A method for classifying the urgency of acute abdominal conditions using artificial intelligence models.

An AI model integrates medical image and patient information to accurately predict and classify the urgency of acute abdominal conditions, addressing the challenges of inconsistent visual interpretation and delayed assessments in emergency settings.

JP2026182735APending Publication Date: 2026-11-19BEAMWORKS INC
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
JP2026023032
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Priority Date
2026-02-11
Filing Date
2026-02-16
Publication Date
2026-11-19

AI Technical Summary

Technical Problem

Existing emergency medical systems struggle to accurately and rapidly determine the necessity and urgency of emergency surgery for acute abdominal conditions due to reliance on visual interpretation, inconsistent judgments among medical personnel, and lack of integration of medical image analysis with patient clinical information, leading to delayed and inconsistent assessments.

Method used

An artificial intelligence model that integrates medical image analysis with patient information to detect abnormal findings, predict severity, and classify surgical necessity and urgency, using deep learning models to extract features and classify urgency levels based on both imaging and clinical data.

Benefits of technology

Enables rapid and consistent classification of emergency patients, improving decision-making efficiency by integrating severity prediction and patient clinical information in real-time, and providing a visual basis for medical professionals to make informed judgments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2026182735000001_ABST
    Figure 2026182735000001_ABST
Patent Text Reader

Abstract

This invention provides a method, apparatus, and program that comprehensively analyze medical images and patient information using an artificial intelligence model to detect abnormal findings related to acute abdominal diseases, quantitatively predict their severity, and automatically classify the patient's surgical necessity and urgency based on the prediction results. [Solution] A method for classifying the urgency of abdominal diseases using an artificial intelligence model executed by a computing device, comprising the steps of: acquiring medical images and patient information (S110); predicting the severity of the abdominal disease using the medical images (S120); and classifying the urgency (triage) based on the predicted severity and patient information (S130).
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to a technology for supporting decision-making related to emergency diagnosis and treatment, and more specifically, to a technology that analyzes medical images and patient clinical information using an artificial intelligence model, detects abnormal findings of acute abdominal diseases, and classifies the necessity of surgery and urgency level.

Background Art

[0002] In an emergency medical care environment, rapidly and accurately determining the necessity of emergency surgery for patients with Acute Abdomen has a significant impact on patient prognosis. Causative diseases of acute abdomen are widely distributed, ranging from mild conditions such as appendicitis to extremely urgent diseases such as Bowel Ischemia, Perforation, intra-abdominal hemorrhage, and Aortic Dissection, and their initial clinical symptoms are ambiguous, making differentiation often difficult even for skilled specialists.

[0003] In particular, medical images such as abdominal CT are used as an important diagnostic tool for confirming intra-abdominal abnormal findings (hemorrhage, bowel ischemia, intestinal perforation, free gas, free body fluid, organ damage, etc.). However, conventional emergency CT interpretation tends to rely heavily on the visual interpretation and experience of radiology specialists, and there is a limitation that interpretation delays or judgment discrepancies among medical personnel may occur during interpretation under conditions of night shifts or excessive workload.

[0004] In particular, for diseases such as bowel ischemia and intestinal perforation, even a diagnosis delay of several hours can lead to progression to intestinal necrosis and septic shock, resulting in a sharp increase in mortality. During time periods such as nights and weekends, or in environments with insufficient medical staff, it is difficult for radiology specialists to immediately perform interpretation and respond, which may cause obstacles in medical responses such as delays in determining the necessity of surgery or deciding on patient transfer.

[0005] Furthermore, existing artificial intelligence-based medical image analysis technologies often focus on individually detecting or classifying specific lesions or findings, limiting their ability to integrate with comprehensive assessments of the severity of acute abdominal diseases and the classification of the patient's surgical necessity and urgency, which are required in actual clinical settings. Moreover, while it is essential to consider not only medical image information but also patient clinical information such as age, vital signs, and level of consciousness when determining urgency, there is currently a lack of systematic technologies that can automatically integrate medical image analysis results and patient information for classification (triage).

[0006] For these reasons, in actual emergency room settings, CT scan results, patient clinical information, and the subjective judgments of healthcare professionals are often used individually without being coordinated. As a result, problems such as inconsistent or delayed assessments of the severity of acute abdominal diseases and the need for surgery are frequently pointed out.

[0007] Patent Document 1 discloses a vehicle-based emergency medical system and method for classifying and treating emergency patients and guiding their transport to a hospital. [Prior art documents] [Patent Documents]

[0008] [Patent Document 1] Korean Published Patent No. 10-2025-0083115 [Overview of the Initiative] [Problems that the invention aims to solve]

[0009] This disclosure aims to provide a method for detecting abnormal findings related to acute abdominal diseases, quantitatively predicting their severity, and automatically classifying the patient's surgical necessity and urgency based on the prediction results, by comprehensively analyzing medical images and patient information using an artificial intelligence model.

