Ultrasound diagnostic apparatus and control method thereof

Abstract

The invention discloses ultrasonic diagnosis equipment and a control method thereof. The control method of the ultrasonic diagnosis equipment comprises the steps that an animal ultrasonic image is obtained; extracting, from the animal ultrasound image, a cross-sectional image including cross-sections of a plurality of target structures; obtaining a processed image including a contour drawing image of a plurality of target structures from the cross-sectional image; obtaining the center of gravity of each contour drawing image in the contour drawing images; and obtaining length information of the target structure based on an intersection point between at least one line passing through the center of gravity of the contour drawing image and the boundary of the contour drawing image.

Description

Technical Field

[0001] This disclosure relates to an ultrasound diagnostic device for acquiring ultrasound images and a control method thereof. Background Technology

[0002] In the medical field, various medical imaging devices have recently been widely used to image and obtain information about human biological tissues for the purpose of early diagnosis of various diseases or surgery. Representative examples of such medical imaging devices include ultrasound diagnostic equipment, CT equipment, and MRI equipment.

[0003] An ultrasound imaging device is a means of transmitting ultrasound signals generated by a transducer of a probe to an object and obtaining, non-invasively, at least one image of an internal region of the object (e.g., soft tissue or blood flow) by receiving information from signals reflected from the object. In particular, ultrasound diagnostic devices are used for medical purposes (such as observing the interior of an object, detecting foreign bodies, and measuring damage). Such ultrasound diagnostic devices are widely used alongside other imaging diagnostic devices because ultrasound imaging devices are more stable than diagnostic devices that use X-rays, can display images in real time, and are safe because there is no radiation exposure.

[0004] In animal echocardiography, the ratio of the diameter of the left atrium (LA) to the diameter of the aorta (AO) is an important indicator for assessing heart health (especially whether the left atrium is enlarged). Therefore, it is necessary to accurately measure the ratio of the diameter of the left atrium to the aorta.

[0005] However, because each veterinarian defines the boundaries of the heart structure based on different standards, errors can occur in the measurement results between veterinarians, which may lead to inaccurate diagnoses. Summary of the Invention

[0006] One aspect of this disclosure is to provide an ultrasound diagnostic device and a control method thereof, the ultrasound diagnostic device and control method being designed to use a machine learning model to determine the boundary between the left atrium and the aorta in an animal cardiac ultrasound image and to measure the diameter ratio of the left atrium to the aorta according to standardized criteria.

[0007] Other aspects of this disclosure will be set forth in part in the description which follows, and will be readily understood from the description, or may be learned by practicing this disclosure.

[0008] According to one aspect of this disclosure, a method for controlling an ultrasound diagnostic device may include: acquiring an animal ultrasound image; extracting a cross-sectional image from the animal ultrasound image comprising a cross-section of a plurality of target structures; obtaining a processed image from the cross-sectional image, the processed image comprising a contour drawing image of the plurality of target structures; obtaining the centroid of each of the contour drawing images; and obtaining length information of the target structures based on the intersection point between at least one line passing through the centroid of the contour drawing image and the boundary of the contour drawing image.

[0009] According to one aspect of this disclosure, an ultrasound diagnostic device may include: a probe configured to send ultrasound signals to an object and receive echo signals reflected from the object; a display configured to display an ultrasound image obtained based on echo information received by the probe; an input interface configured to receive user input; and a processor configured to: control the probe to obtain an animal ultrasound image; extract a cross-sectional image from the animal ultrasound image comprising a cross-section of a plurality of target structures; obtain a processed image from the cross-sectional image comprising a contour drawing image of the plurality of target structures; obtain the centroid of each of the contour drawing images; and obtain length information of the target structures based on the intersection point between at least one line passing through the centroid of the contour drawing image and the boundary of the contour drawing image. Attached Figure Description

[0010] These and / or other aspects of this disclosure will become clearer and more readily understood from the following description of embodiments taken in conjunction with the accompanying drawings, in which: Figure 1 The control block diagram of the ultrasound imaging system is shown when the probe is a wired probe or a hybrid probe. Figure 2 A control block diagram of an ultrasound imaging system is shown when the probe is a wireless probe or a hybrid probe. Figures 3 to 6 This is a view illustrating an ultrasound imaging system according to an embodiment; Figure 7 Examples of cross-sectional echocardiographic images of animals and methods for manually measuring the diameter ratio of the left atrium to the aorta are shown in the embodiments; Figure 8 An example of input data to a machine learning model according to an embodiment is shown; Figure 9 An example of output data from a machine learning model, according to an embodiment, is shown; Figure 10 This is an illustration of a method for obtaining a processed image of a contour drawing image including multiple target structures using a machine learning model according to an embodiment; Figure 11An example of training data used to train a machine learning model is shown; Figure 12 Another example of training data used to train a machine learning model is shown; Figure 13 and Figure 14 This is an illustration of an example of a standardized method for obtaining length information of multiple target structures from a processed image; Figure 15 This is an illustration of another example of a standardized method for obtaining length information of multiple target structures from a processed image; Figure 16 This is an illustration of an image displayed on a monitor of an ultrasound diagnostic device, showing an example of a standardized method for obtaining length information of multiple target structures from a processed image; and Figure 17 This is a control flowchart of an ultrasound diagnostic device according to an embodiment. Detailed Implementation

[0011] This disclosure will illustrate the principles of the disclosure and disclose embodiments thereof in order to clarify the scope of the claims of the disclosure and enable those skilled in the art to practice the embodiments thereof. Embodiments of the disclosure may be implemented in various forms.

[0012] Throughout this specification, the same reference numerals refer to the same elements. This specification does not describe all components of the embodiments, and will not describe general content or repetition between embodiments within the art to which this disclosure pertains. As used in this specification, a “module” or “unit” may be implemented as one or more of software, hardware, and firmware, or a combination of two or more of these. According to embodiments, multiple “modules” or “units” may be implemented as a single element, or a single “module” or “unit” may include multiple elements.

[0013] Unless the relevant context clearly indicates otherwise, the singular form of a noun corresponding to an item may include a single item or multiple items.

[0014] In this disclosure, each of the phrases such as “A or B”, “at least one of A and B”, “at least one of A or B”, “A, B or C”, “at least one of A, B and C”, and “at least one of A, B or C” may include any one of the items listed together in the corresponding phrase or all possible combinations thereof.

[0015] The term “and / or” includes any combination of multiple related components or any one of multiple related components.

[0016] Terms such as “first,” “second,” “primary,” and “secondary” can be simply used to distinguish a given component from other corresponding components without limiting the corresponding components in any other way (e.g., importance or order).

[0017] The terms “front surface”, “rear surface”, “upper surface”, “lower surface”, “side surface”, “left side”, “right side”, “upper part”, “lower part”, etc., used in this disclosure are defined with reference to the accompanying drawings, and the shape and position of each component are not limited by these terms.

[0018] The terms “comprising,” “having,” etc., are intended to indicate the presence of the features, quantities, steps, operations, components, parts, or combinations thereof described in this disclosure, and do not exclude the presence or addition of one or more other features, quantities, steps, operations, components, parts, or combinations thereof.

[0019] When any component is referred to as being “connected,” “joined,” “supported,” or “in contact” with another component, this includes situations where the components are indirectly connected, joined, supported, or in contact with each other through a third component, as well as situations where they are directly connected, joined, supported, or in contact with each other.

[0020] When any component is said to be "on" or "above" another component, this includes not only cases where any component is in contact with the other component, but also cases where the other component exists between the two components.

[0021] In the following, an ultrasonic device according to various embodiments will be described in detail with reference to the accompanying drawings. When describing with reference to the drawings, similar reference numerals may be assigned to the same or corresponding components, and redundant descriptions may be omitted.

[0022] In this disclosure, images may include medical images obtained by medical imaging devices such as magnetic resonance imaging (MRI) devices, computed tomography (CT) devices, ultrasound imaging devices, and X-ray imaging devices.

[0023] In this disclosure, the “object” being photographed may include a person, an animal, or a part thereof. For example, an object may include a part of a human body (organs, etc.) or a phantom.

[0024] Throughout this disclosure, "ultrasound image" means an image of an object that has been processed based on ultrasound signals emitted to and reflected from the object.

