WO2025198331A1 - Dispositif électronique pour déterminer le risque de nodules sur la base d'une pluralité d'ensembles d'images d'entrée, procédé de fonctionnement associé et support d'enregistrement pour la mise en œuvre dudit procédé - Google Patents

Dispositif électronique pour déterminer le risque de nodules sur la base d'une pluralité d'ensembles d'images d'entrée, procédé de fonctionnement associé et support d'enregistrement pour la mise en œuvre dudit procédé

Info

Publication number
WO2025198331A1
WO2025198331A1 PCT/KR2025/003567 KR2025003567W WO2025198331A1 WO 2025198331 A1 WO2025198331 A1 WO 2025198331A1 KR 2025003567 W KR2025003567 W KR 2025003567W WO 2025198331 A1 WO2025198331 A1 WO 2025198331A1
Authority
WO
WIPO (PCT)
Prior art keywords
nodule
electronic device
information
input image
image set
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
PCT/KR2025/003567
Other languages
English (en)
Korean (ko)
Inventor
오동렬
박재현
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Monitor Corp Co Ltd
Original Assignee
Monitor Corp Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from KR1020240200282A external-priority patent/KR20250142180A/ko
Application filed by Monitor Corp Co Ltd filed Critical Monitor Corp Co Ltd
Publication of WO2025198331A1 publication Critical patent/WO2025198331A1/fr
Pending legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
    • A61B6/02—Arrangements for diagnosis sequentially in different planes; Stereoscopic radiation diagnosis
    • A61B6/03—Computed tomography [CT]
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H30/00—ICT specially adapted for the handling or processing of medical images
    • G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H30/00—ICT specially adapted for the handling or processing of medical images
    • G16H30/40—ICT specially adapted for the handling or processing of medical images for processing medical images, e.g. editing
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/50—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for simulation or modelling of medical disorders
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients

Definitions

  • Various embodiments disclosed in this document relate to an electronic device for determining the risk of a nodule based on a plurality of sets of input images, a method of operating the same, and a recording medium for performing the method.
  • CT computed tomography
  • Various embodiments disclosed in this document aim to provide an electronic device that detects at least one nodule contained in at least a part of a user's body using artificial intelligence.
  • the various embodiments disclosed in this document aim to provide an electronic device that accurately determines the risk level of at least one nodule contained in at least a part of a user's body using artificial intelligence.
  • Various embodiments disclosed in this document aim to provide an electronic device that provides a user with information about the risk of at least one nodule contained in at least a part of the user's body using artificial intelligence.
  • an electronic device includes a communication circuit for transmitting and receiving data with an external device, at least one processor, and a memory for storing commands, wherein the commands are individually or collectively executed by the at least one processor so that the electronic device obtains a first input image set in which at least a part of a user's body is photographed at a first point in time, detects at least one first nodule with respect to the first input image set, determines a similarity between the at least one first nodule and the at least one second nodule, and the at least one second nodule is obtained based on a second input image set stored in the memory, and when the similarity is equal to or greater than a preset value, inputs the first input image set and the second input image set into a risk assessment model to obtain risk information about the at least one first nodule, and the risk assessment model may be an artificial intelligence model trained to output risk information about the at least one first nodule based on first size information about the at least one first nodule, first type information about
  • a method of operating an electronic device may include: acquiring a first input image set in which at least a part of a user's body is captured at a first point in time; detecting at least one first nodule from the first input image set; determining a similarity between the at least one first nodule and the at least one second nodule; wherein the at least one second nodule is acquired based on a second input image set stored in the electronic device, and when the similarity is equal to or greater than a preset value, inputting the first input image set and the second input image set into a risk assessment model to obtain risk information for the at least one first nodule, wherein the risk assessment model may be an artificial intelligence model trained to output risk information for the at least one first nodule based on first size information for the at least one first nodule, first type information for the at least one first nodule, and variation information for the at least one first nodule and the at least one second nodule.
  • a non-transitory computer-readable recording medium including a program for executing a control method of an electronic device includes the steps of: obtaining a first input image set in which at least a part of a user's body is photographed at a first point in time; detecting at least one first nodule with respect to the first input image set; determining a similarity between the at least one first nodule and the at least one second nodule; obtaining the at least one second nodule based on the second input image set stored in the electronic device; and, if the similarity is equal to or greater than a preset value, inputting the first input image set and the second input image set into a risk assessment model to obtain risk information for the at least one first nodule, wherein the risk assessment model may be an artificial intelligence model trained to output risk information for the at least one first nodule based on first size information of the at least one first nodule, first type information of the at least one first nodule, and variation information of the at least one first nodu
  • An electronic device can detect at least one nodule included in at least a part of a user's body using artificial intelligence.
  • An electronic device can accurately determine the risk level of at least one nodule included in at least a part of a user's body using artificial intelligence.
  • An electronic device may provide a user with information about the risk of at least one nodule included in at least a part of the user's body using artificial intelligence.
  • An electronic device can use artificial intelligence to determine the risk of at least one nodule contained in at least a part of a user's body, thereby increasing the success rate of early detection and treatment of diseases such as lung cancer.
  • FIG. 1 is a diagram illustrating a system in various embodiments.
  • FIG. 2 is a block diagram of an electronic device according to various embodiments.
  • FIG. 3 illustrates a concept for controlling a function related to risk assessment for at least one node of an electronic device, according to various embodiments.
  • FIG. 4 is a flowchart illustrating an operation of an electronic device according to various embodiments to obtain risk information based on lung disease information.
  • FIG. 5 is a diagram illustrating a set of input images according to various embodiments.
  • FIG. 6 is a diagram illustrating at least one artificial intelligence model according to various embodiments.
  • FIG. 7 is a flowchart illustrating an operation of an electronic device according to various embodiments to correct primary risk information based on lung disease information.
  • FIG. 8 is a diagram illustrating an electronic device obtaining risk information according to various embodiments.
  • FIG. 9 is a flowchart illustrating an operation of an electronic device according to various embodiments to display risk information depending on whether or not a patient has a lung disease.
  • Figure 10 is a flowchart illustrating an operation of obtaining risk information based on multiple sets of input images.
  • FIG. 11 is a diagram illustrating a method for an electronic device to determine a set of nodules according to various embodiments.
  • FIG. 12A is a flowchart illustrating an operation of an electronic device to determine similarity according to various embodiments.
  • FIG. 12B is a diagram illustrating a method for an electronic device to determine similarity according to various embodiments.
  • FIG. 12c is a diagram illustrating a method for an electronic device to determine similarity according to various embodiments.
  • FIG. 13 is a flowchart illustrating an operation of an electronic device according to various embodiments to obtain risk information of a nodule by considering whether or not it has a lung disease.
  • FIG. 14 is a diagram illustrating an electronic device according to various embodiments obtaining risk information based on a plurality of input image sets.
  • first and/or “second” may be used to describe various components, these components should not be limited by these terms. These terms are only intended to distinguish one component from another; for example, without departing from the scope of the present disclosure, a first component may be referred to as a "second component,” and similarly, a second component may also be referred to as a "first component.”
  • FIG. 1 is a diagram illustrating a system in various embodiments.
  • the system (100) may include a user terminal (110), an electronic device (130), and a database (150).
  • the user terminal (110) may be a variety of devices capable of verifying data generated by the electronic device (130) and metadata analyzed therefrom.
  • the user terminal (110) may include a portable communication device (e.g., a smartphone), a computer device, a portable multimedia device, a medical device, a camera, a wearable device, or a home appliance.
  • the electronic device according to the embodiments of the present document is not limited to the aforementioned devices.
  • the user terminal (110) may be connected to the electronic device (130) via a network, and multiple user terminals (110) may be connected to the electronic device (130) simultaneously.
  • the user terminal (110) may include a display.
  • the display may visually provide information to an external party (e.g., a user) of the user terminal (110).
  • the display may display various contents (e.g., text, images, videos, icons, and/or symbols). For example, it may display risk information for a node obtained from an electronic device (130).
  • the display may include a liquid crystal display (LCD), a light emitting diode (LED) display, or an organic light emitting diode (OLED) display for this purpose.
  • LCD liquid crystal display
  • LED light emitting diode
  • OLED organic light emitting diode
  • the electronic device (130) may include a device that determines the risk level of at least one nodule included in at least a part of the user's body based on a set of input images that have been obtained by capturing at least a part of the user's body.
  • the electronic device (130) may be provided by being included in a computer-readable recording medium by tangibly implementing a program of commands for implementing the same.
  • the electronic device (130) may be implemented in the form of program commands that can be executed through various computer means and may be recorded in a computer-readable recording medium.
  • the electronic device (130) may be configured as a computer program that sequentially or non-sequentially performs operations of receiving an image (or a set of images) that has captured at least a part of the user's body and analyzing the same, and the computer program may be stored in a computer-readable recording medium.
  • the database (150) may correspond to a storage device that stores various pieces of information generated through an electronic device (130) and/or a user terminal (110).
  • FIG. 2 is a block diagram of an electronic device according to various embodiments.
  • the electronic device (130) may be implemented by including a processor (210), a memory (230), a user input/output unit (250), and a communication circuit (270).
  • the components listed above may be operatively or electrically connected to each other.
  • the components of the electronic device (130) illustrated in FIG. 2 may be modified, deleted, or added in part, as an example.
  • the electronic device (130) may further include an output device (e.g., a display (e.g., the display described with reference to FIG. 1)).
  • an output device e.g., a display (e.g., the display described with reference to FIG. 1)
  • the electronic device (130) includes an output device, various information acquired through the electronic device (130) may be provided through the output device.
  • the processor (210) may include at least one processor implemented to provide at least some different functions.
  • the processor (210) may control the overall operation of the electronic device (130) and may be electrically connected to the memory (230), the user input/output unit (250), and the communication circuit (270) to control data flow therebetween.
  • the processor (210) may be implemented as a CPU (Central Processing Unit) of the electronic device (130).
  • the processor (210) may include a main processor (e.g., a central processing unit or an application processor) or an auxiliary processor (e.g., a graphics processing unit, a neural processing unit (NPU), an image signal processor, a sensor hub processor, or a communication processor) that can operate independently of or together with the main processor.
  • a main processor e.g., a central processing unit or an application processor
  • auxiliary processor e.g., a graphics processing unit, a neural processing unit (NPU), an image signal processor, a sensor hub processor, or a communication processor
  • the auxiliary processor may be configured to use lower power than the main processor or to be specialized for a given function.
  • the auxiliary processor may be implemented separately from the main processor or as a part thereof.
  • the auxiliary processor may control at least a portion of functions or states associated with at least one of the components of the electronic device (130), for example, on behalf of the main processor while the main processor is in an inactive (e.g., sleep) state, or together with the main processor while the main processor is in an active (e.g., application execution) state.
  • the auxiliary processor e.g., an image signal processor or a communication processor
  • the auxiliary processor may include a hardware structure specialized for processing artificial intelligence models.
  • the artificial intelligence models may be generated through machine learning. Such learning may be performed, for example, in the electronic device (130) itself on which the artificial intelligence model is executed, or may be performed through a separate server.
  • the learning algorithm may include, but is not limited to, supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning, for example.
  • the artificial intelligence model may include multiple artificial neural network layers.
  • the artificial neural network may be one of a deep neural network (DNN), a convolutional neural network (CNN), a recurrent neural network (RNN), a restricted Boltzmann machine (RBM), a deep belief network (DBN), a bidirectional recurrent deep neural network (BRDNN), a deep Q-network, or a combination of two or more of the above, but is not limited to the examples described above.