[0010] However, the technical challenges that this disclosure aims to solve are not limited to those mentioned above, and may include a variety of technical challenges that are obvious to an average engineer, as described below. [Means for solving the problem]

[0011] Based on one embodiment of the present disclosure for achieving the aforementioned objectives, a method for classifying the urgency of abdominal diseases using an artificial intelligence model executed by a computing device is disclosed. The method may include the steps of: acquiring medical images and patient information; predicting the severity of the abdominal disease using the medical images; and triaging the urgency based on the predicted severity and patient information.

[0012] In one embodiment, the patient information may include at least one of the following: age, sex, body mass index, mechanism of injury, blood pressure, heart rate, respiratory rate, oxygen saturation, or Glasgow Coma Scale (GCS).

[0013] In one embodiment, the step of predicting the severity of an abdominal disease using the above medical image may include the step of extracting abnormal finding feature information from the above medical image using the first artificial intelligence model, and predicting the severity based on the extracted abnormal finding feature information.

[0014] In one embodiment, the step of classifying the urgency based on the predicted severity and patient information may include the step of classifying the urgency based on the predicted severity and patient information using the second artificial intelligence model.

[0015] In one embodiment, the first artificial intelligence model includes a deep learning model trained to extract feature information from medical images, and the second artificial intelligence model may include a model trained to classify the urgency by integrating the predicted severity and patient information.

[0016] In one embodiment, the first artificial intelligence model may include a model trained using training medical images labeled with abnormal findings related to the abdominal disease, and the second artificial intelligence model may include a model trained using training patient clinical information labeled with surgical necessity and urgency grade, and severity extracted from the first artificial intelligence model.

[0017] In one embodiment, the abnormal findings related to the above abdominal disease may include at least one of the following: intraperitoneal hemorrhage, bowel ischemia, intestinal perforation, organ damage, free air, free fluid, vascular abnormalities, inflammatory changes, or mass.

[0018] In one embodiment, the method may further include the steps of: generating an abnormality estimation area map based on the predicted severity using the first artificial intelligence model; and generating an emergency image interpretation report using the medical image, patient information, urgency, and the abnormality estimation area map.

[0019] In one embodiment, the first artificial intelligence model may include an encoder for extracting abnormal finding feature information from the medical image; a classification module for calculating the severity based on the extracted abnormal finding feature information; and a computation module for generating an abnormal finding estimation region map based on the calculated severity.

[0020] In one embodiment, the step of generating an emergency image interpretation report using the above medical image, patient information, and abnormality estimation area map may include the step of overlaying the above abnormality estimation area map on the medical image as a heat map indicating information with colors and outlines; and the step of further displaying the above severity, above urgency, and surgical necessity.

[0021] In one embodiment, the method may comprise detecting a modification in patient information while the emergency interpretation report is displayed; reclassifying the urgency level based on the severity level and the modified patient information; and dynamically reflecting the reclassified urgency level on the report in response to the modification in the patient information.

[0022] In one embodiment, the step of predicting the severity of the abdominal disease using the medical image may further comprise segmenting an anatomical structure in the medical image; and weighting and analyzing the abnormal finding feature information in a region corresponding to the segmented anatomical structure using the first artificial intelligence model.

[0023] In one embodiment, the first artificial intelligence model may be configured to apply an Attention Mechanism to the segmentation result of the anatomical structure, and assign a weight to the region corresponding to the anatomical structure related to the abnormal finding for analysis.

[0024] In one embodiment, the medical image includes three-dimensional volume data composed of a plurality of slices, and the first artificial intelligence model may be configured to learn the Spatial Context of the three-dimensional volume data to detect the abnormal finding.

[0025] The step of classifying the urgency level based on the predicted severity level and the patient information may include a step of classifying into any one of a Surgical Emergency grade or a conservative treatment feasible grade.

[0026] According to an embodiment of the present disclosure for achieving the above-described problems, a computer program stored in a computer-readable storage medium is disclosed. When executed by one or more processors, the computer program causes the one or more processors to perform the following operations for classifying the urgency of abdominal diseases using an artificial intelligence model, where the operations include: an operation of acquiring medical images and patient information; an operation of predicting the severity of the abdominal disease using the medical images; and an operation of triaging the urgency based on the predicted severity and the patient information.

[0027] In one embodiment, the patient information may include at least any one of age, sex, body mass index, mechanism of injury, blood pressure, heart rate, respiratory rate, oxygen saturation, or Glasgow Coma Scale (GCS).