[0025] In the following description, embodiments will be described in detail with reference to the accompanying drawings.

[0026] Reference Figure 1 and Figure 2 The ultrasound imaging system 100 may include a probe 20 and an ultrasound diagnostic device 40.

[0027] The ultrasound diagnostic device 40 can be implemented not only as a cart-type device but also as a portable device. Portable ultrasound imaging devices may include, for example, smartphones, laptops, PDAs, tablet PCs, etc., including probes and applications, but are not limited to these.

[0028] The probe 20 may include a wired probe that is connected to the ultrasound diagnostic device 40 via a wire to communicate with the ultrasound diagnostic device 40 via a wire, a wireless probe that is wirelessly connected to the ultrasound diagnostic device 40 to communicate with the ultrasound diagnostic device 40 via a wire, and / or a hybrid probe that is connected to the ultrasound diagnostic device 40 via a wire or wirelessly to communicate with the ultrasound diagnostic device 40 via a wire or to communicate with the ultrasound diagnostic device 40 via a wire.

[0029] According to various embodiments, such as Figure 1 As shown, the ultrasound diagnostic device 40 may include an ultrasound transmitter / receiver module 110, or as... Figure 2 As shown, probe 20 may include an ultrasound transmitter / receiver module 110. According to various embodiments, both the ultrasound diagnostic device 40 and probe 20 may also include the ultrasound transmitter / receiver module 110.

[0030] According to various embodiments, the ultrasound diagnostic device 40 may further include an image processor 130, a display 140, and / or an input interface 170. Alternatively, the image processor 130, display 140, and / or input interface 170 may also be included in the probe 20.

[0031] Therefore, the description of the ultrasound transmitter / receiver module 110, image processor 130, display 140 and / or input interface 170 included in the ultrasound diagnostic device 40 can also be applied to the ultrasound transmitter / receiver module 110, image processor 130, display 140 and / or input interface 170 included in the probe 20.

[0032] Figure 1 A control block diagram of the ultrasound imaging system 100 is shown when probe 20 is a wired probe or a hybrid probe.

[0033] The probe 20 may include multiple transducers. These transducers can transmit ultrasound signals to the object 10 in response to a transmission signal applied from the transmission module 113. The multiple transducers can form a received signal by receiving ultrasound signals reflected from the object 10 (echo signals). The probe 20 may be implemented as an integral part of the ultrasound diagnostic device 40, or as a separate device connected to the ultrasound diagnostic device 40 via wires. The ultrasound diagnostic device 40 may be connected to one or more probes 20, depending on the implementation type.

[0034] In the case where probe 20 is a wired probe or a hybrid probe, probe 20 may include a cable and connector capable of being connected to a connector of ultrasound diagnostic equipment 40.

[0035] The probe 20 according to the embodiment can be implemented as a two-dimensional probe. When the probe 20 is implemented as a two-dimensional probe, a plurality of transducers included in the probe 20 can be arranged in a two-dimensional manner to form a two-dimensional transducer array.

[0036] For example, a two-dimensional transducer array may have the form in which multiple subarrays of multiple transducers arranged in a first direction are arranged in a second direction different from the first direction.

[0037] Additionally, when the probe 20 according to the embodiment is implemented as a two-dimensional probe, the ultrasonic transmitter / receiver module 110 may include an analog beamformer and a digital beamformer. Optionally, the two-dimensional probe may include one or both of an analog beamformer and a digital beamformer, depending on the implementation type.

[0038] The processor 120 controls the transmitting module 113 to form a transmitting signal to be applied to each of the multiple transducers included in the probe 20, taking into account the position and focus of the multiple transducers.

[0039] The processor 120 can control the receiving module 117 to generate ultrasound data by performing analog-to-digital conversion on the received signal received from the probe 20 and summing the digitally converted received signal, taking into account the position and focus of multiple transducers.

[0040] When probe 20 is implemented as a two-dimensional probe, processor 120 can calculate a time delay value for digital beamforming for each of the multiple subarrays included in the two-dimensional transducer array. Processor 120 can also calculate a time delay value for analog beamforming for each transducer included in one of the multiple subarrays. Processor 120 can control the analog beamformer and the digital beamformer to form a transmission signal to be applied to each of the multiple transducers based on the time delay values ​​for analog and digital beamforming. Processor 120 can also control the analog beamformer to sum the signals received from the multiple transducers for each subarray based on the time delay values ​​for analog beamforming. Processor 120 can also control the ultrasound transmitter / receiver module 110 to perform analog-to-digital conversion on the summed signal for each subarray. Processor 120 can also control the digital beamformer to generate ultrasound data by summing the digitally converted signal based on the time delay values ​​for digital beamforming.

[0041] The image processor 130 uses the generated ultrasound data to generate ultrasound images.

[0042] The display 140 can display the generated ultrasound images and various information processed by the ultrasound diagnostic device 40 and / or probe 20. Depending on the implementation type, the probe 20 and / or ultrasound diagnostic device 40 may include one or more displays 140. The display 140 may also include a touch panel or touch screen.

[0043] Display 140 can output four-dimensional ultrasound images according to control commands from processor 120. The four-dimensional ultrasound images can represent three-dimensional images provided in real time by adding a time dimension. For example, a four-dimensional ultrasound image can be an ultrasound image that includes fetal movement, heartbeat, or other movements of biological tissue over time. The four-dimensional ultrasound image can be based on ultrasound image data acquired in real time or ultrasound image data previously stored in memory 150.

[0044] The processor 120 controls the overall operation of the ultrasound diagnostic device 40 and the signal flow between its internal components. The processor 120 can execute or control various operations or functions of the ultrasound diagnostic device 40 by running programs or instructions stored in the memory 150. The processor 120 can also control the operation of the ultrasound diagnostic device 40 by receiving control signals from the input interface 170 or external devices.

[0045] The ultrasound diagnostic device 40 may include a communication module 160 and can be connected to external devices (e.g., probe 20, server, medical device, portable device (smartphone, tablet PC, wearable device, etc.)) via the communication module 160.

[0046] The communication module 160 may include one or more components that enable communication with external devices, and may include at least one of, for example, a short-range communication module, a wired communication module, and a wireless communication module.

[0047] The communication module 160 can receive control signals and data from external devices, and can send the received control signals to the processor 120 so that the processor 120 can control the ultrasound diagnostic device 40 in response to the received control signals.

[0048] Optionally, the processor 120 can send control signals to an external device via the communication module 160 to control the external device in response to the processor's control signals.

[0049] For example, an external device can process data in the external device in response to control signals received from a processor via a communication module.

[0050] The program that controls the ultrasound diagnostic device 40 can be installed in an external device, and the program may include instructions for performing some or all of the operations of the processor 120.

[0051] The program may be pre-installed on an external device, or the user of the external device may download and install the program from the server providing the application. The server providing the application may include a recording medium storing the program.

[0052] The memory 150 can store various data or programs used to drive and control the ultrasound diagnostic device 40, as well as input and output ultrasound data, ultrasound images, etc.

[0053] Input interface 170 can receive user input for controlling ultrasound diagnostic equipment 40. For example, user input may include, but is not limited to, input via operation buttons, keypad, mouse, trackball, micro switch, knob, etc., input via touchpad or touch screen, voice input, motion input, biometric information input (e.g., iris recognition, fingerprint recognition, etc.).

[0054] Figure 2 A control block diagram of an ultrasound imaging system 100 is shown when probe 20 is a wireless probe or a hybrid probe.

[0055] According to various embodiments, Figure 2 The ultrasound diagnostic device 40 shown can be used for reference. Figure 1 The described ultrasound diagnostic equipment 40 is replaced.

[0056] According to various embodiments, Figure 1 The probe 20 shown can be used as a reference. Figure 2 The described probe 20 is replaced.

[0057] The probe 20 may include a transmitting module 113, a battery 114, a transducer 115, a charging module 116, a receiving module 117, a processor 118, and a communication module 119. Although Figure 1 The probe 20 is shown to include both a transmitting module 113 and a receiving module 117; however, the probe 20 may include only a portion of the configuration of the transmitting module 113 and the receiving module 117, depending on the implementation type, and this portion of the configuration of the transmitting module 113 and the receiving module 117 may be included in the ultrasound diagnostic device 40. Optionally, the probe 20 may also include an image processor 130.