  • the artificial intelligence model may additionally or alternatively include a software structure.
  • the operation of the electronic device (130) described below may be understood as the operation of the processor (210).
  • the memory (230) may include an auxiliary memory device implemented with a non-volatile memory such as an SSD (Solid State Drive) or an HDD (Hard Disk Drive) and used to store all data required for the electronic device (130), and may include a main memory device implemented with a volatile memory such as a RAM (Random Access Memory).
  • the memory (230) may include a plurality of instructions that direct the operations of the processor (210) to implement the functions provided by the service.
  • the processor (210) may include a software server that executes the functions provided by the service based on the plurality of instructions stored in the memory (230).
  • the user input/output unit (250) may include an environment for receiving user input and an environment for outputting specific information to the user.
  • the user input/output unit (250) may include an input device including an adapter such as a touchpad, a touch screen, a virtual keyboard, or a pointing device, and an output device including an adapter such as a monitor or a touch screen.
  • the user input/output unit (250) may correspond to a computing device accessed via a remote connection, in which case the electronic device (130) may function as a server.
  • the electronic device (130) may include a communication circuit (270).
  • the communication circuit (270) may support the establishment of a direct (e.g., wired) communication channel or a wireless communication channel between the electronic device (130) and an external electronic device (e.g., the user terminal (110) of FIG. 1), and the performance of communication through the established communication channel.
  • the communication circuit (270) may operate independently from the processor (210) and may include one or more communication processors that support direct (e.g., wired) communication or wireless communication.
  • the communication circuit (270) may include a wireless communication module (e.g., a cellular communication module, a short-range wireless communication module, or a global navigation satellite system (GNSS) communication module) or a wired communication module (e.g., a local area network (LAN) communication module, or a power line communication module).
  • a wireless communication module e.g., a cellular communication module, a short-range wireless communication module, or a global navigation satellite system (GNSS) communication module
  • GNSS global navigation satellite system
  • wired communication module e.g., a local area network (LAN) communication module, or a power line communication module.
  • any of these communication modules may communicate with an external electronic device via a first network (e.g., a short-range communication network such as Bluetooth, WiFi Direct (wireless fidelity direct), or IrDA (infrared data association)) or a second network (e.g., a long-range communication network such as a legacy cellular network, a 5G network, a next-generation communication network, the Internet, or a computer network (e.g., a LAN or WAN)).
  • a first network e.g., a short-range communication network such as Bluetooth, WiFi Direct (wireless fidelity direct), or IrDA (infrared data association)
  • a second network e.g., a long-range communication network such as a legacy cellular network, a 5G network, a next-generation communication network, the Internet, or a computer network (e.g., a LAN or WAN)
  • a first network e.g., a short-range communication network such as Bluetooth, WiFi Direct (
  • FIG. 3 illustrates a concept for controlling a function related to risk assessment for at least one node of an electronic device, according to various embodiments.
  • the electronic device (130) may utilize hardware and/or software modules to support functions related to assessing the risk of at least one nodule.
  • the processor (210) may drive at least one of the input image receiving unit (310), the nodule detection unit (320), the size information obtaining unit (330), the type information obtaining unit (340), the lung disease information obtaining unit (350), the similarity determining unit (360), the variation information obtaining unit (370), and the risk information obtaining unit (380) by executing commands stored in the memory (230).
  • software modules other than those illustrated in FIG. 3 may be implemented. For example, at least two modules may be integrated into one module, or one module may be split into two or more modules.
  • the electronic device (130) may include both an encoder implemented as hardware and an encoder implemented as a software module, and some of the data acquired through at least one camera module may be processed by the hardware encoder and some of the data may be processed by the software encoder.
  • the input image receiving unit (310) may receive a set of input images.
  • the set of input images may be images capturing at least a portion of the user's body.
  • the set of input images may be images capturing the user's chest.
  • the input image receiving unit (310) may obtain sets of input images capturing at least a portion of the user's body at various points in time. For example, the input image receiving unit (310) may obtain a first set of input images capturing at least a portion of the user's body at a first point in time, and a second set of input images of the user captured at least one point in time that is distinct from the first point in time.
  • the nodule detection unit (320) can detect at least one nodule from a set of input images that capture at least a part of a user's body. For example, the nodule detection unit (320) can obtain a nodule set by detecting at least one nodule from a set of input images through at least one artificial intelligence model (e.g., the first artificial intelligence model (610) described with reference to FIG. 6, the nodule detection model (1113) described with reference to FIG. 11).
  • the first artificial intelligence model (610) described with reference to FIG. 6
  • the nodule detection unit (320) can detect at least one nodule through a segmentation model trained to detect at least one nodule from a set of input images that capture at least a part of a user's body (e.g., the first artificial intelligence model (610) described with reference to FIG. 6, the nodule detection model (1113) described with reference to FIG. 11).
  • a segmentation model trained to detect at least one nodule from a set of input images that capture at least a part of a user's body (e.g., the first artificial intelligence model (610) described with reference to FIG. 6, the nodule detection model (1113) described with reference to FIG. 11).
  • the size information acquisition unit (330) may acquire size information of at least one detected nodule.
  • the size information acquisition unit (330) may input the input image set into at least one artificial intelligence model (e.g., the first artificial intelligence model (610) described with reference to FIG. 6) to acquire size information for each of at least one nodule identified in the input image set.
  • the size information for the at least one nodule may be determined based on the volume of the at least one nodule.
  • the size information acquisition unit (330) may utilize at least one artificial intelligence model.
  • the at least one artificial intelligence model may include a model that specifies an evaluation area included in an input image set, a model that determines an abnormal area (e.g., an area determined to include at least one nodule) included in the evaluation area, and a model that determines the volume of the abnormal area to acquire size information.
  • the type information acquisition unit (340) can determine the type of at least one nodule detected from a set of input images that capture at least a part of the user's body. For example, the type information acquisition unit (340) can acquire type information on at least one nodule in an area determined to be at least one nodule through at least one artificial intelligence model (e.g., the second artificial intelligence model (620) described with reference to FIG. 6, a type detection model). For example, the type information acquisition unit (340) can acquire type information on at least one nodule through a type classification model (type detection model) learned to classify the type of at least one nodule detected by the nodule detection unit (320).
  • a type classification model type detection model learned to classify the type of at least one nodule detected by the nodule detection unit (320).
  • the above type information may include various information related to the characteristics of at least one of the nodules, such as classification according to morphological characteristics (solid nodules, partially solid nodules, non-solid nodules), classification according to size (small nodules, medium nodules, giant nodules), classification according to growth and boundary characteristics (nodules with clear boundaries, nodules with unclear boundaries), classification according to pathological characteristics (benign nodules, malignant nodules, metastatic nodules), and other classifications (calcified nodules, air-containing nodules).
  • morphological characteristics solid nodules, partially solid nodules, non-solid nodules
  • classification according to size small nodules, medium nodules, giant nodules
  • classification according to growth and boundary characteristics nodules with clear boundaries, nodules with unclear boundaries
  • classification according to pathological characteristics benign nodules, malignant nodules, metastatic nodules
  • other classifications calcified nodules, air-containing nodules.
  • the risk information acquisition unit (380) may acquire risk information for at least one nodule detected in an input image set that captures at least a portion of the user's body. For example, the risk information acquisition unit (380) may determine the risk for the at least one nodule based on at least one of size information, type information, and lung disease information for the at least one nodule. For example, the risk information acquisition unit (380) may acquire risk information indicating risk information for the at least one nodule using at least one artificial intelligence model (e.g., the risk determination model (830) described with reference to FIG. 8, the risk determination model (1430) described with reference to FIG. 14).
  • the risk information acquisition unit (380) may acquire risk information indicating risk information for the at least one nodule using at least one artificial intelligence model (e.g., the risk determination model (830) described with reference to FIG. 8, the risk determination model (1430) described with reference to FIG. 14).
  • the risk information acquisition unit (380) may acquire primary risk information based on size information and type information for the at least one nodule, and may correct the primary risk information based on the lung disease information to acquire risk information.
  • the risk information acquisition unit (380) may receive size information, type information, and lung disease information for at least one nodule and use an artificial intelligence model trained to determine the risk for at least one nodule to acquire risk information for at least one nodule.
  • the primary risk information may be determined based on a designated auxiliary indicator (e.g., Lung-RADS (Lung Imaging Reporting and Data System) category).
  • a designated auxiliary indicator e.g., Lung-RADS (Lung Imaging Reporting and Data System) category.
  • the risk information acquisition unit (380) may acquire risk information on at least one first nodule detected in the first input image set based on a first input image set in which at least a part of the user's body is captured at a first point in time, and a second input image set in which at least a part of the user's body is captured at at least one point in time that is distinct from the first point in time.
  • the risk information acquisition unit (380) may determine the risk on the at least one first nodule based on first size information and first type information on the at least one first nodule detected based on the first input image set, and variation amount information on the at least one second nodule and the at least one first nodule detected based on the second input image set.
  • the risk information acquisition unit (380) can acquire risk information indicating risk information for the at least one first nodule by using at least one artificial intelligence model (e.g., the risk judgment model (830) described with reference to FIG. 8, the risk judgment model (1430) described with reference to FIG. 14).
  • the risk information acquisition unit (380) can acquire primary risk information based on first size information and first type information for the at least one first nodule, and can acquire risk information for the at least one first nodule by considering change amount information between the at least one first nodule and the at least one second nodule.
  • the risk information acquisition unit (380) may consider the user's lung disease information when acquiring risk information for the at least one first nodule. For example, the risk information acquisition unit (380) may acquire risk information for the at least one first nodule based on the user's lung disease information, the first size information of the at least one first nodule, the first type information, and the change amount information. For example, the risk information acquisition unit (380) may receive the first size information, the first type information, the change amount information, and the lung disease information for the at least one first nodule, and may acquire risk information for the at least one first nodule using an artificial intelligence model trained to determine the risk of the at least one first nodule.