[0028] In one embodiment, the operation of predicting the severity of the abdominal disease using the medical images may include an operation of extracting abnormal finding feature information from the medical images using the first artificial intelligence model, and predicting the severity based on the extracted abnormal finding feature information.

[0029] In one embodiment, the operation of triaging the urgency based on the predicted severity and the patient information may include an operation of classifying the urgency based on the predicted severity and the patient information using the second artificial intelligence model.

[0030] In one embodiment, the first artificial intelligence model may include a deep learning model trained to extract feature information from medical images, and the second artificial intelligence model may include a model trained to integrate the predicted severity and the patient information to classify the urgency.

[0031] In one embodiment, the first artificial intelligence model may include a model trained using training medical images labeled with abnormal findings related to the abdominal disease, and the second artificial intelligence model may include a model trained using training patient clinical information labeled with surgical necessity and urgency grade, and severity extracted from the first artificial intelligence model.

[0032] In one embodiment, the above operation may further include the operation of generating an abnormality estimation area map based on the predicted severity using the first artificial intelligence model; and the operation of generating an emergency image interpretation report using the medical image, the patient information, the urgency, and the abnormality estimation area map.

[0033] In one embodiment, the first artificial intelligence model may include an encoder for extracting abnormal finding feature information from the medical image; a classification module for calculating the severity based on the extracted abnormal finding feature information; and a computation module for generating an abnormal finding estimation region map based on the calculated severity.

[0034] In one embodiment, the operation of generating an emergency image interpretation report using the above medical image, patient information, and abnormality estimation area map may include the operation of overlaying the above abnormality estimation area map on the medical image as a heat map that shows information with colors and outlines; and the operation of further displaying the above severity and the above urgency.

[0035] In one embodiment, the above operation may include: detecting modifications to patient information while the emergency medical imaging report is displayed; reclassifying the urgency based on the severity and the modified patient information; and dynamically reflecting the reclassified urgency information on the report in response to the modifications to the patient information.

[0036] A computing device based on one embodiment of the present disclosure is disclosed to solve the aforementioned problems. The device includes at least one processor and memory, the at least one processor may be configured to acquire medical images and patient information; predict the severity of the abdominal disease using the medical images; and triage the urgency based on the predicted severity and the patient information.

[0037] In one embodiment, the patient information may include at least one of the following: age, sex, body mass index, mechanism of injury, blood pressure, heart rate, respiratory rate, oxygen saturation, or Glasgow Coma Scale (GCS).

[0038] In one embodiment, the at least one processor may be configured to use the first artificial intelligence model to extract abnormal finding feature information from the medical image and to predict the severity based on the extracted abnormal finding feature information.

[0039] In one embodiment, the at least one processor may be configured to classify the urgency based on the predicted severity and patient information using the second artificial intelligence model.

[0040] In one embodiment, the first artificial intelligence model includes a deep learning model trained to extract feature information from medical images, and the second artificial intelligence model may include a model trained to classify the urgency by integrating the predicted severity and patient information.

[0041] In one embodiment, the first artificial intelligence model may include a model trained using training medical images labeled with abnormal findings related to the abdominal disease, and the second artificial intelligence model may include a model trained using training patient clinical information labeled with surgical necessity and urgency grade, and severity extracted from the first artificial intelligence model.

[0042] In one embodiment, the at least one processor may be further configured to generate an abnormality estimation area map based on the predicted severity using the first artificial intelligence model, and to generate an emergency image interpretation report using the medical image, the patient information, the urgency, and the abnormality estimation area map.

[0043] In one embodiment, the first artificial intelligence model may include an encoder for extracting abnormal finding feature information from the medical image; a classification module for calculating the severity based on the extracted abnormal finding feature information; and a computation module for generating an abnormal finding estimation region map based on the calculated severity.

[0044] In one embodiment, the at least one processor may be configured to overlay the abnormality estimation area map onto the medical image as a heatmap indicating information with color and outlines, and to further display the severity and urgency.

[0045] In one embodiment, the at least one processor may be further configured to detect modifications in patient information while the emergency radiology report is displayed; to reclassify the urgency based on the severity and the modified patient information; and to dynamically reflect the reclassified urgency on the report in response to the modifications in patient information. [Effects of the Invention]

[0046] According to this disclosure, by linking a first artificial intelligence model that detects abnormal findings related to acute abdominal diseases from medical images and predicts their severity with a second artificial intelligence model that classifies the need for surgery and urgency based on the predicted severity and patient information, it becomes possible to make urgency decisions that integrate severity prediction results and patient clinical information for various acute abdominal diseases in real time. This has the effect of enabling rapid and consistent classification (triage) of emergency patients and improving the efficiency of decision-making regarding surgery.

