Figure 10 The system performs vascular structure detection. The vascular structure detection model 72A, for example, uses a deep learning-based object detection algorithm. As the structure detection model 72A, for example, an object detection model composed of R-CNN, a type of CNN, can be used.

[0093] The structure detection unit 72 detects a structure region Rb containing a structure that is an object from the ultrasonic image U. The information representing the structure region Rb corresponds to the aforementioned structure detection information D2. Figure 9 In the example shown, the structure detection unit 72 detects the region containing ECRL and ECRB as the structure region Rb. Alternatively, the structure detection unit 72 can detect the region containing ECRL and the region containing ECRB as structure regions Rb respectively.

[0094] Figure 10 This diagram illustrates an example of the learning phase of the structure detection model 72A using machine learning. The structure detection model 72A learns using teacher data TD2. Teacher data TD2 contains multiple teacher images P assigned the correct label L. The teacher images P in teacher data TD2 are sample images of structures. Teacher data TD2 contains teacher images P of various structures with different types (tendons, bones, nerves, or muscles, etc.), quantities, shapes, sizes, and textures.

[0095] During the learning phase, a teacher image P is input into the structure detection model 72A. The structure detection model 72A outputs a judgment result A for the teacher image P. Based on this judgment result A and the correct label L, a loss calculation using a loss function is performed. Then, based on the result of the loss calculation, the various coefficients of the structure detection model 72A are updated, and the structure detection model 72A is updated according to the updated settings.

[0096] During the learning phase, a series of processes are repeatedly performed, including inputting the teacher image P to the structure detection model 72A, outputting the judgment result A from the structure detection model 72A, loss calculation, updating settings, and updating the structure detection model 72A. This series of processes is repeated until the detection accuracy reaches a predetermined set level. After the structure detection model 72A with the detection accuracy reaching the set level is stored in the storage device 24, it is used by the structure detection unit 72 in the application phase, i.e., the structure detection processing.

[0097] Figure 11 This describes an example of arteriovenous determination processing based on the arteriovenous determination unit 73. The arteriovenous determination unit 73 determines the arteriovenous status of blood vessel B within blood vessel region Ra based on the anatomical relative position of blood vessel region Ra contained in blood vessel detection information D1 and structural region Rb contained in structural detection information D2. The arteriovenous determination unit 73 calculates scores for "artery" and "vein," which serve as labels for blood vessel B, and selects the label with the larger score. The arteriovenous determination unit 73 generates arteriovenous determination information D3 by calculating the label and score for blood vessel B within blood vessel region Ra.

[0098] The arteriovenous determination unit 73, for example, uses a learning-completed model for arteriovenous determination processing. This learning-completed model is learned using teacher data representing the anatomical relative positional relationship between the vascular region Ra and the structural region Rb. Alternatively, the arteriovenous determination unit 73 can use known data representing the anatomical relative positional relationship for arteriovenous determination.

[0099] Figure 11 The vascular region Ra shown is located on the radial side (i.e., the radial side) relative to the structural region Rb containing ECRL and ECRB. Therefore, the vessel B contained in the vascular region Ra is identified as a vein (specifically, a radial cutaneous vein). Thus, arteriovenous identification of vessels is improved by using their anatomical relative position to the structure.

[0100] Figure 12 This illustrates an example of correction processing based on correction unit 74. For the corresponding blood vessel B, correction unit 74 compares the score contained in the arteriovenous determination information D3 with the score contained in the blood vessel detection information D1, selecting the label with the higher score. For example, in... Figure 12 In the example shown, for blood vessel B, the score (0.90) contained in the arteriovenous determination information D3 is higher than the score (0.75) contained in the blood vessel detection information D1. Therefore, the correction unit 74 selects the label (vein) contained in the arteriovenous determination information D3 instead of the label (artery) contained in the blood vessel detection information D1 as the label for blood vessel B.

[0101] Thus, if the score contained in the arteriovenous determination information D3 is higher than the score contained in the vascular detection information D1, the label contained in the vascular detection information D1 is corrected.