[0058] Transducer 115 may include multiple transducers. The multiple transducers may transmit ultrasonic signals to object 10 in response to a transmission signal applied from transmission module 113. The multiple transducers may receive ultrasonic signals reflected from object 10 to form a received signal.

[0059] The charging module 116 can charge the battery 114. The charging module 116 can receive power from an external source. The charging module 116 can receive power wirelessly. However, this disclosure is not limited to this, and the charging module 116 can also receive power via a wire. The charging module 116 can transfer the received power to the battery 114.

[0060] The processor 118 controls the transmitting module 113 to form a transmitting signal to be applied to each of the multiple transducers, taking into account the position and focus of the multiple transducers.

[0061] Processor 118 controls receiver module 117 to generate ultrasound data by performing analog-to-digital conversion on the received signals received from transducer 115, taking into account the positions and focal points of multiple transducers, and summing the digitally converted received signals. Optionally, if probe 20 includes image processor 130, probe 20 can use the generated ultrasound data to generate ultrasound images.

[0062] When probe 20 is implemented as a two-dimensional probe, processor 118 can calculate a time delay value for digital beamforming for each of the multiple subarrays included in the two-dimensional transducer array. Processor 118 can also calculate a time delay value for analog beamforming for each transducer included in one of the multiple subarrays. Processor 118 can control the analog beamformer and the digital beamformer to form a transmission signal to be applied to each of the multiple transducers based on the time delay values ​​for analog and digital beamforming. Processor 118 can also control the analog beamformer to sum the signals received from the multiple transducers for each subarray based on the time delay values ​​for analog beamforming. Processor 118 can also control the ultrasound transmitter / receiver module 110 to perform analog-to-digital conversion on the summed signal for each subarray. Processor 118 can also control the digital beamformer to generate ultrasound data by summing the digitally converted signal based on the time delay values ​​for digital beamforming.

[0063] The processor 118 can control the overall operation of the probe 20 and the signal flow between the internal components of the probe 20. The processor 118 can execute or control various operations or functions of the probe 20 by running programs or instructions stored in the memory 111. The processor 118 can also control the operation of the probe 20 by receiving control signals from the probe 20's input interface 170 or external devices (e.g., ultrasound diagnostic equipment 40).

[0064] The communication module 119 can wirelessly transmit the generated ultrasound data or ultrasound images to the ultrasound diagnostic device 40 via a wireless network. The communication module 119 can also receive control signals and data from the ultrasound diagnostic device 40.

[0065] The ultrasound diagnostic device 40 can receive ultrasound data or ultrasound images from the probe 20.

[0066] In an embodiment, where the probe 20 includes an image processor 130 capable of generating ultrasound images from ultrasound data, the probe 20 can send ultrasound data and / or ultrasound images generated by the image processor 130 to the ultrasound diagnostic device 40.

[0067] In an embodiment, where the probe 20 does not include an image processor 130 capable of generating ultrasound images from ultrasound data, the probe 20 may transmit ultrasound data to the ultrasound diagnostic device 40. The ultrasound data may include raw ultrasound data, and the ultrasound image may refer to ultrasound image data.

[0068] The ultrasound diagnostic device 40 may include a processor 120, an image processor 130, a display 140, a memory 150, a communication module 160, and an input interface 170.

[0069] The image processor 130 uses ultrasound data received from the probe 20 to generate an ultrasound image.

[0070] The display 140 can display ultrasound images received from the probe 20, ultrasound images generated by processing the ultrasound data received from the probe 20, and various information processed by the ultrasound imaging system 100. The ultrasound diagnostic device 40 may include one or more displays 140, depending on the implementation type. The display 140 may also include a touch panel or a touch screen.

[0071] The processor 120 controls the overall operation of the ultrasound diagnostic device 40 and the signal flow between its internal components. The processor 120 can execute or control various operations or functions of the ultrasound diagnostic device 40 by running programs or applications stored in the memory 150. The processor 120 can also control the operation of the ultrasound diagnostic device 40 by receiving control signals from the input interface 170 or external devices.

[0072] The ultrasound diagnostic device 40 may include a communication module 160 and can be connected to external devices (e.g., probe 20, server, medical device, portable device (smartphone, tablet PC, wearable device, etc.)) via the communication module 160.

[0073] The communication module 160 may include one or more components that enable communication with external devices, and may include at least one of, for example, a short-range communication module, a wired communication module, and a wireless communication module.

[0074] The communication module 160 of the ultrasound diagnostic device 40 and the communication module 119 of the probe 20 can communicate using a network or short-range wireless communication method. For example, the communication module 160 of the ultrasound diagnostic device 40 and the communication module 119 of the probe 20 can communicate using any of the following: wireless LAN, Wi-Fi, Bluetooth, ZigBee, Wi-Fi Direct (WFD), Infrared Data Association (IrDA), Bluetooth Low Energy (BLE), Near Field Communication (NFC), Wireless Broadband Internet (WiBro), Global Microwave Access Interoperability (WiMAX), Shared Wireless Access Protocol (SWAP), Wireless Gigabit Alliance (WiGig), RF communication, and wireless data communication methods including 60 GHz millimeter wave (mm wave) short-range communication, etc.

[0075] Therefore, the communication module 160 of the ultrasound diagnostic device 40 and the communication module 119 of the probe 20 may include at least one of the following modules: wireless LAN communication module, Wi-Fi communication module, Bluetooth communication module, ZigBee communication module, Wi-Fi Direct (WFD) communication module, Infrared Data Association (IrDA) communication module, Bluetooth Low Energy (BLE) communication module, Near Field Communication (NFC) module, Wireless Broadband Internet (WiBro) communication module, Global Microwave Access Interoperability (WiMAX) communication module, Shared Wireless Access Protocol (SWAP) communication module, Wireless Gigabit Alliance (WiGig) communication module, RF communication module, and 60 GHz millimeter wave (mm wave) short-range communication module.

[0076] In an embodiment, probe 20 may use a first communication method (e.g., BLE) to send device information (e.g., ID information) of probe 20, may wirelessly pair with ultrasound diagnostic device 40, and may send ultrasound data and / or ultrasound images to the paired ultrasound diagnostic device 40.

[0077] The device information of probe 20 may include various information related to the serial number, model name and battery status of probe 20.

[0078] The ultrasound diagnostic device 40 can receive device information (e.g., ID information) of the probe 20 from the probe 20 using a first communication method (e.g., BLE), can wirelessly pair with the probe 20, can send an activation signal to the paired probe 20, and can receive ultrasound data and / or ultrasound images from the probe 20. In this case, the activation signal may include a signal for controlling the operation of the probe 20.

[0079] In an embodiment, probe 20 can use a first communication method (e.g., BLE) to send device information (e.g., ID information) of probe 20, can wirelessly pair with ultrasound diagnostic device 40, and can use a second communication method (e.g., 60 GHz millimeter wave and Wi-Fi) to send ultrasound data and / or ultrasound images to ultrasound diagnostic device 40 paired via the first communication method.

[0080] The ultrasound diagnostic device 40 can receive device information (e.g., ID information) of the probe 20 from the probe 20 using a first communication method (e.g., BLE), can wirelessly pair with the probe 20, can send an activation signal to the paired probe 20, and can receive ultrasound data and / or ultrasound images from the probe 20 using a second communication method (e.g., 60 GHz millimeter wave and Wi-Fi).

[0081] According to various embodiments, a first communication method for pairing the probe 20 and the ultrasound diagnostic device 40 with each other may have a lower frequency band than a second communication method used by the probe 20 to transmit ultrasound data and / or ultrasound images to the ultrasound diagnostic device 40.

[0082] The display 140 of the ultrasound diagnostic device 40 can display a UI indicating device information of the probe 20. For example, the display 140 can display the UI indicating the identification information of the wireless probe 20, the pairing method indicating how to pair with the probe 20, the data communication status between the probe 20 and the ultrasound diagnostic device 40, the method of performing data communication with the ultrasound diagnostic device 40, and the battery status of the probe 20.