[0047] Furthermore, by dividing the location of anatomical structures and assigning weights to those regions for analysis, this disclosure can achieve high detection accuracy even for diseases where subtle imaging changes are important, such as intestinal ischemia and intestinal perforation, and can visualize the estimated abnormality area to provide medical professionals with an explainable basis for judgment.

[0048] Furthermore, according to this disclosure, by utilizing a foundation model pre-trained on large-scale medical images, it becomes possible to maintain consistent performance even in diverse medical environments with different imaging equipment and protocols (contrast-enhanced / non-contrast-enhanced).

[0049] On the other hand, the effects of this disclosure are not limited to those described above, and may include a variety of effects that are obvious to an ordinary engineer, as described below. [Brief explanation of the drawing]

[0050] [Figure 1] Figure 1 is a block diagram of a computing device for classifying the urgency of abdominal diseases using an artificial intelligence model, based on one embodiment of the present disclosure. [Figure 2] Figure 2 shows an example of the structure of an artificial intelligence platform model based on one embodiment of the present disclosure. [Figure 3] Figure 3 is a flowchart illustrating a method for classifying the urgency of abdominal diseases using an artificial intelligence model, based on one embodiment of the present disclosure. [Figure 4] Figure 4 is a schematic diagram illustrating the operation of classifying the urgency of abdominal diseases using an artificial intelligence model, based on one embodiment of the present disclosure. [Figure 5] Figure 5 is a schematic diagram showing the generation results of an emergency medical imaging report in one embodiment of the present disclosure. [Figure 6] Figure 6 is a schematic diagram of an exemplary computing environment that can embody the embodiments of this disclosure. [Modes for carrying out the invention]

[0051] Various embodiments will be described herewith reference to the drawings. Various descriptions are provided herein to facilitate understanding of this disclosure. However, these embodiments can certainly be carried out without these specific descriptions.

[0052] In this specification, terms such as “component,” “module,” and “system” refer to computer-related entities, hardware, firmware, software, combinations of software and hardware, or software execution. For example, a component may be, but is not limited to, a processing procedure executed on a processor, a processor, an object, an execution thread, a program, and / or a computer. For example, both an application running on a computing device and the computing device itself can be components. One or more components may reside in a processor and / or an execution thread, and one component may be localized within one computer or distributed across two or more computers. Such components may also be executed from a variety of computer-readable media, each containing a variety of data structures. Components may communicate locally and / or remotely, for example, by signals containing one or more data packets (e.g., data transmitted over a network such as the Internet, through data and / or signals from one component interacting with other components in a local system or distributed system).

[0053] The term "or" is used with the intention of meaning an implicational "or," not an exclusive "or." That is, unless specifically specified and contextually clear, "X uses A or B" means one of the natural implicational substitutions. In other words, if X uses A; X uses B; or X uses both A and B, then "X uses A or B" can be any of these. Furthermore, the terms "and / or" in this specification should be understood to refer to and include all possible combinations of one or more of the related items discussed.

[0054] Furthermore, the term “includes” as a predicate and / or as a modifier should be understood to mean that the feature and / or component in question exists. However, the term “includes” as a predicate and / or as a modifier should be understood not to exclude the existence or addition of one or more other further features, components and / or groups thereof. Also, where the number is not specifically identified or where it is not clear from the context to indicate a singular form, “singular” in this specification and claims should generally be interpreted to mean “one or more.”

[0055] Furthermore, the phrase "at least one of A or B" should be interpreted as meaning "including only A," "including only B," or "a combination of A and B."

[0056] Those skilled in the art should further recognize that the various exemplary logical blocks, configurations, modules, circuits, means, logic, and algorithmic stages described herein as relating to the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of both. To clearly illustrate the interoperability between hardware and software, various exemplary components, blocks, configurations, means, logic, modules, circuits, and stages have been generally described above in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the specific application and design constraints of the overall system. Skilled technicians can implement the described functionality in various ways for individual specific applications; however, decisions regarding such implementation should not be construed as departing from the scope of this disclosure.

[0057] The descriptions relating to the embodiments shown herein are provided so that a person with ordinary skill in the art of this disclosure may utilize or practice the invention. Various modifications to such embodiments are obvious to a person with ordinary skill in the art of this disclosure, and the general principles defined herein can be applied to other embodiments without departing from the scope of this disclosure. Therefore, this disclosure is not limited by the embodiments shown herein, but should be interpreted in the broadest sense consistent with the principles and novel features shown herein.

[0058] Figure 1 is a block diagram of a computing device for classifying the urgency of abdominal diseases using an artificial intelligence model, based on one embodiment of the present disclosure.