  • the data set may be implemented on a user interface (UI) and displayed to the user via a display. Furthermore, the data set may be displayed to the user via an external device (e.g., the user terminal (110) of FIG. 1, a display device) connected to the electronic device (130).
  • UI user interface
  • an external device e.g., the user terminal (110) of FIG. 1, a display device
  • FIG. 4 is a flowchart (400) showing an operation of an electronic device according to various embodiments to obtain risk information based on lung disease information.
  • FIG. 6 is a diagram illustrating at least one artificial intelligence model according to various embodiments.
  • the operations described below may be performed in combination with one another.
  • the operations performed by the electronic device (130) may refer to operations performed by the processor (210) of the electronic device (130).
  • the operations illustrated in FIG. 4 may be performed in various orders, not limited to the order illustrated. Furthermore, according to various embodiments, more operations may be performed than those illustrated in FIG. 4, or at least one operation may be performed less than those illustrated in FIG.
  • the electronic device (130) may obtain a set of input images capturing at least a portion of the user's body in operation 401.
  • the input image set (510) may be an image taken of the user's body, such as an MRI, CT, or X-ray.
  • the input image set (510) may include a plurality of slices.
  • Meta-information related to the image information such as slice thickness, resolution, patient information, scan date and time, equipment information, scan direction, use of contrast agent, radiation dose, image format and size, reconstruction parameters, scan range, sequence, and protocol information of the input image set (510), may be stored in the form of DICOM (Digital Imaging and Communications in Medicine) data. That is, the input image set (510) may be stored as a DICOM file, and may be stored in a form that includes metadata together with the image itself.
  • DICOM Digital Imaging and Communications in Medicine
  • the electronic device (130) can classify at least one nodule included in the input image set (510) into an inactive area (603) or an active area (605) through a first artificial intelligence model (610) (e.g., a nodule detection model).
  • the first artificial intelligence model (610) can specify an evaluation area based on at least one nodule included in the input image set (510).
  • the evaluation area can be a slice that includes at least a portion of at least one nodule.
  • the evaluation area can include a plurality of slices, and a single continuous nodule can include a plurality of slices.
  • the first artificial intelligence model (610) may determine at least one nodule portion included in the evaluation area and determine size information (601) of the at least one nodule. For example, the first artificial intelligence model (610) may determine a volume of the at least one nodule and output size information (601) for the at least one nodule based on the volume. In addition, the first artificial intelligence model (610) may classify the evaluation area as an inactive area (603) when the volume of the at least one nodule included in the evaluation area is less than or equal to a cutoff value.
  • the cutoff value is a set value and may be set differently depending on the user.
  • the first artificial intelligence model (610) may be a model trained to identify an evaluation area including at least one nodule and determine the volume of at least one nodule. That is, the first artificial intelligence model (610) may be trained in various ways and is not limited to any one method. Furthermore, the first artificial intelligence model (610) may include at least one sub-AI model, and each of the sub-AI models may be trained in various ways.
  • the first sub-AI model of the first AI model (610) may be trained to automatically identify an evaluation region containing at least one nodule.
  • the first sub-AI model may be trained to identify the presence of pulmonary nodules, tumors, or abnormal tissues in medical images such as computed tomography (CT) or magnetic resonance imaging (MRI).
  • CT computed tomography
  • MRI magnetic resonance imaging
  • the first sub-AI model may learn the characteristics of abnormal areas determined to be nodules in an image and distinguishing factors from surrounding normal tissues through a deep learning method on an annotated medical image data set.
  • the second sub-AI model of the first AI model (610) may be trained to output size information of at least one detected nodule.
  • the electronic device (130) may determine size information of at least one nodule based on the number of pixels of at least one nodule on the slice through the first AI model (610). For example, the boundary of at least one nodule may be detected through the second sub-AI model of the first AI model (610), and the volume of at least one nodule may be determined.
  • the electronic device (130) can determine size information of at least one nodule using DICOM data.
  • the electronic device (130) can determine size information (e.g., volume) of at least one nodule using data such as slice interval and resolution included in the DICOM data.
  • the electronic device (130) may obtain type information about the at least one node.
  • the electronic device (130) obtaining type information about the at least one node may be identical to or similar to the operation of the type information obtaining unit (340) described with reference to FIG. 3.
  • redundant descriptions are omitted.
  • the electronic device (130) can obtain type information (607) for at least one nodule included in the input image set (510) through the second artificial intelligence model (620) (e.g., type detection model).
  • the electronic device (130) determines the evaluation area as an activation area (605) and inputs the activation area (605) into the second artificial intelligence model (620) to obtain type information (607) of the at least one nodule.
  • the second artificial intelligence model (620) may be a model trained to classify the type of at least one nodule included in the activation area (605) and output type information.
  • the second artificial intelligence model may be a model trained to receive a nodule set including at least one nodule as input and output type information for the at least one nodule.
  • the inactive region (603) may be excluded from analysis by the second artificial intelligence model (620).
  • the inactive region (603) may be classified as not a nodule in the results for a patient provided through the implementation of the present disclosure, and may thus be excluded from further analysis by an expert.
  • the active region (605) may be classified as a nodule in the results for a patient provided through the implementation of the present disclosure, and may thus be requested for further analysis by an expert.
  • classification as an inactive region (6603) or an active region (605) may provide convenience in terms of accuracy and speed in patient analysis to experts.
  • the electronic device (130) may obtain lung disease information regarding whether the user has a lung-related disease.
  • the electronic device (130) may obtain information about lung diseases, such as whether the user has a lung-related disease, and if so, what type of disease, based on a set of input images that capture at least a part of the user's body.
  • the electronic device (130) may obtain lung disease information including whether the user has a lung disease (and/or the type of lung disease) based on a set of input images through at least one artificial intelligence model (e.g., the third artificial intelligence model described with reference to FIG. 8, a lung disease detection model).
  • an artificial intelligence model e.g., the third artificial intelligence model described with reference to FIG. 8, a lung disease detection model.
  • the electronic device (130) can obtain the lung disease information through a lung disease detection model, which is an artificial intelligence model trained to detect whether a lung disease such as pneumonia, tuberculosis, sarcoidosis, fungal infection, pulmonary vascular disease, airway disease, pleural disease, etc. exists and the type of lung disease based on an input image set that captures at least a part of the user's body.
  • a lung disease detection model which is an artificial intelligence model trained to detect whether a lung disease such as pneumonia, tuberculosis, sarcoidosis, fungal infection, pulmonary vascular disease, airway disease, pleural disease, etc. exists and the type of lung disease based on an input image set that captures at least a part of the user's body.
  • the electronic device (130) may obtain risk information for at least one nodule based on the size information, the type information, and the lung disease information. For example, the electronic device (130) may obtain risk information for at least one nodule detected in a set of input images that capture at least a part of the user's body. For example, the electronic device (130) may determine the risk for the at least one nodule based on at least one of the size information, the type information, and the lung disease information for the at least one nodule. For example, the electronic device (130) may obtain risk information indicating risk information for the at least one nodule using at least one artificial intelligence model (e.g., the risk determination model (830) described with reference to FIG. 8, the risk determination model (1430) described with reference to FIG. 14).
  • the electronic device (130) may obtain risk information indicating risk information for the at least one nodule using at least one artificial intelligence model (e.g., the risk determination model (830) described with reference to FIG. 8, the risk determination model (1430) described with reference
  • FIG. 7 is a flowchart (700) showing an operation of an electronic device according to various embodiments to correct primary risk information based on lung disease information.
  • FIG. 8 is a diagram illustrating an electronic device obtaining risk information according to various embodiments.
  • the operations described below may be performed in combination with one another.
  • the operations performed by the electronic device (130) may refer to operations performed by the processor (210) of the electronic device (130).
  • the operations illustrated in FIG. 7 may be performed in various orders, not limited to the order illustrated. Furthermore, according to various embodiments, more operations may be performed than those illustrated in FIG. 7, or at least one operation may be performed less than those illustrated in FIG.
  • the electronic device (130) may obtain primary risk information for at least one nodule based on size information and type information in operation 701.
  • the electronic device (130) may obtain size information and type information for at least one nodule based on a set of input images that capture at least a portion of the user's body through the operation described with reference to FIG. 4.
  • the electronic device (130) can obtain an input image set (810) that captures at least a portion of the user's body (e.g., chest).
  • the electronic device (130) can input the input image set (810) into at least one artificial intelligence model (820) to obtain size information (821), type information (823), and lung disease information (825).
  • the electronic device (130) may obtain primary risk information for each of at least one target node based on size information (821)) and type information (823)).
  • the electronic device (130) may obtain risk information by correcting the primary risk information based on the lung disease information.
  • the electronic device (130) may obtain risk information (840) using the risk judgment model (830).
  • the electronic device (130) may obtain the primary risk information (831) based on the size information (821) and the type information (823).
  • the electronic device (130) may obtain risk information (840) by correcting the primary risk information (831) based on whether or not the device has a lung disease (833).
  • the electronic device (130) can determine primary risk information (831) based on auxiliary indicators (e.g., Lung-RADS (Lung Imaging Reporting and Data System) category) specified according to size information (821) and type information (823) through a risk assessment model (830), and can obtain risk information (840) by correcting (e.g., increasing) the numerical value of the primary risk information according to whether or not there is a lung disease (833).
  • auxiliary indicators e.g., Lung-RADS (Lung Imaging Reporting and Data System) category
  • the at least one artificial intelligence model (820) and the risk assessment model (830) may be configured as a single artificial intelligence model.
  • the electronic device (130) may input an input image set (810) into the artificial intelligence model to obtain risk information (840).
  • the risk information (840) may be obtained through an operation similar to the operation described above.
  • FIG. 9 is a flowchart (900) showing an operation of an electronic device according to various embodiments to display risk information depending on whether or not a patient has a lung disease.
  • the operations described below may be performed in combination with one another.
  • the operations performed by the electronic device (130) may refer to operations performed by the processor (210) of the electronic device (130).