[0083] When the probe 20 includes a display 140, the display 140 of the probe 20 may display a UI indicating device information of the probe 20. For example, the display 140 may display a UI indicating identification information of the wireless probe 20, a pairing method indicating how to pair with the probe 20, the data communication status between the probe 20 and the ultrasound diagnostic device 40, the method of performing data communication with the ultrasound diagnostic device 40, and the battery status of the probe 20.

[0084] The communication module 160 can also receive control signals and data from external devices and send the received control signals to the processor 120, so that the processor 120 controls the ultrasound diagnostic device 40 in response to the received control signals.

[0085] Optionally, the processor 120 can send control signals to an external device via the communication module 160 to control the external device in response to the control signals of the processor 120.

[0086] For example, an external device can process data in response to control signals received from the processor 120 via the communication module.

[0087] The program that controls the ultrasound diagnostic device 40 can be installed in an external device, and the program may include instructions for performing some or all of the operations of the processor 120.

[0088] The program may be pre-installed on an external device, or the user of the external device may download and install the program from the server providing the application. The server providing the application may include a recording medium storing the program.

[0089] The memory 150 can store various data or programs used to drive and control the ultrasound diagnostic device 40, as well as input and output ultrasound data, ultrasound images, etc.

[0090] It will be passed later. Figure 3 , Figure 4 , Figure 5 and Figure 6 An example of an ultrasound imaging system 100 according to an embodiment of the present disclosure is described.

[0091] Figure 3 , Figure 4 , Figure 5 and Figure 6 This is a view showing an ultrasound imaging apparatus according to an embodiment.

[0092] Reference Figure 3 and Figure 4 The ultrasound imaging devices 40a and 40b may include a main display 121 (140) and a secondary display 122 (140). At least one of the main display 121 and the secondary display 122 may be implemented as a touchscreen. At least one of the main display 121 and the secondary display 122 may display various information or ultrasound images processed by the ultrasound imaging devices 40a and 40b. In addition, at least one of the main display 121 and the secondary display 122 may be implemented as a touchscreen and provide a GUI, allowing user input of data for controlling the ultrasound imaging devices 40a and 40b. For example, the main display 121 may display ultrasound images, and the secondary display 122 may display a control panel for controlling the display of ultrasound images in a GUI format (e.g., Figure 4 The secondary display 122 may provide data for controlling the display of images via a control panel displayed in a GUI format. For example, time gain compensation (TGC) buttons, freeze buttons, trackballs, microswitches, knobs, etc., may be provided as GUI settings on the secondary display 122.

[0093] Ultrasonic imaging devices 40a and 40b can use input control data to control the display of ultrasound images on the main display 121. Ultrasonic imaging devices 40a and 40b can also be connected to probe 20 via wired or wireless means to send and receive ultrasound signals to and from the object.

[0094] Reference Figure 4 In addition to the main display 121 and the secondary display 122, the ultrasound imaging device 40b may also include a control panel 165. The control panel 165 may include buttons, a trackball, microswitches, knobs, etc., and may be provided by the user with data for controlling the ultrasound imaging device 40b. For example, the control panel 165 may include a Time Gain Compensation (TGC) button 171, a freeze button 172, etc. The TGC button 171 is used to set the TGC value for each depth of the ultrasound image. When the input of the freeze button 172 is sensed while scanning the ultrasound image, the ultrasound imaging device 40b may maintain the display of the frame image at that time point.

[0095] Buttons, trackballs, microswitches, knobs, etc., included in the control panel 165 can be configured as a GUI on the main display 121 or the secondary display 122. Ultrasonic imaging devices 40a and 40b can be connected to probe 20 to send and receive ultrasonic signals to and from the object.

[0096] Reference Figure 5 and Figure 6 Ultrasound diagnostic devices 40c and 40d can be implemented in a portable form. Portable ultrasound imaging devices 40c and 40d may include, for example, smartphones, laptops, PDAs, tablet PCs, etc., including probes and applications, but are not limited to these.

[0097] The ultrasound imaging device 40c may include a main body 41. (Refer to...) Figure 5 The probe 20 can be connected to one side of the body 41 via a wire. For this purpose, the body 41 may include a connection terminal to which the cable connected to the probe 20 can be attached and detached, and the probe 20 may include a connection terminal to which the cable connected to the body 41 can be attached and detached.

[0098] Reference Figure 6 The probe 20 can be wirelessly connected to the ultrasound diagnostic device 40d. The main body 41 may include an input / output interface (e.g., a touch screen) 145 (display 140 and input interface 170). Ultrasound images, various information processed by the ultrasound imaging device, GUI, etc., can be displayed on the input / output interface 145.

[0099] Ultrasound images can be displayed on input / output interface 145. Ultrasound imaging devices 40c and 40d can use AI to correct the ultrasound images displayed on input / output interface 145. Ultrasound imaging devices 40c and 40d can provide alarms to notify users of information about lesions in the ultrasound images displayed on input / output interface 145 using various audiovisual tools such as graphics, sound, and vibration.

[0100] The ultrasound imaging devices 40c and 40d can output a control panel displayed in GUI form via input / output interface 145.

[0101] The ultrasound imaging device 40d and the probe 20 can establish communication or pairing using short-range wireless communication. For example, the ultrasound imaging device 40d and the probe 20 can communicate using Bluetooth, BLE, Wi-Fi, or Wi-Fi Direct.

[0102] Ultrasonic imaging devices 40c and 40d can run programs or applications related to probe 20 to control probe 20 and output information related to probe 20. Ultrasonic imaging devices 40c and 40d can perform operations related to probe 20 while communicating with a predetermined server. Probe 20 can register with ultrasonic imaging devices 40c and 40d, or it can register with a predetermined server. Ultrasonic imaging devices 40c and 40d can communicate with registered probe 20 and perform operations related to probe 20.

[0103] Ultrasonic imaging devices 40c and 40d may include various types of input / output interfaces, such as speakers, LEDs, and vibration devices. For example, ultrasonic imaging devices 40c and 40d can output various information in the form of graphics, sound, or vibration through the input / output interfaces. Ultrasonic imaging devices 40c and 40d can also output various notifications or data through the input / output interfaces.

[0104] According to embodiments of this disclosure, ultrasound imaging devices 40a, 40b, 40c, or 40d may use artificial intelligence (AI) models to process ultrasound images or obtain additional information from ultrasound images. According to embodiments of this disclosure, ultrasound imaging devices 40a, 40b, 40c, or 40d may use AI models to generate ultrasound images or perform processes such as correction, image quality improvement, encoding, and decoding on ultrasound images. Furthermore, according to embodiments of this disclosure, ultrasound imaging devices 40a, 40b, 40c, or 40d may use AI models to perform processes such as baseline definition, anatomical information acquisition, lesion information acquisition, surface extraction, boundary definition, length measurement, area measurement, volume measurement, and annotation creation on ultrasound images.

[0105] The AI ​​model can be set on ultrasound imaging devices 40a, 40b, 40c or 40d, or it can be set on a server.

[0106] AI models can be implemented using various artificial neural network models or deep neural network models. Additionally, various machine learning algorithms or deep learning algorithms can be used to learn (train) and create AI models. AI models can be implemented using models such as convolutional neural networks (CNNs), recurrent neural networks (RNNs), generative adversarial networks (GANs), and long short-term memory (LSTM) networks.

[0107] Figure 7 A method for manually measuring the diameter ratio of the left atrium to the aorta is shown in the embodiment.

[0108] According to various embodiments, a user can use buttons, a keypad, a mouse, a trackball, etc., of the input interface 170 to specify two points on an ultrasound cross-sectional image and measure the distance between the two points. For example, a user can use a caliper function to obtain length information (e.g., length, area, or width) of structures included in an animal echocardiographic cross-sectional image.

[0109] Specifically, users can view cross-sectional images of animal echocardiography (e.g., Figure 7 Marking is performed on D1 in the input interface 170. In other words, the user can use buttons, keypad, mouse, trackball, etc. of the input interface 170 to place the mark on the animal echocardiogram cross-sectional image. In this case, marking may involve selecting and marking specific points on the animal echocardiogram cross-sectional image. When the user inputs a command about marking through the input interface 170, the processor 120, which has received the command, can display at least one indicator related to the marking (e.g., ...) at the specific point on the animal echocardiogram cross-sectional image based on the user's command. Figure 7 (d1, d2, d3, or d4 in the text).