[0059] The configuration of the computing device (100) shown in Figure 1 is merely a simplified example. In one embodiment of this disclosure, the computing device (100) may include other configurations for implementing the computing environment of the computing device (100), and the computing device (100) may be composed of only some of the disclosed configurations.

[0060] The computing device (100) may include a processor (110), memory (130), and a network unit (150).

[0061] The processor (110) can consist of one or more cores and may include processors for data analysis and deep learning, such as a central processing unit (CPU), a general-purpose graphics processing unit (GPGPU), and a tensor processing unit (TPU) of a computing device. The processor (110) can read computer programs stored in memory (130) and execute data processing for machine learning in one embodiment of the present disclosure. Based on one embodiment of the present disclosure, the processor (110) can perform calculations for training a neural network. In deep learning (DL), the processor (110) can perform calculations for training a neural network, such as processing input data for training, extracting features from input data, calculating errors, and updating the weights of the neural network using backpropagation. At least one of the CPU, GPGPU, and TPU of the processor (110) can process the training of a network function. For example, both the CPU and GPGPU can train a network function or classify data using a network function. Furthermore, in one embodiment of this disclosure, the processors of multiple computing devices can be used together to train network functions and perform data classification using network functions. In addition, the computer program executed in the computing device in one embodiment of this disclosure may be a program that can be executed on a CPU, GPGPU, or TPU.

[0062] In one embodiment of the present disclosure, the memory (130) can store information in any form generated or determined by the processor (110) and information in any form received by the network unit (150).

[0063] In several embodiments of this disclosure, the memory (130) may include at least one type of storage medium from among flash memory type, hard disk type, multimedia card micro type, card type memory (e.g., SD or XD memory), RAM (Random Access Memory), SRAM (Static Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), PROM (Programmable Read-Only Memory), magnetic memory, magnetic disk, and optical disk. The computing device (100) may also operate in conjunction with web storage that performs the storage function of the memory (130) over the internet. The foregoing descriptions of memory are illustrative and this disclosure is not limited thereto.

[0064] In one embodiment of the present disclosure, the network unit (150) can use a variety of wired communication systems such as public switched telephone networks (PSTN), xDSL (x Digital Subscriber Line), RADSL (Rate Adaptive DSL), MDSL (Multi Rate DSL), VDSL (Very High Speed ​​DSL), UADSL (Universal Asymmetric DSL), HDSL (High Bit Rate DSL), and local area networks (LANs).

[0065] Furthermore, the network unit (150) in this specification can utilize a variety of wireless communication systems such as CDMA (Code Division Multi Access), TDMA (Time Division Multi Access), FDMA (Frequency Division Multi Access), OFDMA (Orthogonal Frequency Division Multi Access), SC-FDMA (Single Carrier-FDMA), and other systems.

[0066] The network (150) in this disclosure can be configured regardless of the mode of communication, such as wired or wireless, and can consist of various communication networks such as local area networks (LANs), personal area networks (PANs), and wide area networks (WANs). Furthermore, the aforementioned network can be the well-known World Wide Web (WWW), and it can also utilize wireless transmission technologies used for short-range communication, such as infrared data association (IrDA) or Bluetooth®.

[0067] The technologies described herein can be used not only in the aforementioned networks but also in other networks.

[0068] Figure 2 shows an example of the structure of an artificial intelligence platform model based on one embodiment of the present disclosure.

[0069] Throughout this specification, the terms artificial intelligence model, artificial intelligence infrastructure model, computational model, neural network, network function, and neural network can be used interchangeably.

[0070] A neural network can be constructed as a collection of interconnected computational units, generally called nodes. These nodes may also be referred to as neurons. A neural network consists of at least one node. The nodes (or neurons) that make up a neural network can be interconnected by one or more links.

[0071] In a neural network, one or more nodes connected via links can, relatively speaking, be in an input-output node relationship. The concepts of input and output nodes are relative; any node that is an output node to another node can be an input node in relation to another node, and vice versa. As mentioned above, the relationship between input and output nodes can be established around links. One input node can be connected to one or more output nodes via links, and vice versa.

[0072] In a relationship between input and output nodes connected via a single link, the output node's data can be determined based on the data input to the input node. In this case, the link connecting the input and output nodes can have weights. These weights can be variable, and can be adjusted according to the user or algorithm to perform the functions required by the neural network. For example, if one or more input nodes are interconnected to one output node by their respective links, the output node can determine its value based on the values ​​input to the input nodes connected to it and the weights set for the links corresponding to each input node.