[0102] The correction unit 74 outputs the corrected vascular detection information D1 with the corrected label as the aforementioned vascular information DB to the emphasis display unit 54. The vascular information DB contains the location information of the vascular region Ra within the ultrasound image U, as well as the label and score for the vascular region Ra.

[0103] Figure 13 This illustrates an example of emphasis display processing based on the emphasis display unit 54. The emphasis display unit 54 displays a blood vessel region Ra within the ultrasound image U displayed on the display device 21 of the device body 20, using a rectangular frame based on blood vessel information DB. Furthermore, the emphasis display unit 54 displays, in a recognizable manner, whether the blood vessel contained within the blood vessel region Ra is an artery or a vein, based on the artery / vein determination result. Figure 13In the example shown, solid lines represent the vascular region Ra containing veins, and dashed lines represent the vascular region Ra containing arteries. Furthermore, it is emphasized that the display unit 54 is not limited to the type of line; it can also display the vascular region Ra in a recognizable manner based on line thickness, color, brightness, etc.

[0104] Next, use Figure 14 The flowchart shown illustrates an example of the operation of the ultrasound diagnostic device 2. First, the main control unit 50 uses the input device 22, etc., and the surgeon determines whether a start operation has been performed (step S10). If the main control unit 50 determines that a start operation has been performed (step S10: Yes), it activates the transceiver circuit 14 of the ultrasound probe 10 and the image generation unit 51 to generate an ultrasound image U (step S11). The generated ultrasound image U is displayed on the display device 21 via the display control unit 52.

[0105] At this time, as Figure 2 As shown, the surgeon places the ultrasonic probe 10 into the surface of the living body 30. The ultrasonic beam UB is transmitted from the transducer array 13 into the living body 30 according to the drive signal input from the transmitting circuit 16. The ultrasonic echoes from within the living body 30 are received by the transducer array 13, and the received signal is output to the receiving circuit 17. The received signal received by the receiving circuit 17 is converted into an acoustic signal via the amplification unit 41, the A / D conversion unit 42, and the beam shaper 43. This acoustic signal is output to the device body 20 via the communication unit 15.

[0106] The main body 20 receives the acoustic signal output from the ultrasonic probe 10 via the communication unit 23. The acoustic signal received by the main body 20 is output to the image generation unit 51. In the image generation unit 51, the acoustic signal undergoes envelope detection processing in the signal processing unit 61 to become a B-mode image signal. After passing through the DSC 62 and the image processing unit 63, it is output as an ultrasonic image U to the display control unit 52. Furthermore, the ultrasonic image U is output to the image analysis unit 53.

[0107] In the image analysis unit 53, the aforementioned blood vessel detection processing is performed by the blood vessel detection unit 71 (see reference). Figure 7 (Step S12). The vascular detection information D1 generated by the vascular detection processing is output to the correction unit 74 and the arteriovenous determination unit 73.

[0108] Furthermore, step S13 is executed in parallel with step S12. In step S13, the aforementioned structure detection process is performed by the structure detection unit 72 (see reference). Figure 9 The structure detection information D2 generated by this structure detection processing is output to the arteriovenous determination unit 73. In step S14, the arteriovenous determination unit 73 performs the above-mentioned arteriovenous determination processing (see reference). Figure 11The arterial and vein determination information D3 generated by this arterial and vein determination process is output to the correction unit 74.

[0109] Next, the aforementioned correction process is performed by the correction unit 74 (see reference). Figure 12 (Step S15). In this correction process, the label of the vascular region Ra contained in the vascular detection information D1 is corrected based on the arteriovenous determination information D3. As a result of this correction process, the vascular information DB is output to the emphasis display unit 54.

[0110] Then, the above-mentioned emphasis display processing is performed by the emphasis display unit 54 (see reference). Figure 13 (Step S16). Through this emphasis display process, the vascular region Ra is emphasized within the ultrasound image U displayed on the display device 21. Furthermore, the vascular region Ra clearly indicates whether the blood vessel contained within it is an artery or a vein. Thus, by emphasizing the display, the surgeon can accurately determine the location of the blood vessel present in the ultrasound image U and accurately determine whether the blood vessel is an artery or a vein.