[0110] Therefore, processor 120 can measure the distance between two points specified by the user. For example, when the user specifies two points d1 and d2 to measure the diameter of the left atrium (LA), processor 120 can measure the distance between the two points d1 and d2. Similarly, when the user specifies two points d3 and d4 to measure the diameter of the aorta (AO), processor 120 can measure the distance between the two points d3 and d4. Therefore, processor 120 can measure the ratio of the diameter of the left atrium to the diameter of the aorta. In this case, processor 120 can display a line connecting the two points used as a reference for measuring each diameter on the animal echocardiographic cross-sectional image (e.g., Figure 7 (C1 and C2 in the text) and at least one indicator associated with a specific point marker (e.g., Figure 7 (d1, d2, d3, or d4 in the text).

[0111] However, when manually measuring the length information of a structure included in a cross-sectional image using a cross-sectional image, errors may occur even when measuring the same structure included in the same cross-sectional image, because the positions of the two points used as the basis for the length measurement are determined based on the user's experience or subjective judgment.

[0112] Figure 8 An example of input data to a machine learning model, according to an embodiment, is shown.

[0113] According to an embodiment, the processor 120 can acquire animal ultrasound images and extract cross-sectional images included in the animal ultrasound images.

[0114] A cross-sectional image can be obtained from one of the image frames included in the acquired ultrasound image. In this case, the ultrasound image can be a real-time ultrasound image obtained from the echo signal received by the probe 20 when the ultrasound signal emitted from the probe 20 to the object is reflected and returns as an echo signal. Alternatively, the ultrasound image can be an ultrasound image stored in the memory 150.

[0115] According to an embodiment, the ultrasound image obtained by the processor 120 may include a parasternal short-axis (PSAX) view image obtained by placing the probe 20 next to the sternum of an animal and emitting ultrasound waves in the transverse direction of the heart. Therefore, the cross-sectional image obtained from the ultrasound image may include a cross-sectional image based on the short axis of the animal's heart (e.g., Figure 8 (D21 in the middle).

[0116] Cross-sectional images obtained from parasternal short-axis (PSAX) views show a cross-section of the heart and allow for observation of several cardiac structures at once. Therefore, because several important cardiac structures (such as the left ventricle, aorta, and left atrium) can be identified, cross-sectional images can be useful for assessing the condition of heart valves or myocardium.

[0117] In addition, cross-sectional images can be used to observe the heart's contraction and relaxation states as well as blood flow.

[0118] Users can use cross-sectional images obtained from parasternal short axis (PSAX) view images to measure the length, area, or width of the left atrium or aorta included in the cross-sectional image.

[0119] In addition, when diagnosing cardiac function and valvular abnormalities, users can use cross-sectional images obtained from parasternal short axis (PSAX) view images to measure the diameter ratio of the left atrium to the aorta.

[0120] Figure 9 An example of output data from a machine learning model, according to an embodiment, is shown.

[0121] According to an embodiment, the processor 120 can input the cross-sectional image as input data into the AI ​​model and obtain the processed image (including the outline drawing image of multiple target structures included in the cross-sectional image) as output data.

[0122] In this context, the target structure may include a structure that serves as the object of measurement for obtaining length information. For example, when obtaining information about the diameter of the left atrium, the target structure may be a cross-section of the left atrium. Similarly, when obtaining information about the diameter of the aorta, the target structure may be a cross-section of the aorta. As another example, when obtaining information about the ratio of the diameters of the left atrium to the aorta, the target structure may be a cross-section of both the left atrium and the aorta. The target structure can be determined based on user input received from the input interface 170. For example, the user may directly set the target structure (e.g., organ, lesion, tumor, etc.) through the input interface 170. Additionally, the processor 120 may determine a preset target structure as the target structure based on the mode or application selected by the user through the input interface 170.

[0123] Contour rendering images are used to trace the boundaries of target structures (e.g., organs, lesions, tumors, etc.) and visually convey information about the boundaries of the target structure.

[0124] For example, refer to Figure 9 Based on the contents of the left atrial and aortic sections identified as target structures, the processed images may include a contour-mapped image CT1 of the left atrial section and a contour-mapped image CT2 of the aortic section. Although Figure 9 The processed image only shows the outline drawing image without any indication of the cross-sectional image used as input data, but the processed image can be implemented by overlaying the outline drawing image with the cross-sectional image used as input data.

[0125] According to an embodiment, the processor 120 can display various information about the target structure and a contour drawing image of the target structure on a processed image obtained from a machine learning model.

[0126] For example, such as Figure 9 As shown, the processor 120 can also calculate the centroid of each contour drawing image and display an indicator indicating the centroid (e.g., Figure 9 (CG1 and CG2 in the image). In this case, a known image processing method can be used as the method by which processor 120 calculates the centroid of each contour drawing image. As an example, processor 120 can binarize the contour drawing image and calculate the coordinate values ​​of all pixels belonging to the binarized contour drawing image by taking a weighted average. That is, processor 120 can calculate the centroid by adding the product of the coordinates of each pixel with the pixel values ​​of each in the x-axis and y-axis and dividing it by the total number of pixels in the structure.

[0127] In the following text, reference will be made to Figure 10 The method for obtaining processed data by using a machine learning model that takes a cross-sectional image obtained by processor 120 as input data is described in detail.

[0128] Figure 10 This is an illustration of a method for obtaining a processed image, including a contour drawing image of multiple target structures, using an AI model according to an embodiment.

[0129] The AI ​​model according to the embodiment may include an encoder 800 and a decoder 900.

[0130] According to an embodiment, processor 120 can input input data (e.g., D21) to encoder 800 to extract features from a cross-sectional image. In other words, when processor 120 inputs input data to encoder 800, a feature map M4 can be created that includes features extracted from the cross-sectional image as input data. In other words, processor 120 can utilize an AI model that uses a cross-sectional image obtained from ultrasound images as input data to perform a segmentation task.

[0131] The encoder 800 may include a CNN backend 810, atrous convolution 820, atrous spatial pyramid pooling (ASPP) 830, and 1×1 convolution 840.

[0132] The CNN backend 810 can perform multi-layer convolutional operations to transform the input image into abstract features. That is, low-level features can be extracted from the initial layers of the CNN backend 810, and these low-level features can contain much detailed boundary or texture information of the input image. Therefore, these low-level features can be used to recover important details in the decoder 900.

[0133] Dilated convolution 820 is similar to basic convolution, but it expands the receiving region through each filter by leaving gaps between the filters in the kernel. Therefore, information can be obtained from a large area without reducing image resolution. Dilated convolution 820 can be included in the CNN backend 810. In other words, dilated convolution 820 can be used in specific layers of the CNN backend 810.

[0134] For example, while general convolutions extract features from small regions, dilated convolutions (820) can effectively handle contextual information in large regions by increasing the spacing between pixels. Therefore, important object boundaries and contextual information can be obtained simultaneously in image segmentation tasks.

[0135] The ASPP 830 can use dilated convolutional filters of different sizes to train object boundaries and contextual information. The ASPP 830 can obtain information at various scales of an image through filters of various sizes and dilation ratios. For example, the ASPP 830 can include a 1×1 convolution 831, a 3×3 dilated convolution 832 with a dilation ratio of 6, a 3×3 dilated convolution 833 with a dilation ratio of 12, and a 3×3 dilated convolution 834 with a dilation ratio of 18. The ASPP 830 can use small filters to extract detailed features from small regions and large filters to extract abstract and broad features from large regions.

[0136] The ASPP 830 can also add global context information for the entire image through the operations of the image pooling 835.

[0137] ASPP 830 can create combined feature maps by combining these multi-scale feature maps M1, M21, M22, M23 and M3.

[0138] A 1×1 convolution 840 can reduce the complex information from combining feature maps from multiple channels into a single channel. Therefore, the 1×1 convolution 840 improves computational efficiency and creates a more concise representation before passing complex information to a multi-scale information decoder.

[0139] A 1×1 convolution 840 can compress and summarize information obtained from multiple scales by combining all feature map channels at each location. The feature map M4 created in this way can be fed to the decoder 900 and detailed segmentation can be performed.

[0140] Decoder 900 can perform detailed image segmentation by converting low-resolution feature maps extracted from encoder 800 into high-resolution maps.