[0073] As mentioned above, a neural network consists of one or more nodes interconnected via one or more links, forming an input-output node relationship within the neural network. In a neural network, the characteristics of the network can be determined by the number of nodes and links, the correlation between nodes and links, and the weight values ​​assigned to each link. For example, if there are two neural networks with the same number of nodes and links but different link weight values, these two neural networks can be recognized as different from each other.

[0074] A neural network can be constructed as a set of one or more nodes. A subset of the nodes that make up a neural network can form a layer. Some of the nodes that make up a neural network can form a layer based on their distance from the first input node. For example, a set of nodes that are n in distance from the first input node can form an n-th layer. The distance from the first input node can be defined based on the minimum number of links that must be traversed to reach that node from the first input node. However, this definition of a layer is arbitrary for the sake of explanation, and the position of a layer in a neural network can be defined in a way different from the above explanation. For example, the layer of nodes can be defined based on their distance from the final output node.

[0075] In one embodiment of the content of this disclosure, a collection of neurons or nodes can be defined using the terms "hierarchy" or "layer."

[0076] The first input node can refer to one or more nodes in a neural network that receive data directly without going through links in relation to other nodes. Alternatively, it can refer to a node in a neural network that does not have other input nodes connected via links in relation to other nodes based on links. Similarly, the final output node can refer to one or more nodes in a neural network that do not have an output node in relation to other nodes. Furthermore, a hidden node can refer to a node that does not fall under the category of first input node or final output node, but is still a node that makes up the neural network.

[0077] In one embodiment of this disclosure, the neural network may have the same number of nodes in the input layer as in the output layer, with the number of nodes decreasing once before increasing again as one progresses from the input layer to the hidden layer. In another embodiment of this disclosure, the neural network may have fewer nodes in the input layer than in the output layer, with the number of nodes increasing as one progresses from the input layer to the hidden layer. In yet another embodiment of this disclosure, the neural network may have more nodes in the input layer than in the output layer, with the number of nodes decreasing as one progresses from the input layer to the hidden layer. In yet another embodiment of this disclosure, the neural network may be a combination of the above-described neural networks.

[0078] An artificial intelligence platform model based on one embodiment of the disclosed material may include a deep neural network (DNN). A deep neural network can mean a neural network that includes multiple hidden layers in addition to the input and output layers. By using a deep neural network, it is possible to grasp the latent structures of data. That is, it is possible to grasp the latent structures of photographs, text, videos, audio, protein sequence structures, gene sequence structures, peptide sequence structures, music (for example, what is depicted in a photograph, what is the content and emotion of a text, what is the content and emotion of an audio recording, etc.), and / or the binding affinity between peptides and MHC. Deep neural networks can include convolutional neural networks (CNNs), recurrent neural networks (RNNs), autoencoders, restricted Boltzmann machines (RBMs), deep belief networks (DBNs), Q networks, U networks, Siam networks, generative adversarial networks (GANs), transformers, and the like. The deep neural networks mentioned above are merely examples, and this disclosure is not limited to them.

[0079] The artificial intelligence platform model described herein can be represented by any of the aforementioned network structures, including an input layer, a hidden layer, and an output layer.

[0080] The neural networks used in the artificial intelligence platform models described herein may be trained using at least one of the following methods: supervised learning, unsupervised learning, semi-supervised learning, transfer learning, active learning, or reinforcement learning. Learning a neural network can be the process of applying knowledge to the neural network to enable it to perform specific actions.

[0081] Neural networks can be trained to minimize the error in their output. During neural network training, training data is repeatedly input into the network, the network's output and target error for the training data are calculated, and the error is updated by backpropagating the network's error from the output layer to the input layer to reduce the error. In supervised learning, training data with the correct answer labeled is used (i.e., labeled training data), while in unsupervised learning, the correct answer may not be labeled for each training data point. For example, in supervised learning for data classification, the training data could be data with a category labeled for each data point. Labeled training data is input into the neural network, and the error can be calculated by comparing the neural network's output (category) with the labels on the training data. As another example, in unsupervised learning for data classification, the error can be calculated by comparing the input training data with the output of the neural network. The calculated error is backpropagated in the neural network in the reverse direction (i.e., from the output layer to the input layer), and through backpropagation, the connection weights of each node in each layer of the neural network can be updated. The amount of change in the connection weights of each node that are updated can be determined by the learning rate. The calculation of the neural network on the input data and the backpropagation of the error can constitute a learning cycle (epoch). The learning rate can be applied in a way that changes depending on the number of iterations of the neural network's learning cycle. For example, in the early stages of learning the neural network, a high learning rate can be used to increase efficiency by enabling the neural network to quickly achieve a certain level of performance, while in the later stages of learning, a low learning rate can be used to improve accuracy.