  • the operations illustrated in FIG. 9 may be performed in various orders, not limited to the order illustrated. Furthermore, according to various embodiments, more operations may be performed than those illustrated in FIG. 9, or at least one operation may be performed less than those illustrated in FIG.
  • the electronic device (130) may output risk information acquired based on a set of input images capturing at least a portion of the user's body. For example, the electronic device (130) may transmit the risk information to an external device (e.g., the user terminal (110) of FIG. 1) so that the risk information is output through the display of the external device.
  • an external device e.g., the user terminal (110) of FIG. 1
  • the electronic device (130) can determine whether the user has a lung disease in operation 901. For example, the electronic device (130) can determine whether the risk information has been corrected based on whether the user has a lung disease when calculating the risk information to be displayed.
  • the electronic device (130) may transmit the risk information to an external device so as to display risk information based on a first visual object. For example, if the risk information for at least one nodule is determined by taking into account that the user has a lung disease, that is, if the risk information is determined by correcting the primary risk information for at least one nodule, the electronic device (130) may display the risk information based on the first visual object. For example, the electronic device (130) may display the risk information based on a first visual object that includes a mark (e.g., a distinguished mark) indicating that the risk information is corrected information depending on whether the user has a lung disease.
  • a mark e.g., a distinguished mark
  • the electronic device (130) may transmit the risk information to an external device in operation 905 so as to display the risk information based on a second visual object if the user does not have a lung disease (901-No).
  • the electronic device (130) may display the risk information based on a second visual object if the risk information for at least one nodule is determined by taking into account that the user does not have a lung disease, i.e., if the risk information is determined based on size information and type information for at least one nodule.
  • the first visual object and the second visual object can be distinguished. Accordingly, the user can distinguish whether the risk for at least one nodule is a risk adjusted due to lung disease. For example, if the risk is determined to be 3 based on the size information and type information of at least one nodule, and if the initial risk information is determined to be 2 based on the size information and type information of at least one nodule and then the risk is adjusted to 3 due to lung disease, the two can be distinguished by using distinct visual objects, even though the risk values are the same.
  • Figure 10 is a flowchart (1000) showing an operation of obtaining risk information based on multiple input image sets.
  • FIG. 11 is a diagram illustrating a method for an electronic device to determine a set of nodules according to various embodiments.
  • the operations described below may be performed in combination with one another.
  • the operations performed by the electronic device (130) may refer to operations performed by the processor (210) of the electronic device (130).
  • the operations illustrated in FIG. 10 may be performed in various orders, not limited to the order illustrated. Furthermore, according to various embodiments, more operations may be performed than those illustrated in FIG. 10, or at least one operation may be performed less than those illustrated in FIG.
  • the electronic device (130) may acquire a first set of input images taken of at least a portion of the user's body at a first point in time.
  • the electronic device (130) can detect at least one first nodule for a first set of input images.
  • the electronic device (130) can detect at least one second nodule from a second set of input images acquired at at least one time point distinct from the first time point.
  • the electronic device (130) detects at least one first nodule and at least one second nodule based on the first input image set and the second input image set.
  • the electronic device (130) can input a first input image set (1111) captured at a first point in time into a nodule detection model (1113) to obtain a first activation area (1115) including at least one first nodule.
  • the electronic device (130) can input a second input image set (1121) captured at at least one point in time distinct from the first point in time into the nodule detection model (1113) to obtain a second activation area (1125) including at least one second nodule.
  • the nodule detection model (1113) (e.g., the first artificial intelligence model (610) of FIG. 6, the nodule detection model described with reference to FIG. 3) may be an artificial intelligence model trained to detect at least one nodule included in at least a part of the user's body based on an input image set.
  • the first input image set and the second input image set may include images taken of at least a part of the user's body (e.g., the chest).
  • the first input image set (1111) may be an image of the user's body taken of a first time point, such as an MRI, CT, or X-ray.
  • the second input image set (1121) may be an image of the user's body taken of at least one time point (e.g., a past time point) different from the first time point.
  • the first input image set (1111) and the second input image set (1121) may each include a plurality of slides, and may be 3D images.
  • the first input image set (1111) and the second input image set (1121) may each include a plurality of slides including at least one nodule.
  • the first input image set (1111) and the second input image set (1121) may be images captured in the same format.
  • the first input image set (1111) may be images captured at a first point in time
  • the second input image set (1121) may be images captured at least at a point in time that is distinct from the first point in time, and may be images captured in the past that is temporally earlier than the first point in time.
  • the electronic device (130) can determine a first activation region (1115) including at least one first nodule included in a first input image set (1111) and a second activation region (1125) including at least one second nodule included in a second input image set (1121) through a nodule detection model (1113).
  • the nodule detection model (1113) can be composed of at least one artificial intelligence model.
  • the nodule detection model (1113) can specify evaluation regions included in the first input image set (1111) and the second input image set (1121).
  • the nodule detection model (1113) can determine at least one nodule portion included in the evaluation region and determine a volume of the at least one nodule.
  • the nodule detection model (1113) can classify the evaluation region as an inactive region when the volume of at least one nodule included in the evaluation region is less than or equal to a cutoff value. Conversely, the nodule detection model (1113) can classify the evaluation area as an activated area if the volume of at least one nodule is greater than the cutoff value.
  • the activated area can be the image itself cropped to the evaluation area and can be determined to a preset size.
  • at least one nodule can be a set of activated area data including nodules.
  • Each of the at least one first nodule and the at least one second nodule can include at least one piece of activated area data as data determined from images captured at a first time point and at least one time point distinct from the first time point.
  • the electronic device (130) can determine at least one nodule including an active area through the nodule detection model (1113).
  • the active area can include at least one segmented nodule portion (e.g., at least one first nodule, at least one second nodule).
  • the active area can include at least one nodule portion included within a preset size and a surrounding area thereof. That is, the electronic device (130) can determine an active area including at least one nodule portion and a surrounding area through the nodule detection model (1113).
  • the activation area can be generated with a preset size.
  • the activation area (520) can include at least one nodule and a surrounding area, and can be generated with a preset size.
  • the activation area e.g., the first activation area (1115), the second activation area (1125)
  • the activation area can be generated with a size of 10 mm x 10 mm.
  • the activation area e.g., the first activation area (1115), the second activation area (1125)
  • the activation area can be generated with a size of 15 mm x 5 mm.
  • the activation area (e.g., the first activation area (1115), the second activation area (1125)) generated in this way can be a target for similarity determination.
  • the electronic device (130) may determine a similarity between at least one first nodule and at least one second nodule. For example, the electronic device (130) may acquire at least one second nodule detected from a second input image set acquired at at least one time point distinct from the first time point, and determine a similarity between the at least one first nodule and the at least one second nodule.
  • the electronic device (130) determines the similarity between at least one first nodule and at least one second nodule as described below with reference to FIGS. 12a, 12b and 12c.
  • the electronic device (130) may input the first input image set (1111) and the second input image set (1121) into a risk assessment model to obtain risk information for at least one first nodule. For example, if the electronic device (130) determines that at least one first nodule and at least one second nodule are the same nodule, the electronic device (130) may use the second input image set (1121) together to determine the risk for the at least one first nodule.
  • the electronic device (130) may determine the risk for the at least one first nodule based on first size information and first type information for the at least one first nodule detected based on the first input image set (1111), and change amount information for the at least one second nodule and the at least one first nodule detected based on the second input image set (1121).
  • the electronic device (130) may obtain risk information indicating risk information for the at least one first nodule by using at least one artificial intelligence model (e.g., the risk judgment model (1430) described with reference to FIG. 14).
  • the electronic device (130) when determining the risk of at least one first nodule detected from a first input image set (1111) photographed at a first point in time, the electronic device (130) can more accurately determine the risk of at least one first nodule by considering information (e.g., change amount information) based on at least one second nodule detected from a second input image set (1121) photographed at a different point in time (e.g., past point in time).
  • information e.g., change amount information
  • FIG. 12a is a flowchart illustrating an operation of an electronic device determining similarity according to various embodiments.
  • FIG. 12b is a diagram illustrating a method for an electronic device to determine similarity according to various embodiments.
  • FIG. 12c is a diagram illustrating a method for an electronic device to determine similarity according to various embodiments.
  • the operations described below may be performed in combination with one another.
  • the operations performed by the electronic device (130) may refer to operations performed by the processor (210) of the electronic device (130).
  • the operations illustrated in FIG. 12A may be performed in various orders, not limited to the order illustrated. Furthermore, according to various embodiments, more operations may be performed than those illustrated in FIG. 12A, or at least one operation may be performed less than those illustrated in FIG.
  • the electronic device (130) may generate a feature map of an activated area in operation 1210.
  • the electronic device (130) may generate a first feature map of a first activated area (1115) including at least one first nodule detected in a first input image set (1111).
  • the electronic device (130) may generate a second feature map of a second activated area (1125) including at least one second nodule detected in a second input image set (1121).
  • the electronic device (130) may determine the similarity between feature maps in operation 1230. For example, the electronic device (130) may determine the similarity between the first feature map and the second feature map.
  • the electronic device (130) can determine the similarity between at least one first nodule included in the first activation area (1115) and at least one second nodule included in the second activation area (1125) through the similarity judgment model (1201). In one embodiment, the electronic device (130) can determine the similarity between at least one first nodule and at least one second nodule by comparing at least one first activation area (1115) and at least one second activation area (1125), respectively.
  • the electronic device (130) may classify the activation areas determined to be similar to the activation areas of at least one first nodule among at least one second nodule into a similar nodule set (1203), and may classify the activation areas determined to be dissimilar to all activation areas included in at least one first nodule among at least one second nodule into a dissimilar nodule set (1205).