[0141] Decoder 900 can combine low-level features extracted from the initial layer of encoder 800 into a high-level feature map M4. Because low-level features contain detailed information such as the boundaries or texture information of the cross-sectional image input as input data, the boundaries of the segmentation can be recovered more accurately by combining low-level features.

[0142] However, because low-level features contain a lot of unnecessary information, decoder 900 can refine the information or remove unimportant channels through a 1×1 convolution operation. Therefore, decoder 900 can extract a refined low-level feature map M5.

[0143] The decoder 900 can perform upsampling 920 and 950, which progressively enlarge the low-resolution feature map M4. Therefore, a processed image of the same size as the cross-sectional image input as input data can be generated as output data.

[0144] For example, decoder 900 can convert low-resolution features to medium-resolution features by upsampling feature map M4 via first upsampling 920.

[0145] In 930, decoder 900 can create a combined feature map by combining low-level features (a low-level feature map M5 in a refined form obtained through the operation of 1×1 convolution 910) and a medium-resolution feature map that has already been first upsampled.

[0146] Subsequently, the decoder 900 can supplement the image with detailed information through the operation of a 3×3 convolution 940.

[0147] The decoder 900 can obtain a processed image (e.g., D22) with the same resolution as the cross-sectional image input as input data via a second upsampling 950.

[0148] In other words, the decoder 900 can generate a high-resolution processed image that matches the size of the cross-sectional image that is input as input data, based on the low-level and high-level features extracted from the encoder 800.

[0149] Figure 11 An example of training data used to train a machine learning model is shown.

[0150] Figure 12 Another example of training data used to train a machine learning model is shown.

[0151] According to an embodiment, the processor 120 can train an AI model through supervised learning. In other words, the processor 120 can use input data and corresponding correct answer data to train the AI ​​model to generate the desired output data. In this case, the correct answer data can be referred to as labeled data.

[0152] In this scenario, the processor 120 can prepare input data and correct answer data, measure the loss function (which is the difference between the AI ​​model's output data and the correct answer data) during AI model training, and repeatedly adjust the AI ​​model's parameters based on the measured loss function multiple times (multiple cycles). Therefore, the error between the AI ​​model's output data and the correct answer data can be reduced, and accurate output data can be output. During this process, the processor 120 can use test data to evaluate the AI ​​model's learning level. In this case, the generalization ability of the AI ​​model can be checked by using data not input during the AI ​​model's training (i.e., third-party data) as test data.

[0153] In AI model training, correct answer data is crucial for improving the model's generalization ability. Correct answer data serves as the foundation for AI model training, and it's essential to ensure that this data is accurate and diverse so that the AI ​​model can output accurate results in various situations.

[0154] In particular, when performing segmentation tasks, the input data must be accurately labeled in the correct class (e.g., person, background). When the correct answer data is incomplete or contains incorrect labels, the AI ​​model may be unable to learn pixel by pixel.

[0155] In this case, labeling the input data may include: labeling the boundaries of the structure to obtain a contour drawing image of the structure in the cross-sectional image.

[0156] For example, labeling the input data may include: having a veterinarian mark the boundaries of the left atrial section and the aortic section to obtain a contour drawing image of the left atrial section and the aortic section identified as the target structures in the cross-sectional image.

[0157] In an embodiment, a cross-sectional image obtained from an ultrasound image (e.g., Figure 8 D21 in the image can be used as input data for training an AI model. In this case, the cross-sectional image may include the target structure.

[0158] Reference Figure 11 As correct answer data for training the AI ​​model, it can include in the cross-sectional image D31 an image L on which the boundaries of each target structure that can be examined in the cross-sectional image are marked. 11 and L 12 In other words, the data used by veterinarians to mark boundaries by reflecting the inspectable portions in cross-sectional images as accurately as possible can be used as training data.

[0159] Additionally, refer to Figure 12 During AI model training, the image L indicates the boundary of the corresponding target structure (where the feature factors of the target structure are reflected). 21 and L 22 This data can also be included in the cross-sectional image D32 as correct answer data. That is, the data in which the veterinarian marks the boundaries by considering the original form of the target structure and even parts that cannot be detected in the cross-sectional image can be used as additional training data.

[0160] According to an embodiment, when training an AI model, more generalized output data can be obtained from the AI ​​model by using correct answer data that utilizes two standard labels as correct answer data.

[0161] Figure 13 and Figure 14This is an illustration of an example of a standardized method for obtaining length information of multiple target structures from a processed image.

[0162] According to an embodiment, the processor 120 can obtain the length information of the target structure based on the intersection point between at least one line passing through the centroid of the contour drawing image and the boundary of the contour drawing image.

[0163] In this case, obtaining the length information of the target structure may include obtaining the diameter, width, area, or diameter ratio of each target structure in the target structure.

[0164] For example, obtaining length information of the target structure may include obtaining the diameter of the left atrium and aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta.

[0165] According to an embodiment, the processor 120 can also obtain length information of the target structure based on lines passing through all centroids of the contour drawing image.

[0166] Reference Figure 13 and Figure 14 The processor 120 can draw a line L of the centroids CG1 and CG2 of the cross-sectional image D23 based on the corresponding contours passing through the left atrium and aorta. CG The intersection point d between the contour drawing image and the boundary of the image. 1' d 2' d 3' and d 4' To calculate the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta.

[0167] Specifically, when calculating the diameter of the left atrium, processor 120 can use d 1' and d 2' d is calculated using a reference point. 1' and d 2' The distance between them, d 1' and d 2' It is a line L that passes through both the centers of gravity CG1 and CG2. CG The intersection point between the image and the boundary of the left atrial contour drawing. The processor 120 can also use d when calculating the diameter of the aorta. 3' and d 4' d is calculated using a reference point. 3' and d 4' The distance between them, d 3' and d 4' It is a line L that passes through both the centers of gravity CG1 and CG2. CG The intersection point between the boundary of the aortic contour drawing image and the boundary of the image. Subsequently, the processor 120 can calculate d when calculating the diameter ratio of the left atrium to the aorta. 1' and d 2'The distance between and d 3' and d 4' The ratio of the distances between them.

[0168] According to an embodiment, by using a standardized method based on the centroid of the connected contour drawing image and the boundary of the contour drawing image to extract the length information of the target structure, factors that cause errors between veterinarians and between measurement attempts can be eliminated.

[0169] Figure 15 This is an illustration of another example of a standardized method for obtaining length information of multiple target structures from a processed image.

[0170] According to an embodiment, the processor 120 can obtain the length information of the target structure by using multiple lines passing through the centroids of the contour drawing image, rather than by using lines passing through all the centroids of the contour drawing image.

[0171] Reference Figure 15 The processor 120 can draw the first line L, the centroid of the contour drawing image CT1, through the first target structure on the cross-sectional image D4. CG1 And draw the second line L of the centroid of the image CT2, which is the outline of the second target. CG2 With the first line L CG1 It has a preset angle θ.

[0172] Therefore, processor 120 can be based on the first line L CG1 The length information of the first target structure is obtained by finding the intersection points between the boundary of the contour rendering image CT1 and the first target structure. The processor 120 can also obtain the length information of the first target structure based on the second line L. CG2 The length information of the second target structure is obtained by finding the intersection point between the boundary of the contour drawing image CT2 and the second target structure.

[0173] Processor 120 can determine the first line L based on user input received through input interface 170. CG1 Second line L CG2 A preset angle θ is defined between the two angles. For example, processor 120 may receive the magnitude of the preset angle θ in advance via input interface 170. As another example, when processor 120 receives input related to a mode or application from a user via input interface 170, processor 120 may determine the magnitude of the preset angle based on preset data stored in memory 150.

[0174] According to the embodiment, the length information of the target structure is obtained by drawing multiple lines passing through the center of gravity, and appropriate length information can be obtained according to the characteristics or diagnostic conditions of each structure.

[0175] Figure 16This is an illustration of an image displayed on the screen of an ultrasound diagnostic device, showing an example of a standardized method for obtaining length information of multiple target structures from a processed image.

[0176] According to an embodiment, the processor 120 can display the obtained length information together with the cross-sectional image. In this case, the cross-sectional image may include a cross-sectional image obtained from an ultrasound image and input as input data to the AI ​​model.