[0082] In neural network training, training data can generally be a subset of real-world data (i.e., data that the trained neural network intends to process). This can lead to a training cycle where errors on the training data decrease, but errors on the real-world data increase. Overfitting is the phenomenon where errors on real-world data increase due to excessive training on the training data. For example, a neural network that has learned to recognize cats by observing yellow cats may fail to recognize cats of other colors as cats; this is a type of overfitting. Overfitting can increase errors in machine learning algorithms. Various optimization methods can be used to prevent such overfitting. These methods include increasing the amount of training data, regularization, dropout (deactivating some of the network's nodes during training), and the use of a batch normalization layer.

[0083] This disclosure relates to artificial intelligence-based diagnostic support software that receives abdominal CT images and patient clinical information as input, automatically detects areas suspected of acute abdominal disease, and predicts and provides severity and urgency (Triage) grades to healthcare professionals. This disclosure can comprehensively analyze patient clinical information along with CT images taken in environments such as emergency rooms to quantitatively evaluate the severity of acute abdominal disease and automatically classify the patient's need for surgery and urgency.

[0084] On the other hand, this disclosure may also be implemented in the form of standalone software, which can receive abdominal CT images as input in accordance with the DICOM protocol, a standard medical imaging protocol, and transmit images displaying estimated abnormality areas, along with the results of the severity and urgency assessment of abdominal diseases, to a storage system or medical information system specified by the user. In this way, this disclosure can effectively support rapid emergency diagnosis and treatment decision-making by healthcare professionals by visualizing candidate abnormality areas in CT images taken in emergency rooms, etc., using contour lines and colors, and by providing information on the severity and triage classification of abdominal diseases.

[0085] Figure 3 is a flowchart showing a method for classifying the urgency of abdominal diseases using an artificial intelligence model, based on one embodiment of the present disclosure, and Figure 4 is a schematic diagram of the operation of classifying the urgency of abdominal diseases using an artificial intelligence model, based on one embodiment of the present disclosure. The method for classifying the urgency of abdominal diseases using an artificial intelligence model can be performed by a computing device (100).

[0086] On the other hand, the following description will focus on one embodiment for detecting abnormal findings related to acute abdominal diseases, quantitatively predicting their severity, and automatically classifying the patient's surgical necessity and urgency based on the prediction results. However, this disclosure is not limited to this, and the same or similar methods can be applied to detecting abnormal findings related to acute diseases or injuries occurring in areas other than the abdomen, such as the chest, head, pelvis, spine, or other body parts, and predicting their severity and urgency.

[0087] In this case, the medical image (1) can be an abdominal CT image suitable for determining the urgency of an acute abdominal disease. The medical image (1) described below may also be an abdominal CT image taken in an emergency setting and acquired by a computing device (100), and an artificial intelligence model may be applied to the abdominal CT image to analyze characteristic information related to abnormal findings in the abdominal disease and classify the severity. However, this is merely an example for the sake of explanation, and this disclosure is not limited thereto, and the same or similar methods may be applied to other medical images capable of analyzing the severity of abdominal diseases.

[0088] According to one embodiment of the present disclosure, a computing device (100) can acquire medical images (1) and patient information (2) (S110). For example, the medical image (1) is an image taken in an emergency medical environment and may include at least one of computed tomography (CT) images, magnetic resonance imaging (MRI) images, ultrasound images, or X-ray images. The medical image (1) may be received directly from imaging equipment in the emergency room or received via a network from a hospital information system (PACS) or a medical image storage server.

[0089] For example, a medical image (1) may include three-dimensional volume data composed of multiple slices, and a computing device (100) may reconstruct the multiple slices into a single continuous volume or process the medical image taking into account the spatial relationships between the slices. For instance, an abdominal CT image may include multiple cross-sectional slices taken sequentially along the abdominal region of the human body, and the slices may be composed of three-dimensional volume data arranged at regular intervals.

[0090] On the other hand, patient information (2) may include at least one of the following clinical information usable for urgency classification: age, sex, body mass index, mechanism of injury, blood pressure, heart rate, respiratory rate, oxygen saturation, or Glasgow Coma Scale (GCS). Patient information (2) may be obtained through an electronic medical record (EMR) system, an emergency medical information system, or a healthcare professional input interface. The computing device (100) can match and store medical images and patient information to correspond to the same patient, and can be used as input data for predicting the severity and classifying the urgency of abdominal disease in subsequent stages.

[0091] As another example, the patient information described above may include both standardized and atypical data. Standardized data may include at least one of the following: age, sex, body mass index, blood pressure, heart rate, respiratory rate, oxygen saturation, or Glasgow Coma Scale (GCS). Atypical data may include at least one of the following: emergency room initial visit record, nursing record, or chief complaint text.