  • the electronic device (130) may classify the activation areas determined to be similar to the activation areas included in at least one second nodule among at least one first nodule into a similar nodule set (1203), and may classify the activation areas determined to be dissimilar to the activation areas included in at least one second nodule among at least one first nodule into a dissimilar nodule set (1205).
  • the electronic device (130) can generate a feature map of an activated area and determine the similarity between feature maps through a similarity judgment model (1201).
  • the feature map can be a set of information that emphasizes a specific feature of an image and can be extracted through a convolutional neural network.
  • the similarity judgment model (1201) can determine the similarity between the generated feature maps.
  • the similarity judgment model (1201) can generate feature maps for each of a first activated area (1115) and a second activated area (1125) and compare them to determine the similarity.
  • the method for determining the similarity is not limited to a specific method and can be performed in various ways, such as a statistical method, a distance-based method, or a machine learning method.
  • the similarity judgment model (1201) can determine the similarity between feature maps through geometric matching.
  • the electronic device (130) can determine the similarity between feature maps based on geometric matching through the similarity judgment model (1201). That is, the electronic device (130) can determine the similarity between the first activation area (1115) and the second activation area (1125) through geometric matching.
  • the geometric matching can determine the similarity between feature maps using various transformation models.
  • the similarity judgment model (1201) can determine the similarity between the activation areas by considering the similarity of at least one nodule included in the activation area as well as the similarity of the surrounding area through geometric matching.
  • the similarity judgment model (1201) can determine the similarity between a plurality of feature maps by geometrically transforming the feature map through a transformation model of the feature map.
  • the electronic device (130) may determine at least one nodule included in an activation area as the same nodule if the similarity is greater than or equal to a reference similarity. For example, the electronic device (130) may determine the at least one first nodule and the at least one second nodule as the same nodule if the similarity between a first activation area including at least one first nodule and a second activation area including at least one second nodule is greater than or equal to a preset value.
  • the electronic device (130) may consider change amount information confirmed through the at least one second nodule when determining the risk of the at least one first nodule.
  • FIG. 13 is a flowchart illustrating an operation of an electronic device according to various embodiments to obtain risk information of a nodule by considering whether or not it has a lung disease.
  • FIG. 14 is a diagram illustrating an electronic device according to various embodiments obtaining risk information based on a plurality of input image sets.
  • the operations described below may be performed in combination with one another.
  • the operations performed by the electronic device (130) may refer to operations performed by the processor (210) of the electronic device (130).
  • the operations illustrated in FIG. 13 may be performed in various orders, not limited to the order illustrated. Furthermore, according to various embodiments, more operations may be performed than those illustrated in FIG. 13, or at least one operation may be performed less than those illustrated in FIG.
  • the electronic device (130) can input a first input image set into a lung disease detection model to obtain lung disease information indicating whether the user has a lung-related disease.
  • the electronic device (130) may obtain risk information of the at least one first nodule based on the first size information, the first type information, the variation information, and the lung disease information in operation 1303.
  • the electronic device (130) may obtain risk information of the at least one first nodule based on the first size information of the at least one first nodule, the first type information of the at least one first nodule, the variation information based on the difference between the at least one first nodule and the at least one second nodule, and the lung disease information.
  • the electronic device (130) may acquire a first input image set (1411) captured at least a portion of the user's body (e.g., chest) at a first point in time.
  • the electronic device (130) may acquire a second input image set (1413) captured at least one point in time distinct from the first point in time (e.g., a point in time prior to the first point in time).
  • the electronic device (130) may input a first input image set (1411) into at least one artificial intelligence model (1420) to obtain first size information (1421) indicating a size of at least one first nodule, first type information (1423) indicating a type of the at least one first nodule, and lung disease information (1425).
  • the electronic device (130) may input a second input image set (1413) into at least one artificial intelligence model to obtain size information (1427) indicating a size of at least one second nodule, and second type information (1429) indicating a type of the at least one second nodule.
  • the electronic device (130) may input each of the first input image set (1411) and the second input image set (1413) into at least one artificial intelligence model (e.g., the nodule detection model (1113) of FIG. 11, the first artificial intelligence model (610) of FIG. 6) to obtain first size information (1421) and second size information (1427).
  • at least one artificial intelligence model e.g., the nodule detection model (1113) of FIG. 11, the first artificial intelligence model (610) of FIG. 6) to obtain first size information (1421) and second size information (1427).
  • the electronic device (130) may input each of the first input image set (1411) and the second input image set (1413) into at least one artificial intelligence model (e.g., the second artificial intelligence model (620) of FIG. 6) to obtain first type information (1423) and second size information (1429).
  • at least one artificial intelligence model e.g., the second artificial intelligence model (620) of FIG. 6
  • the electronic device (130) can obtain lung disease information (1425) by inputting a first input image set (1411) into at least one artificial intelligence model (e.g., a lung disease detection model described with reference to FIG. 3).
  • a lung disease detection model described with reference to FIG. 3
  • the electronic device (130) may obtain risk information (1440) of at least one first nodule based on the first size information (1421), the first type information (1423), the change amount information (1435) obtained based on the first size information (1421) and the second size information (1427) and/or the first type information (1423) and the second type information (1429), and the lung disease information (1425) through the risk judgment model (1430).
  • the electronic device (130) may obtain primary risk information (1431) for each of at least one first node being targeted, based on the first size information (1421)) and the first type information (1423)).
  • the electronic device (130) can obtain risk information (1440) by correcting the primary risk information (1431) based on the change information (1435) and the presence or absence of lung disease (1433).
  • the electronic device (130) can obtain the risk information (1440) using the risk judgment model (1430).
  • the electronic device (130) can obtain the primary risk information (1431) based on the first size information (1421) and the first type information (1423).
  • the electronic device (130) can obtain the risk information (1440) by correcting the primary risk information (1431) based on the change information (1435) and the presence or absence of lung disease (1433).
  • the at least one artificial intelligence model (1420) and the risk assessment model (1430) may be configured as a single artificial intelligence model.
  • the electronic device (130) may input a first input image set (1411) and a second input image set (1413) into the artificial intelligence model to obtain risk information (1440).
  • the risk information (1440) may be obtained through an operation similar to the operation described above.
  • the electronic device (130) can obtain the first input image set (1411) and the second input image set (1413) for the user through user registration information, etc., and input the first input image set (1411) and the second input image set (1413) into at least one artificial intelligence model to obtain risk information (1440) of at least one first nodule included in the first input image set (1411).
  • the electronic device (130) determines that at least one first nodule included in the first input image set (1411) and at least one second nodule included in the second input image set (1413) are the same nodule through a similarity judgment model (e.g., the similarity judgment model (1201) of 12b), the electronic device (130) can obtain risk information (1440) by using not only the first size information (1421) and the first type information (1423) for the at least one first nodule, but also the change amount information (1435) between the at least one second nodule and the at least one first nodule.
  • a similarity judgment model e.g., the similarity judgment model (1201) of 12b
  • the electronic device (130) may output risk information (1440) acquired based on a first input image set that captures at least a portion of the user's body. For example, the electronic device (130) may transmit the risk information (1440) to an external device (e.g., the user terminal (110) of FIG. 1) so that the risk information (1440) is output through a display of the external device. In addition, for example, the electronic device (130) may display the risk information (1440) through a display included in (or connected to) the electronic device (130).
  • an external device e.g., the user terminal (110) of FIG. 1
  • the electronic device (130) may display the risk information (1440) through a display included in (or connected to) the electronic device (130).
  • the electronic device (130) includes a communication circuit for transmitting and receiving data with an external device, at least one processor, and a memory for storing commands, wherein the commands are individually or collectively executed by the at least one processor so that the electronic device obtains a first input image set in which at least a part of the user's body is photographed at a first point in time, detects at least one first nodule with respect to the first input image set, determines a similarity between the at least one first nodule and at least one second nodule, and the at least one second nodule is obtained based on a second input image set stored in the memory, and when the similarity is equal to or greater than a preset value, inputs the first input image set and the second input image set into a risk assessment model to obtain risk information about the at least one first nodule, and the risk assessment model may be configured to obtain the risk information about the at least one first nodule based on first size information about the at least one first nodule, first type information about the at least one
  • the instructions are individually or collectively executed by the at least one processor to cause the electronic device to input the first set of input images into a nodule detection model to detect the at least one first nodule of at least a portion of the user's body, and to input the second set of input images into the nodule detection model to detect the at least one second nodule, wherein the nodule detection model may be an artificial intelligence model trained to detect at least one nodule included in at least a portion of the user's body based on the input set of images.
  • the instructions may be individually or collectively executed by the at least one processor to cause the electronic device to, through the nodule detection model, specify a first evaluation area in the first input image set, obtain first size information of at least one first nodule included in the first evaluation area, and classify the evaluation area as a first activation area if a volume of the at least one first nodule is greater than a cutoff value based on the first size information, and, through the nodule detection model, specify a second evaluation area in the second input image set, obtain second size information of at least one second nodule included in the second evaluation area, and classify the evaluation area as a second activation area if a volume of the at least one second nodule is greater than a cutoff value based on the second size information.
  • the instructions are individually or collectively executed by the at least one processor to cause the electronic device to generate a first feature map of the first activation area, generate a second feature map of the second activation area, and obtain a similarity between the first feature map and the second feature map through a similarity judgment model, wherein the similarity judgment model may be a model learned to determine a similarity based on activation areas including at least one nodule.