[0177] Reference Figure 16 The processor 120 can control the display to show the diameter of the left atrium section, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta along with the cross-sectional image.

[0178] In this case, the processor 120 may display the diameter of the left atrial section, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta as annotation 10b on one side of the cross-sectional image. In this disclosure, the annotation regarding the diameter of the left atrial section, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta is displayed in the upper left corner of the screen of the display 140, but this is only an example and its position may be changed as long as it does not cover the cross-sectional image.

[0179] The processor 120 may also display the diameter of the left atrial section, the diameter of the aorta, or the ratio of the diameters of the left atrium and the aorta together with icon indicators and numerical values. For example, the icon indicator 10a displayed using numerical values ​​may include: a first indicator indicating the intersection point between a line passing through the centroid of the respective contour drawing image of the left atrium and the aorta and the respective boundary of the contour drawing image; and a second indicator indicating a connecting line connecting the intersection points in the respective contour drawing images.

[0180] Figure 17 This is a control flowchart of an ultrasound diagnostic device according to an embodiment.

[0181] According to an embodiment, processor 120 can acquire an animal ultrasound image (1001). In this case, the animal ultrasound image may include a real-time ultrasound image based on an echo signal returned by reflecting an ultrasound signal emitted from probe 20 to the object. The animal ultrasound image may also be an ultrasound image stored in memory 150.

[0182] Processor 120 can extract a cross-sectional image (1002) including a cross-section of a target structure from an animal ultrasound image. In this case, the target structure may include a structure that is measured as a target for obtaining length information. For example, when obtaining information about the diameter of the left atrium, the target structure may be a cross-section of the left atrium. Furthermore, when obtaining information about the diameter of the aorta, the target structure may be a cross-section of the aorta. As another example, when obtaining information about the ratio of the diameters of the left atrium to the aorta, the target structure may be a cross-section of both the left atrium and the aorta. The target structure may be determined based on user input received from input interface 170. For example, the user may directly set the target structure (e.g., organ, lesion, tumor, etc.) through input interface 170. Processor 120 may also determine a preset target structure as the target structure based on a mode or application selected by the user through input interface 170.

[0183] The processor 120 can obtain a processed image (1003) from the extracted cross-sectional image, including a contour drawing image of the target structure. Specifically, the processor 120 can input the extracted cross-sectional image as input data into the AI ​​model and obtain the processed image as output data. The AI ​​model can be stored in the memory 150.

[0184] In this context, when training an AI model to obtain processed images, a machine learning model can be learned by using cross-sectional images of sections including multiple target structures as input data and images indicating the boundaries of the corresponding target structures that can be checked in the cross-sectional images of sections including multiple target structures as training data.

[0185] Additionally, a machine learning model can be learned by adding images indicating the boundaries of the corresponding target structures (images reflecting the feature factors of the target structures) as training data to cross-sectional images that include multiple target structures. Therefore, a more generalized output data can be obtained by using correct answer data, which utilizes two standard labels, as correct answer data to learn the AI ​​model.

[0186] Processor 120 can obtain the centroid (1004) of each contour-drawn image included in the obtained processed image. In this case, a known image processing method can be used as the method by which processor 120 calculates the centroid of each contour-drawn image. As an example, processor 120 can binarize the contour-drawn image and calculate the coordinate values ​​of all pixels belonging to the binarized contour-drawn image by taking a weighted average. That is, processor 120 can calculate the centroid by adding the product of the coordinates of each pixel with the pixel values ​​of each in the x-axis and y-axis and dividing it by the total number of pixels in the structure.

[0187] Therefore, processor 120 can obtain length information of the target structure based on the intersection point between at least one line passing through the centroid of the contour-drawn image and the boundary of the contour-drawn image (1005). In this case, the length information may include length, diameter, width, or diameter ratio. For example, processor 120 can calculate the diameter of the left atrium, the diameter of the aorta, or the diameter ratio of the left atrium to the aorta based on the intersection point between a line passing through the centroid of the respective contour-drawn images of the left atrium and the aorta and the boundary of the contour-drawn image.

[0188] Specifically, processor 120 can use the two intersection points between a line passing through all centroids and the boundary of the left atrial contour image as reference points to calculate the diameter of the left atrium. Processor 120 can also use the two intersection points between a line passing through all centroids and the boundary of the aortic contour image as reference points to calculate the diameter of the aorta. Subsequently, processor 120 can calculate the diameter ratio of the left atrium to the aorta.

[0189] The control method of the ultrasound diagnostic device according to the embodiment may include: acquiring an animal ultrasound image; extracting a cross-sectional image from the animal ultrasound image that includes a cross section of a plurality of target structures; obtaining a processed image from the cross-sectional image that includes a contour drawing image of a plurality of target structures; obtaining the centroid of each contour drawing image in the contour drawing image; and obtaining length information of the target structure based on the intersection point between at least one line passing through the centroid of the contour drawing image and the boundary of the contour drawing image.

[0190] The target structures may include the left atrium and the aorta.

[0191] Obtaining length information of the target structure may include: obtaining the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta.

[0192] Obtaining the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta may include: calculating the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta based on the intersection points between a line through all centroids of the corresponding contour drawing images of the left atrium and the aorta and the boundaries of the corresponding contour drawing images.

[0193] Control methods for ultrasound diagnostic equipment may also include displaying the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta together with cross-sectional images.

[0194] Displaying the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the aorta along with the cross-sectional image may include displaying the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the aorta as a note on one side of the cross-sectional image.

[0195] Displaying the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta along with a cross-sectional image may include: displaying icon indicators and numerical values ​​for the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta.

[0196] In this case, the icon indicator may include: a first indicator indicating the intersection of a line through which all centroids of the corresponding contour drawing image pass through the left atrium and the aorta with the boundary of the corresponding contour drawing image; and a second indicator indicating a connecting line connecting the intersections in the corresponding contour drawing image.

[0197] The processed image obtained from the cross-sectional image, which includes contour rendering images of multiple target structures, may include: inputting the cross-sectional image of the cross section including multiple target structures obtained from the animal ultrasound image as input data into the AI ​​model stored in memory, and obtaining the processed image including contour rendering images of multiple target structures as output data.

[0198] The control method for the ultrasound diagnostic device may further include: training an AI model by using a cross-sectional image of a cross section including multiple target structures as input data and using an image indicating the boundary of the corresponding target structure that can be examined in the cross-sectional image of the cross section including multiple target structures as training data.

[0199] The control method for the ultrasound diagnostic equipment may also include: training an AI model by adding an image indicating the boundary of the corresponding target structure (which reflects the characteristic factors of the target structure) as training data to a cross-sectional image including multiple target structures.

[0200] An ultrasound diagnostic device according to an embodiment may include: a probe configured to send an ultrasound signal to an object and receive an echo signal reflected from the object; a display configured to display an ultrasound image obtained based on the echo information received by the probe; an input interface configured to receive user input; and a processor configured to control the probe to obtain an animal ultrasound image, extract a cross-sectional image from the animal ultrasound image including a cross-section of a plurality of target structures, obtain a processed image from the cross-sectional image including contour-drawn images of a plurality of target structures, obtain the centroid of each contour-drawn image in the contour-drawn images, and obtain length information of the target structures based on the intersection of at least one line passing through the centroid of the contour-drawn image with the boundary of the contour-drawn image.

[0201] The target structures may include the left atrium and the aorta.

[0202] The length information of the target structure may include the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta.

[0203] The processor can be configured to calculate the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameters of the left atrium to the aorta based on the intersection of a line through which all centroids of the corresponding contour-drawn image pass through the left atrium and the aorta with the boundary of the corresponding contour-drawn image.

[0204] The processor can be configured to control the display to show the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta, along with a cross-sectional image.

[0205] The processor can be configured to control the display to show the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta as annotations on one side of the cross-sectional image.

[0206] The processor can be configured to control the display to show icon indicators and numerical values ​​for the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta.

[0207] In this case, the icon indicator may include: a first indicator indicating the intersection of a line through which all centroids of the corresponding contour drawing image pass through the left atrium and the aorta with the boundary of the corresponding contour drawing image; and a second indicator indicating a connecting line connecting the intersections in the corresponding contour drawing image.