[0111] Next, the main control unit 50 uses the input device 22, etc., and the surgeon determines whether a termination operation has been performed (step S17). If the main control unit 50 determines that a termination operation has not been performed (step S17: No), the process returns to step S11. A new ultrasound image U is then generated. On the other hand, if the main control unit 50 determines that a termination operation has been performed (step S17: Yes), the operation of the ultrasound diagnostic device 2 is terminated.

[0112] Previously, blood vessels were detected through vascular imaging, and arteries and veins were individually identified from the detected vessels. This method is prone to errors in arteriovenous identification, as the identification varies from frame to frame. When the surgeon performs punctures based on these identification results, they may sometimes mistakenly puncture the wrong vessel.

[0113] In contrast, according to the technology of the present invention, the arteries and veins of blood vessels are determined by utilizing the anatomical relative positional relationship between blood vessels and structures detected from ultrasound images, thus improving the accuracy of arteriovenous identification. As a result, the surgeon can accurately pinpoint the blood vessels (e.g., veins) of the puncture target.

[0114] Next, use Figures 15-19 Explanation for Figure 7 Examples of different ultrasound images U shown are provided for blood vessel detection processing, structure detection processing, arteriovenous identification processing, correction processing, and emphasis display processing.

[0115] like Figure 15As shown, the ultrasound image U in this example displays two blood vessels, B1 and B2. Furthermore, the ultrasound image U shows the abductor pollicis longus muscle (hereinafter referred to as APL) and the radial styloid process (hereinafter simply referred to as the styloid process). APL and the styloid process are examples of structures.

[0116] In the vascular detection processing, the vascular detection unit 71 detects the regions containing blood vessels B1 and B2 within the ultrasound image U as vascular regions Ra, and assigns labels and scores to each vascular region Ra. The two vascular regions Ra assigned labels and scores correspond to the aforementioned vascular detection information D1. In this example, both blood vessels B1 and B2 are labeled as "vein".

[0117] like Figure 16 As shown, in the structure detection processing, the area containing APL and the area containing bony prominences within the ultrasound image U are detected as structure regions Rb by the structure detection unit 72. This indicates that the information of the two structure regions Rb corresponds to the aforementioned structure detection information D2.

[0118] like Figure 17 As shown, in the arteriovenous determination process, the arteriovenous determination unit 73 determines the arteriovenous status of two vascular regions Ra based on their anatomical relative positions with two structural regions Rb. The arteriovenous determination unit 73 generates arteriovenous determination information D3 by calculating the labels and scores for vessels B1 and B2 respectively.

[0119] like Figure 18 As shown, in the correction process, for both blood vessels B1 and B2, the correction unit 74 compares the scores contained in the arteriovenous determination information D3 with the scores contained in the blood vessel detection information D1, and selects the label with the higher score. In this example, for either blood vessel B1 or B2, the score contained in the arteriovenous determination information D3 is higher than the score contained in the blood vessel detection information D1, therefore the label contained in the arteriovenous determination information D3 is selected. In this example, among the labels of blood vessels B1 and B2 contained in the blood vessel detection information D1, the label of blood vessel B1 is corrected from "vein" to "artery".

[0120] like Figure 19 As shown, in the emphasis display process, the vascular region Ra within the ultrasound image U is emphasized according to the corrected label. In this example, the vascular region Ra containing the vessel B1 identified as an "artery" is represented by a dashed line, and the vascular region Ra containing the vessel B2 identified as a "vein" is represented by a solid line.

[0121] [Variation Example]

[0122] Hereinafter, various modifications of the ultrasonic diagnostic apparatus 2 according to the first embodiment described above will be described.

[0123] In the first embodiment, the arteriovenous determination unit 73 performs arteriovenous determination processing (see reference). Figure 11 In the above, scores are calculated for the labels "artery" and "vein" used to label blood vessel B, and the label with the higher score is selected. Alternatively, for example, ... Figure 20 As shown, you can set a score threshold and select labels with scores above the threshold.