  • the similarity judgment model can determine the similarity between the first feature map and the second feature map through geometric matching.
  • the instructions may be individually or collectively executed by the at least one processor to cause the electronic device to input the first activation area into a type detection model to obtain the first type information of the at least one first module, and to input the second activation area into the type detection model to obtain the second type information of the at least one second module, wherein the type detection model may be an artificial intelligence model trained to output a type for at least one nodule.
  • the commands may be individually or collectively executed by the at least one processor to cause the electronic device to input the first input image set into a lung disease detection model to obtain lung disease information indicating whether the user has a lung-related disease
  • the lung disease detection model is an artificial intelligence model trained to receive an image including at least a part of the user's body and detect the presence or absence of lung disease, and to obtain risk information of the at least one first nodule based on the first size information, the first type information, the change amount information, and the lung disease information through the risk determination model.
  • the instructions may be individually or collectively executed by the at least one processor to cause the electronic device to obtain primary risk information of the at least one first nodule based on the first size information, the first type information, and the change amount information through the risk assessment model, and to obtain the risk information by correcting the primary risk information based on the lung disease information.
  • the second input image set may be a set of images acquired at at least one point in time that is distinct from the first point in time.
  • the instructions may be individually or collectively executed by the at least one processor to cause the electronic device to output the risk information through a display of the external device and to transmit the risk information to the external device.
  • the operating method of the electronic device may include an operation of obtaining a first input image set in which at least a part of a user's body is photographed at a first point in time, an operation of detecting at least one first nodule with respect to the first input image set, an operation of determining a similarity between the at least one first nodule and at least one second nodule, an operation of obtaining the at least one second nodule based on a second input image set stored in the electronic device, and, if the similarity is equal to or greater than a preset value, inputting the first input image set and the second input image set into a risk assessment model to obtain risk information for the at least one first nodule, wherein the risk assessment model may be an artificial intelligence model trained to output risk information for the at least one first nodule based on first size information for the at least one first nodule, first type information for the at least one first nodule, and variation information for the at least one first nodule and the at least one second nodule.
  • the operation of detecting the at least one first nodule includes an operation of inputting the first input image set to a nodule detection model to detect the at least one first nodule in at least a part of the body of the user
  • the operation of detecting the at least one second nodule includes an operation of inputting the second input image set to the nodule detection model to detect the at least one second nodule
  • the nodule detection model may be an artificial intelligence model trained to detect at least one nodule included in at least a part of the body of the user based on the input image set.
  • the operation of detecting the at least one first nodule may include: an operation of specifying a first evaluation area in the first input image set through the nodule detection model, obtaining first size information of at least one first nodule included in the first evaluation area, and an operation of classifying the evaluation area as a first activation area based on the first size information when a volume of the at least one first nodule is greater than a cutoff value; and the operation of detecting the at least one second nodule may include: an operation of specifying a second evaluation area in the second input image set through the nodule detection model, obtaining second size information of at least one second nodule included in the second evaluation area, and an operation of classifying the evaluation area as a second activation area based on the second size information when a volume of the at least one second nodule is greater than a cutoff value.
  • the operation of determining the similarity further includes an operation of generating a first feature map of the first activated area, an operation of generating a second feature map of the second activated area, and an operation of obtaining a similarity between the first feature map and the second feature map through a similarity determination model, wherein the similarity determination model may be an artificial intelligence model learned to determine the similarity based on activated areas including at least one nodule.
  • the method further includes an operation of inputting the first activation area into a type detection model to obtain the first type information of the at least one first nodule, and an operation of inputting the second activation area into the type detection model to obtain the second type information of the at least one second nodule, wherein the type detection model may be an artificial intelligence model trained to output a type for at least one nodule.
  • the operation of obtaining the risk information may further include an operation of inputting the first input image set into a lung disease detection model to obtain lung disease information indicating whether the user has a lung-related disease;
  • the lung disease detection model is an artificial intelligence model trained to receive an image including at least a part of the user's body and detect the presence or absence of lung disease; and an operation of obtaining risk information of the at least one first nodule based on the first size information, the first type information, the change amount information, and the lung disease information through the risk judgment model.
  • the operation of obtaining the risk information may further include an operation of obtaining primary risk information of the at least one first nodule based on the first size information, the first type information, and the change amount information through the risk judgment model, and an operation of obtaining the risk information by correcting the primary risk information based on the lung disease information.
  • the second input image set may be a set of images acquired at at least one point in time that is distinct from the first point in time.
  • the method of operating an electronic device may further include an operation of transmitting the risk information to an external device connected to the electronic device so as to output the risk information through a display of the external device.
  • the control method comprises the steps of: obtaining a first input image set in which at least a part of a user's body is photographed at a first point in time; detecting at least one first nodule with respect to the first input image set; determining a similarity between the at least one first nodule and at least one second nodule; wherein the at least one second nodule is obtained based on a second input image set stored in the electronic device, and when the similarity is equal to or greater than a preset value, inputting the first input image set and the second input image set into a risk assessment model to obtain risk information for the at least one first nodule, wherein the risk assessment model may be an artificial intelligence model trained to output risk information for the at least one first nodule based on first size information for the at least one first nodule, first type information for the at least one first nodule, and variation information for the at least one
  • each of the phrases “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.
  • part and module used in various embodiments of the present disclosure may include units implemented in hardware, software, or firmware. For example, they may be used interchangeably with terms such as logic, logic block, component, or circuit.
  • a module may be an integrally formed component or a minimum unit or part of the component that performs one or more functions.
  • the "part” and “module” used in various embodiments of the present disclosure may be stored in an addressable storage medium and implemented by various programs that can be executed by a processor.
  • Various embodiments of the present disclosure may be implemented as software (e.g., a program) including one or more commands stored in a memory (230) (e.g., built-in memory or external memory) readable by a device (e.g., an electronic device (130)).
  • the memory (230) may be represented as a storage medium.
  • the methods according to the various embodiments disclosed in this document may be provided as a computer program product.
  • the computer program product may be traded as a commodity between a seller and a buyer.
  • the computer program product may be distributed in the form of a device-readable storage medium (e.g., a compact disc read-only memory (CD-ROM)), or may be distributed online (e.g., downloaded or uploaded) through an application store or directly between two user devices.
  • a device-readable storage medium e.g., a compact disc read-only memory (CD-ROM)
  • CD-ROM compact disc read-only memory
  • each component e.g., a module or a program of the above-described components may include one or more entities, and some of the entities may be separated and placed in other components.
  • one or more components or operations of the above-described components may be omitted, or one or more other components or operations may be added.
  • a plurality of components e.g., a module or a program
  • the integrated component may perform one or more functions of each component of the plurality of components in a manner identical to or similar to that performed by the corresponding component among the plurality of components prior to the integration.
  • the operations performed by a module, program, or other component may be performed sequentially, in parallel, iteratively, or heuristically, or one or more of the operations may be performed in a different order, omitted, or one or more other operations may be added.
  • a computer-readable storage medium storing one or more programs (software modules) may be provided.
  • the one or more programs stored in the computer-readable storage medium are configured for execution by one or more processors within an electronic device.
  • the one or more programs include instructions that cause the electronic device to execute methods according to the embodiments described in the claims or specification of the present disclosure.
  • a function or operation performed by an electronic device may be performed by one or more processors executing one or more instructions stored in a memory.
  • the function or operation of the electronic device mentioned in the present disclosure may be performed by one processor executing one or more instructions, or may be performed by a combination of multiple processors executing one or more instructions.
  • the processor mentioned in the present disclosure may be understood to include a circuit for performing calculations or controlling other components of the electronic device.
  • the one or more processors may include at least one of a central processing unit (CPU), a graphics processing unit (GPU), a micro controller unit (MCU), a sensor hub, a supplementary processor, a communication processor, an application processor, an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a neural processing unit (NPU), a system on a chip (SoC), or an integrated circuit implemented to execute one or more instructions, and may have a plurality of cores.
  • CPU central processing unit
  • GPU graphics processing unit
  • MCU micro controller unit
  • FPGA field programmable gate array
  • NPU neural processing unit
  • SoC system on a chip
  • a program (software module, software) may be stored in a non-volatile memory including a random access memory (RAM), a flash memory, a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic disc storage device, a compact disc ROM (CD-ROM), a digital versatile disc (DVD) or other forms of optical storage devices, a magnetic cassette. Or, it may be stored in a memory formed by a combination of some or all of these.
  • the memory may be formed by a single storage medium or may be formed by a combination of a plurality of storage media.
  • the one or more commands may be stored in a single storage medium or may be distributed and stored in a plurality of storage media.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Public Health (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Epidemiology (AREA)
  • Primary Health Care (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Pathology (AREA)
  • Databases & Information Systems (AREA)
  • Radiology & Medical Imaging (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Geometry (AREA)
  • Biophysics (AREA)
  • High Energy & Nuclear Physics (AREA)
  • Optics & Photonics (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Molecular Biology (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • Veterinary Medicine (AREA)
  • Image Analysis (AREA)