[0208] An ultrasound diagnostic device may include a memory configured to store an AI model, wherein a processor may be configured to: input a cross-sectional image of a cross section including multiple target structures obtained from an animal ultrasound image as input data to the AI ​​model stored in the memory, and obtain a processed image including contour drawings of multiple target structures as output data.

[0209] The processor can be configured to learn an AI model by using cross-sectional images of cross sections including multiple target structures as input data and images indicating the boundaries of the corresponding target structures that can be examined in the cross-sectional images of cross sections including multiple target structures as training data.

[0210] The processor can be configured to learn an AI model by adding images indicating the boundaries of the corresponding target structures (which reflect the feature factors of the target structures) as training data to a cross-sectional image that includes cross-sections of multiple target structures.

[0211] According to one aspect of this disclosure, factors that cause errors between veterinarians and between measurement attempts can be eliminated by using standardized methods to measure the ratio of the diameter of the left atrium to the aorta.

[0212] According to one aspect of this disclosure, the accuracy of diagnosis can be improved.

[0213] However, the effects achievable by the ultrasound diagnostic device and control method of this disclosure are not limited to those mentioned above, and those skilled in the art to which this disclosure pertains will clearly understand from the above description other effects not mentioned. The disclosed embodiments can be implemented in the form of a recording medium storing computer-executable instructions. The instructions can be stored in the form of program code, and when run by a processor, program modules can be created to perform the operations of the disclosed embodiments. The recording medium can be implemented as a computer-readable recording medium.

[0214] Computer-readable recording media include any type of recording medium in which computer-readable instructions are stored. For example, recording media may include read-only memory (ROM), random access memory (RAM), magnetic tape, magnetic disk, flash memory, optical data storage devices, etc.

[0215] Computer-readable recording media may be provided in the form of non-transitory storage media. Here, "non-transitory storage media" simply means that it is a tangible device and does not contain signals (e.g., electromagnetic waves), and the term does not distinguish between cases where data is stored semi-permanently in a storage medium and cases where data is temporarily stored. For example, a "non-transitory storage medium" may include a buffer for temporarily storing data.

[0216] According to embodiments, the methods according to the various embodiments disclosed in this document may be included and provided in a computer program product. The computer program product is a commodity and can be traded between a seller and a buyer. The computer program product may be distributed in the form of a machine-readable recording medium (e.g., an optical disc read-only memory (CD-ROM)) or distributed online (e.g., downloaded or uploaded) between two user devices (e.g., smartphones) via an app store (e.g., the Play Store™). In the case of online distribution, at least a portion of the computer program product (e.g., a downloadable app) may be at least temporarily stored or temporarily created in a machine-readable recording medium (such as the memory of a manufacturer's server, app store server, and relay server).

[0217] Specific embodiments have been described and illustrated above. However, those skilled in the art will understand that this disclosure is not limited to the above embodiments, and various changes and modifications can be made without departing from the technical concept of this disclosure as described in the claims.

Claims

1. A control method for an ultrasound diagnostic device, comprising: Obtaining ultrasound images of animals; Extract cross-sectional images of cross sections including multiple target structures from the animal ultrasound images; A processed image is obtained from the cross-sectional image, the processed image including contour rendering images of the plurality of target structures; Obtain the centroid of each element in the contour drawing image; as well as The length information of the target structure is obtained based on the intersection point between at least one line passing through the centroid of the contour drawing image and the boundary of the contour drawing image.

2. The control method as described in claim 1, wherein, The target structures include the left atrium and the aorta.

3. The control method as described in claim 2, wherein, Obtaining the length information of the target structure includes: obtaining the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta.

4. The control method as described in claim 3, wherein, Obtaining the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta includes: calculating the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta based on the intersection points between a line passing through all centroids of the corresponding contour drawing images of the left atrium and the aorta and the boundary of the corresponding contour drawing images.

5. The control method as described in claim 4, further comprising: The diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta are displayed together with the cross-sectional image.

6. The control method as described in claim 5, wherein, The step of displaying the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta together with the cross-sectional image includes displaying the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta as annotations on one side of the cross-sectional image.

7. The control method as described in claim 5, wherein, The step of displaying the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta along with the cross-sectional image includes: displaying an icon indicator and the numerical values ​​of the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta, and... The icon indicator includes a first indicator and a second indicator. The first indicator indicates the intersection point between a line passing through all centroids of the corresponding contour drawing image of the left atrium and the aorta and the boundary of the corresponding contour drawing image. The second indicator indicates a connecting line connecting the intersection points in the corresponding contour drawing image.

8. The control method as described in claim 1, wherein, The process of obtaining the processed image from the cross-sectional image includes: inputting a cross-sectional image of the cross-section including the plurality of target structures obtained from the animal ultrasound image as input data into an AI model stored in a memory; and obtaining a processed image including the contour drawing image of the plurality of target structures as output data.

9. The control method as described in claim 8, further comprising: The AI ​​model is trained by using cross-sectional images of the cross-sections including the plurality of target structures as input data, and by using images indicating the boundaries of the corresponding target structures that can be examined in the cross-sectional images including the plurality of target structures as training data.

10. The control method as described in claim 9, further comprising: The AI ​​model is trained by adding images indicating the boundaries of the corresponding target structures as training data to the cross-sectional images that include the multiple target structures, wherein the feature factors of the target structures are reflected in the images indicating the boundaries of the corresponding target structures.

11. An ultrasound diagnostic device, comprising: The probe is configured to send ultrasonic signals to the object and receive echo signals reflected from the object; The display is configured to show an ultrasound image based on echo information received by the probe; The input interface is configured to receive user input. as well as The processor is configured to: control the probe to acquire an animal ultrasound image; extract a cross-sectional image from the animal ultrasound image comprising a cross section of a plurality of target structures; obtain a processed image from the cross-sectional image comprising a contour drawing image of the plurality of target structures; obtain the centroid of each of the contour drawing images; and obtain length information of the target structures based on the intersection point between at least one line passing through the centroid of the contour drawing image and the boundary of the contour drawing image.

12. The ultrasound diagnostic device as described in claim 11, wherein, The target structures include the left atrium and the aorta.

13. The ultrasound diagnostic device as described in claim 12, wherein, The length information of the target structure includes the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta.

14. The ultrasound diagnostic device as described in claim 13, wherein, The processor is configured to calculate the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameters of the left atrium to the aorta based on the intersection of a line through all centroids of the corresponding contour drawing images passing through the left atrium and the aorta with the boundary of the corresponding contour drawing image.

15. The ultrasound diagnostic device as described in claim 14, wherein, The processor is configured to control the display to display the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta, together with the cross-sectional image.

16. The ultrasound diagnostic device as described in claim 15, wherein, The processor is configured to control the display to display the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta as annotations on one side of the cross-sectional image.

17. The ultrasound diagnostic device as described in claim 15, wherein, The processor is configured to: control the display to show icon indicators and numerical values ​​for the diameter of the left atrium, the diameter of the aorta, or the ratio of the diameter of the left atrium to the diameter of the aorta, and The icon indicator includes a first indicator and a second indicator, the first indicator indicating the intersection point between a line passing through all centroids of the corresponding contour drawing image of the left atrium and the aorta and the boundary of the corresponding contour drawing image, and the second indicator indicating a connecting line connecting the intersection points in the corresponding contour drawing image.

18. The ultrasound diagnostic device of claim 11, further comprising a memory configured for storing an AI model, wherein, The processor is configured to input a cross-sectional image of the animal ultrasound image, including the cross-sections of the plurality of target structures, as input data into an AI model stored in the memory, and to obtain a processed image, including the contour drawing images of the plurality of target structures, as output data.

19. The ultrasound diagnostic device as described in claim 18, wherein, The processor is configured to train the AI ​​model by using the cross-sectional image of the cross-section including the plurality of target structures as input data and using an image indicating the boundary of the corresponding target structure that can be checked in the cross-sectional image including the plurality of target structures as training data.

20. The ultrasound diagnostic device as described in claim 19, wherein, The processor is configured to train the AI ​​model by adding images indicating the boundaries of the corresponding target structures as training data to the cross-sectional image including the plurality of target structures, wherein the feature factors of the target structures are reflected in the images indicating the boundaries of the corresponding target structures.

Images