[0124] And, for example, Figure 21 The diagram shows the case where scores are calculated for the labels "artery" and "vein" used to identify blood vessel B, and both scores are less than a threshold. This corresponds to a case where the accuracy of arteriovenous determination based on the relative positional relationship with the structure is low. In this case, the arteriovenous determination unit 73 may also consider the label determination (i.e., arteriovenous determination) difficult and stop the arteriovenous determination. Thus, in cases where arteriovenous determination is difficult, the emphasis display unit 54 can display the blood vessel region Ra within the ultrasound image U without distinguishing whether it is an "artery" or a "vein". In this case, the emphasis display unit 54 can simply display the blood vessel region Ra as a "blood vessel".

[0125] Furthermore, if no structure is found in the ultrasound image U through structure detection processing, the display unit 54 can display the vascular region Ra in the ultrasound image U without distinguishing whether it is an "artery" or a "vein".

[0126] Furthermore, in cases where no structure is found in the ultrasound image U through structure detection processing, and in cases where the accuracy of arteriovenous determination based on the relative positional relationship with the structure is low, the emphasis display unit 54 can emphasize the display based on the tags contained in the blood vessel detection information D1.

[0127] And, for example, such as Figure 22 As shown, the display unit 54 can establish a corresponding association between the score of the label selected by the correction unit 74 (i.e., the confidence level of the judgment result selected by the correction unit 74) and the vascular region Ra. Therefore, the surgeon can grasp the confidence level of the arterial and venous determination for each blood vessel.

[0128] And, for example, such as Figure 23 As shown, if the score of the label selected by the correction unit 74 (i.e., the reliability of the judgment result selected by the correction unit 74) is lower than a certain value, the emphasis display unit 54 can display information reminding the surgeon to pay attention. Therefore, the surgeon can reliably grasp situations where the reliability of the arterial / venous puncture is low and precautions are needed during the puncture.

[0129] Furthermore, the arteriovenous determination unit 73 can change the criteria for arteriovenous determination based on the type of structure detected by the structure detection unit 72. This is because, for example, if a structure has anatomically typical characteristics in relation to a blood vessel, the probability of a correct determination is high even if the arteriovenous determination score is low. The arteriovenous determination unit 73 changes the score threshold based on, for example, the type of structure (see reference). Figure 21 Specifically, when a structure possesses anatomically typical features in relation to blood vessels, the threshold is set lower than when it does not possess anatomically typical features. Furthermore, the threshold is not limited to a fractional threshold; the criteria for arteriovenous identification can be altered by changing the algorithm used for arteriovenous identification.

[0130] Furthermore, in the first embodiment, the labels for blood vessels are set to two types: "artery" and "vein," but the labels can be further subdivided. For example, "vein" can be subdivided into "radial cutaneous vein," "ulnar cutaneous vein," etc. Thus, the arteriovenous determination unit 73 can determine the type of blood vessel in addition to determining the type of artery and vein. In this case, the emphasis display unit 54 can establish a corresponding association between the type of blood vessel and the blood vessel region Ra for display.

[0131] Furthermore, in the first embodiment, the blood vessel detection unit 71 performs arterial and venous determination of the blood vessel; however, it may also skip arterial and venous determination and only detect the blood vessel region Ra. In this case, the correction unit 74 does not need to correct the arterial and venous determination result based on the blood vessel detection unit 71.

[0132] Furthermore, in the first embodiment, the blood vessel detection unit 71 and the structure detection unit 72 are each configured using separate object detection models; however, it is also possible to configure the blood vessel detection unit 71 and the structure detection unit 72 using only one object detection model. In this case, teacher data including teacher images of individual blood vessels and teacher images of structures can be used to learn the object detection model. Furthermore, it is also possible to configure the blood vessel detection unit 71, the structure detection unit 72, and the arteriovenous determination unit 73 using only one object detection model. Additionally, it is also possible to configure the blood vessel detection unit 71, the structure detection unit 72, the arteriovenous determination unit 73, and the correction unit 74 using only one object detection model.

[0133] Furthermore, in the first embodiment, the blood vessel detection unit 71 and the structure detection unit 72 are composed of an object detection model including a CNN. However, the object detection model is not limited to a CNN and may be a separator or other general detection model.