Abstract

Un dispositif électronique selon divers modes de réalisation peut : obtenir un premier ensemble d'images d'entrée obtenu par capture d'une image d'au moins une partie du corps d'un utilisateur à un premier instant ; détecter au moins un premier nodule par rapport au premier ensemble d'images d'entrée ; déterminer une similarité entre le ou les premiers nodules et au moins un second nodule, le ou les seconds nodules étant obtenus sur la base d'un second ensemble d'images d'entrée stocké dans une mémoire ; et obtenir des informations de risque du ou des premiers nodules en entrant le premier ensemble d'images d'entrée et le second ensemble d'images d'entrée dans un modèle de détermination de risque lorsque la similarité a une valeur égale ou supérieure à une valeur préconfigurée, le modèle de détermination de risque étant un modèle d'intelligence artificielle entraîné pour délivrer des informations de risque du ou des premiers nodules sur la base de premières informations de taille du ou des premiers nodules, des informations de premier type du ou des premiers nodules, et des informations de variation du ou des premiers nodules et du ou des seconds nodules.
PCT/KR2025/003567 2024-03-20 2025-03-19 Dispositif électronique pour déterminer le risque de nodules sur la base d'une pluralité d'ensembles d'images d'entrée, procédé de fonctionnement associé et support d'enregistrement pour la mise en œuvre dudit procédé Pending WO2025198331A1 (fr)

Applications Claiming Priority (6)

Application Number Priority Date Filing Date Title
KR10-2024-0038236 2024-03-20
KR20240038236 2024-03-20
KR10-2024-0161636 2024-11-14
KR20240161636 2024-11-14
KR10-2024-0200282 2024-12-30
KR1020240200282A KR20250142180A (ko) 2024-03-20 2024-12-30 복수의 입력 이미지 세트에 기초하여 노듈의 위험도를 결정하는 전자 장치, 그 동작 방법 및 이 방법을 수행하기 위한 기록 매체

Publications (1)

Publication Number Publication Date
WO2025198331A1 true WO2025198331A1 (fr) 2025-09-25

Family

ID=97139759

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/KR2025/003567 Pending WO2025198331A1 (fr) 2024-03-20 2025-03-19 Dispositif électronique pour déterminer le risque de nodules sur la base d'une pluralité d'ensembles d'images d'entrée, procédé de fonctionnement associé et support d'enregistrement pour la mise en œuvre dudit procédé

Country Status (1)

Country Link
WO (1) WO2025198331A1 (fr)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20170004029A (ko) * 2010-12-08 2017-01-10 바이엘 헬스케어 엘엘씨 의료 영상 스캔으로부터 유래한 환자 방사선량 추정을 위한 적정 모델의 생성
KR102097742B1 (ko) * 2019-07-31 2020-04-06 주식회사 딥노이드 인공지능 기반의 의료영상 검색 시스템 및 그 구동방법
KR102274564B1 (ko) * 2018-07-03 2021-07-07 (주) 프로큐라티오 빅데이터분석기반 암진단장치
KR102391934B1 (ko) * 2022-01-19 2022-04-28 주식회사 엘티포 인공지능 기반 갑상선 결절의 암 위험도 진단 시스템 및 방법
KR20220113547A (ko) * 2020-01-03 2022-08-12 페이지.에이아이, 인크. 일반화된 질병 검출을 위해 전자 이미지들을 처리하기 위한 시스템들 및 방법들

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20170004029A (ko) * 2010-12-08 2017-01-10 바이엘 헬스케어 엘엘씨 의료 영상 스캔으로부터 유래한 환자 방사선량 추정을 위한 적정 모델의 생성
KR102274564B1 (ko) * 2018-07-03 2021-07-07 (주) 프로큐라티오 빅데이터분석기반 암진단장치
KR102097742B1 (ko) * 2019-07-31 2020-04-06 주식회사 딥노이드 인공지능 기반의 의료영상 검색 시스템 및 그 구동방법
KR20220113547A (ko) * 2020-01-03 2022-08-12 페이지.에이아이, 인크. 일반화된 질병 검출을 위해 전자 이미지들을 처리하기 위한 시스템들 및 방법들
KR102391934B1 (ko) * 2022-01-19 2022-04-28 주식회사 엘티포 인공지능 기반 갑상선 결절의 암 위험도 진단 시스템 및 방법

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
WANG CHAORAN, WANG GUANYUN, WANG WEI, KAN YING, ZHANG MINGYU, YANG JIGANG: "The role of 18F-FDG PET/CT metabolic parameters in the differential diagnosis of post-transplant lymphoproliferative disorder after pediatric liver transplantation", QUANTITATIVE IMAGING IN MEDICINE AND SURGERY, vol. 14, no. 2, 1 February 2024 (2024-02-01), pages 1323 - 1334, XP093357841, ISSN: 2223-4292, DOI: 10.21037/qims-23-1059 *

Similar Documents

Publication Publication Date Title
WO2021060899A1 (fr) Procédé d'apprentissage pour spécialiser un modèle d'intelligence artificielle dans une institution pour un déploiement et appareil pour l'apprentissage d'un modèle d'intelligence artificielle
WO2020242239A1 (fr) Système de prise en charge de diagnostic basé sur l'intelligence artificielle utilisant un algorithme d'apprentissage d'ensemble
WO2020235966A1 (fr) Dispositif et procédé de traitement d'une image médicale à l'aide de métadonnées prédites
WO2021006522A1 (fr) Appareil de diagnostic d'image utilisant un modèle d'apprentissage profond et son procédé
WO2021261808A1 (fr) Procédé permettant d'afficher un résultat de lecture de lésion
WO2019164232A1 (fr) Dispositif électronique, procédé de traitement d'image associé et support d'enregistrement lisible par ordinateur
WO2019083227A1 (fr) Procédé de traitement d'image médicale, et appareil de traitement d'image médicale mettant en œuvre le procédé
WO2022005090A1 (fr) Méthode et appareil de fourniture de résultat de diagnostic
WO2019235828A1 (fr) Système de diagnostic de maladie à deux faces et méthode associée
WO2021054518A1 (fr) Procédé, dispositif et programme logiciel pour diagnostiquer un cancer du col de l'utérus au moyen d'une analyse d'image médicale basée sur une technologie d'intelligence artificielle
WO2020076133A1 (fr) Dispositif d'évaluation de validité pour la détection de région cancéreuse
EP3008688A1 (fr) Procédé et appareil d'enregistrement d'images
WO2012134106A2 (fr) Procédé et appareil pour le stockage et l'affichage d'informations d'image médicale
WO2020162683A1 (fr) Dispositif de diagnostic à ultrasons fournissant un préréglage d'utilisateur et procédé pour son fonctionnement
WO2022131479A1 (fr) Procédé de diagnostic de lésions
EP3649460A1 (fr) Appareil pour optimiser l'inspection de l'extérieur d'un objet cible et procédé associé
WO2021107471A1 (fr) Procédé de récupération de données médicales
WO2022019616A1 (fr) Procédé de fourniture d'informations concernant le diagnostic d'un polype de la vésicule biliaire et dispositif de fourniture d'informations concernant le diagnostic d'un polype de la vésicule biliaire l'utilisant
WO2022173232A2 (fr) Procédé et système pour prédire le risque d'apparition d'une lésion
WO2020032560A2 (fr) Système et procédé de génération de résultats de diagnostic
WO2023128284A1 (fr) Procédé pour fournir des informations concernant le diagnostic du cancer du col de l'utérus et dispositif pour fournir des informations concernant le diagnostic du cancer du col de l'utérus à l'aide de celui-ci
WO2014208950A1 (fr) Procédé et appareil de gestion de données médicales
WO2021153858A1 (fr) Dispositif d'aide à l'identification à l'aide de données d'image de maladies cutanées atypiques
WO2023153569A1 (fr) Procédé d'analyse de l'état d'une articulation de genou et dispositif pour le réaliser
WO2025198331A1 (fr) Dispositif électronique pour déterminer le risque de nodules sur la base d'une pluralité d'ensembles d'images d'entrée, procédé de fonctionnement associé et support d'enregistrement pour la mise en œuvre dudit procédé

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 25774803

Country of ref document: EP

Kind code of ref document: A1