[0134] Furthermore, the object detection models constituting the blood vessel detection unit 71 and the structure detection unit 72 can be composed of a recognizer that performs object recognition based on image feature quantities such as AdaBoost or SVM (Support Vector Machine). In this case, the teacher image can be converted into a feature vector, and then the recognizer can be learned based on the feature vector.

[0135] Furthermore, the vessel detection unit 71 and the structure detection unit 72 are not limited to object detection models based on machine learning; object detection can be performed through template matching. In this case, the vessel detection unit 71 pre-stores typical pattern data of individual vessels as templates, and while searching within the ultrasound image U using the templates, calculates the similarity with the pattern data. Then, the vessel detection unit 71 identifies the location with the highest similarity (above a specified value) as the vessel region Ra. Similarly, the structure detection unit 72 pre-stores typical pattern data of structures as templates, and while searching within the ultrasound image U using the templates, calculates the similarity with the pattern data. Then, the structure detection unit 72 identifies the location with the highest similarity (above a specified value) as the vessel region Ra. Additionally, the template can be a portion of an actual ultrasound image or an image drawn by patterning a vessel or structure.

[0136] Furthermore, when calculating similarity, in addition to simple template matching, methods can be used such as machine learning methods 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 general image recognition methods 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).

[0137] In the first embodiment, the ultrasonic probe 10 is wirelessly connected to the device body 20, but alternatively, the ultrasonic probe 10 and the device body 20 can also be wired.

[0138] Furthermore, in the first embodiment, the image generation unit 51 that generates an ultrasonic image U based on the acoustic signal is provided in the device main body 20. However, alternatively, the image generation unit 51 may also be provided inside the ultrasonic probe 10. In this case, the ultrasonic probe 10 generates an ultrasonic image U and outputs it to the device main body 20. The processor 25 of the device main body 20 performs image analysis, etc., based on the ultrasonic image U input from the ultrasonic probe 10.

[0139] Furthermore, 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, and the ultrasonic probe 10 can also be indirectly connected to the processor 25 via a network.

[0140] As an example, Figure 24 In the ultrasonic diagnostic apparatus 2A shown, the display device 21, input device 22, and ultrasonic probe 10A are connected to the apparatus main body 20A via a network NW. In the apparatus main body 20A, the display device 21 and input device 22 are removed from the apparatus main body 20 according to the first embodiment, and a transceiver circuit 14 is added, thus constituting the apparatus with the transceiver circuit 14, storage device 24, and processor 25. In the ultrasonic probe 10A, the transceiver circuit 14 is removed from the ultrasonic probe 10 according to the first embodiment.

[0141] Thus, in the ultrasound diagnostic device 2A, the display device 21, input device 22, and ultrasound probe 10A are connected to the device body 20A via a network NW, allowing the device body 20A to be used as a so-called remote server. Consequently, for example, the surgeon can have the display device 21, input device 22, and ultrasound probe 10A readily available, improving convenience. Furthermore, the convenience is further enhanced by using a mobile terminal such as a smartphone or tablet to configure the display device 21 and input device 22.

[0142] As another example, in Figure 25 In the ultrasound diagnostic device 2B shown, the display device 21 and the input device 22 are mounted on the device body 20B, and the ultrasound probe 10A is connected to the device body 20B via a network NW. In this case, the device body 20B can be configured as a remote server. Furthermore, the device body 20B can also be configured as a mobile terminal such as a smartphone or tablet.

[0143] In the first embodiment, the hardware structure of the processing unit, which performs various processes such as the main control unit 50, the image generation unit 51, the display control unit 52, the image analysis unit 53, and the emphasis display unit 54, can use various processors as shown below. These processors include general-purpose processors (CPUs) that execute software (program 26) as described above to function as various processing units, as well as processors such as FPGAs (programmable logic devices, PLDs) whose circuit structure can be changed after manufacturing, and processors such as ASICs (dedicated circuits) with circuit structures specifically designed for performing specific processes.

[0144] A processing unit can consist of one of these various processors, or it can consist of a combination of two or more processors of the same or different types (e.g., a combination of multiple FPGAs and / or a combination of a CPU and an FPGA). Furthermore, multiple processing units can be composed of a single processor.

[0145] As examples of a single processor comprising multiple processing units, there are two main approaches: First, as exemplified by computers such as client machines and servers, a processor is constructed using a combination of one or more CPUs and software, and this processor functions as multiple processing units. Second, as exemplified by System-on-Chip (SoC), a processor that implements the overall system functionality including multiple processing units using a single IC chip. In this way, various processing units are constructed using one or more of the aforementioned processors as the hardware structure.

[0146] Furthermore, the hardware architecture of these various processors, more specifically, can utilize circuits composed of circuit elements such as semiconductor components.

[0147] The above content provides a grasp of the techniques described in notes 1 to 9 below.

[0148] [Note 1]

[0149] An information processing apparatus is provided for processing an ultrasonic image generated by transmitting an ultrasonic beam from an array of transducers toward a living body and receiving ultrasonic echoes generated within the living body. The information processing apparatus includes a processor.

[0150] The processor detects vascular regions containing blood vessels from the ultrasound image.

[0151] Detect structures other than blood vessels from the ultrasound image.

[0152] Based on the relative position of the vascular region to the structure, determine whether the blood vessel contained in the vascular region is an artery or a vein.

[0153] [Note 2]

[0154] According to the information processing apparatus described in Appendix 1, wherein,

[0155] The processor displays the vascular region within the ultrasound image shown on the display device in a manner that allows identification of whether the blood vessels contained within the vascular region are arteries or veins.

[0156] [Note 3]

[0157] According to the information processing apparatus described in Appendix 2, wherein,

[0158] The processor determines, based on the vascular region, whether the blood vessel contained in the vascular region is an artery or a vein.

[0159] [Note 4]

[0160] According to the information processing apparatus described in Appendix 3, wherein...

[0161] The processor corrects the arteriovenous determination result based on the relative positional relationship between the blood vessel region and the structure.

[0162] [Note 5]

[0163] According to the information processing apparatus described in Appendix 4, wherein...

[0164] The processor compares the reliability of arteriovenous determination based on the vascular region with the reliability of arteriovenous determination based on the relative positional relationship between the vascular region and the structure, and selects the determination result with higher reliability.

[0165] [Note 6]

[0166] According to the information processing apparatus described in Appendix 5, wherein...

[0167] The processor displays the confidence level of the determination result for the selection on the display device.

[0168] [Note 7]

[0169] According to the information processing apparatus described in Appendix 6, wherein...

[0170] If the reliability of the determination result for the selection is lower than a certain value, the processor will display a reminder message on the display device.

[0171] The technology of this invention can also be appropriately combined with the various embodiments and / or variations described above. Furthermore, it is not limited to the embodiments described above; various structures can certainly be adopted as long as they do not depart from the spirit of the invention.

[0172] The descriptions and illustrations above are detailed explanations of a portion of the technology involved in this invention, and are merely one example of the technology of this invention. For example, the descriptions related to the above-described structure, function, effect, and effect are examples of the structure, function, effect, and effect of the portion involved in the technology of this invention. Therefore, without departing from the technical spirit of this invention, unnecessary parts of the descriptions and illustrations above may be deleted, or new elements may be added or replaced. Furthermore, to avoid complications and to facilitate understanding of the portion involved in the technology of this invention, descriptions related to common technical knowledge that does not require special explanation in aspects enabling the implementation of this invention have been omitted from the descriptions and illustrations above.

[0173] In this specification, "A and / or B" has the same meaning as "at least one of A and B". That is, "A and / or B" can mean only A, only B, or a combination of A and B.

[0174] All documents, patent applications and technical standards described in this specification, and the specific and separately described documents, patent applications and technical standards incorporated herein by reference, are incorporated herein by reference to the same extent.

[0175] Symbol Explanation

[0176] 2. 2A, 2B - Ultrasonic diagnostic device; 10, 10A - Ultrasonic probe; 11 - Housing; 11A - Array housing; 11B - Handle; 13 - Vibrator array; 14 - Transceiver circuit; 15 - Communication unit; 16 - Transmitting circuit; 17 - Receiving circuit; 20 - Device body; 20A, 20B - Device body; 21 - Display device; 22 - Input device; 23 - Communication unit; 24 - Storage device; 25 - Processor; 26 - Program; 30 - Living organism; 31 - Puncture needle; 41 - Amplification unit; 42 - A / D conversion unit; 43 - Beam shaper; 50 - Main control unit; 51 - Image generation unit; 52 - Display control unit; 53 - Image Analysis Unit, 54 - Emphasis Display Unit, 61 - Signal Processing Unit, 62 - DSC, 63 - Image Processing Unit, 71 - Vessel Detection Unit, 71A - Vessel Detection Model, 72 - Structure Detection Unit, 72A - Structure Detection Model, 73 - Artery and Vein Determination Unit, 74 - Correction Unit, A - Determination Result, B, B1, B2 - Vessels, D1 - Vessel Detection Information, D2 - Structure Detection Information, D3 - Artery and Vein Determination Information, DB - Vessel Information, L - Correct Label, M - Guiding Marker, NW - Network, P - Teacher Image, Ra - Vessel Region, Rb - Structure Region, TD1, TD2 - Teacher Data, U - Ultrasound Image, UB - Ultrasound Beam.

Claims

1. An information processing apparatus for processing an ultrasonic image generated by transmitting an ultrasonic beam from an array of transducers toward a living body and receiving ultrasonic echoes generated within the living body, the information processing apparatus comprising: A vascular detection unit that detects vascular regions containing blood vessels from the ultrasound image; A structure detection unit detects structures other than blood vessels from the ultrasound image; and The artery and vein determination unit uses the structure as a marker and determines whether the blood vessel contained in the blood vessel region is an artery or a vein based on the relative positional relationship between the marker and the blood vessel region.

2. The information processing apparatus according to claim 1, wherein, The information processing device also includes an emphasis display unit that displays the vascular region in the ultrasound image displayed on the display device in a way that allows identification of whether the blood vessel contained in the vascular region is an artery or a vein.

3. The information processing apparatus according to claim 2, wherein, In addition to detecting the vascular region, the vascular detection unit also determines whether the blood vessel contained in the vascular region is an artery or a vein.

4. The information processing apparatus according to claim 3, wherein, The information processing device also includes a correction unit that corrects the arterial and vein determination result determined by the blood vessel detection unit based on the arterial and vein determination result of the arterial and vein determination unit.

5. The information processing apparatus according to claim 4, wherein, The correction unit compares the reliability of the arteriovenous determination performed by the blood vessel detection unit with the reliability of the arteriovenous determination performed by the arteriovenous determination unit, and selects the determination result with higher reliability.

6. The information processing apparatus according to claim 5, wherein, The emphasis display unit displays the credibility of the determination result selected by the correction unit on the display device.

7. The information processing apparatus according to claim 6, wherein, When the confidence level of the determination result selected by the correction unit is lower than a certain value, the emphasis display unit will display a reminder message on the display device.

8. An information processing method comprising processing an ultrasonic image generated by transmitting an ultrasonic beam from a transducer array toward a living body and receiving ultrasonic echoes generated within the living body, wherein in the information processing method, Detect vascular regions containing blood vessels from the ultrasound image. Detect structures other than blood vessels from the ultrasound image. Using the structure as a marker, the relative positional relationship between the marker and the vascular region determines whether the blood vessel contained in the vascular region is an artery or a vein.

9. A computer-readable storage medium storing a program that causes the computer to perform processing on an ultrasonic image generated by transmitting an ultrasonic beam from an array of transducers toward a living body and receiving ultrasonic echoes generated within the living body, the program causing the computer to perform the following processing: Detect vascular regions containing blood vessels from the ultrasound image. Detect structures other than blood vessels from the ultrasound image. Using the structure as a marker, the relative positional relationship between the marker and the vascular region determines whether the blood vessel contained in the vascular region is an artery or a vein.

Images