[0092] According to one embodiment of the present disclosure, a computing device (100) can use a medical image (1) to predict the presence or absence of abnormal findings related to an acute abdominal disease and the severity of the abdominal disease (S120). The severity of the abdominal disease can mean an index calculated to reflect the presence or absence of an abdominal disease or abnormal findings observed in the patient, the degree of progression, or the possibility of deterioration to a clinically serious condition.

[0093] The severity of the above abdominal disease can be calculated by comprehensively considering imaging signs such as the type, extent, density, and distribution of abnormal findings identified on medical imaging, or the presence or absence of multiple abnormal findings. For example, a case where localized inflammatory findings are observed may be evaluated as having a different severity than a case where widespread organ damage or multiple pathological signs are observed simultaneously.

[0094] In one embodiment, the abnormal findings related to the acute abdominal disease may include at least one of the following: intra-abdominal hemorrhage, bowel ischemia, intestinal perforation, organ damage, free air, free fluid, vascular abnormalities, inflammatory changes, or mass. The computing device (100) can identify from medical images whether one or more of the above abnormal findings are present, and can predict the severity by analyzing the spatial characteristics and pathological severity of each abnormal finding.

[0095] For example, the severity of abdominal disease can be expressed in at least one of the following forms: a binary classification result indicating the presence or absence of abnormal findings, a multi-level classification divided into multiple risk levels, or a probability value indicating the possibility of disease or complication bleeding. Furthermore, the above-mentioned severity of abdominal disease can be used as an indicator applicable to a wide range of abdominal diseases, such as bleeding, inflammation, intestinal obstruction, and intestinal perforation, as input information for triage performed at a later stage, and can be provided as a reference indicator to assist the clinical judgment of healthcare professionals.

[0096] According to one embodiment of the present disclosure, the computing device (100) can preprocess a medical image (1) before predicting the presence or absence of abnormal findings related to acute abdominal disease and the severity of the abdominal disease. For example, the preprocessing may be a step to correct the quality and format of the image so that the medical image (1) is suitable for use as input to a first artificial intelligence model (10). For example, the computing device (100) can perform at least one of the following on the medical image (1): resolution adjustment, size normalization, contrast correction, noise reduction, or image alignment. For example, if the medical image (1) is an abdominal CT image, the computing device (100) can perform normalization on the pixel or voxel values ​​of the CT image, or windowing on it to make it suitable for identifying specific tissues or lesions. In the case of a CT image composed of multiple slices, the computing device (100) can correct the spacing between slices or perform time-series or spatial alignment to generate a consistent input data format. The pre-processed medical images are input into a first artificial intelligence model (10) to enable the effective extraction of image feature information related to the characteristics of abnormal findings in acute abdominal diseases, thereby improving the accuracy of predicting the severity of abdominal diseases.

[0097] According to one embodiment, a computing device (100) can use a first artificial intelligence model (10) to extract abnormal finding feature information (11) from medical images and predict the severity of an abdominal disease (111) based on the extracted abnormal finding feature information (11). The abnormal finding feature information (11) may include imaging signs that can be observed in medical images before a specific disease or pathological condition is clearly diagnosed. For example, the abnormal finding feature information (11) may include at least one of the following, or a combination thereof: localized changes in tissue density, abnormal morphological changes in organ or vascular boundaries, initial distribution patterns of free fluid, intestinal wall thickening, inflammatory fatty infiltration, and subtle heterogeneity of intratissue shadows.

[0098] The first artificial intelligence model (10) can be configured to automatically learn and extract the abnormal finding feature information (11) described above from the spatial patterns of medical images. The first artificial intelligence model (10) can take the abnormal finding feature information (11) as input and calculate the severity of abdominal disease (111) which reflects the presence or absence of abdominal disease or abnormal findings, the degree of progression, or the likelihood of deterioration to a clinically serious condition in the patient. The severity of abdominal disease (111) can be expressed as a continuous probability value, one of several risk levels, or as a classification result relating to the presence or absence of abdominal disease or abnormal findings.

[0099] In one embodiment, the first artificial intelligence model (10) may include a deep learning model trained to extract feature information from medical images. For example, the first artificial intelligence model may be based on, but is not limited to, a convolutional neural network (CNN), a transformer, a vision transformer (ViT), a foundation model, or a combination thereof. Exemplary, the first artificial intelligence model (10) may be based on a convolutional neural network (CNN), in which case it may be configured to receive preprocessed medical images as input and hierarchically extract local image features through a plurality of convolutional and pooling layers. For example, the first artificial intelligence model (10) can automatically learn features such as morphological changes in organ boundaries, local changes in tissue density, or abnormal shadow distributions through convolutional operations. Furthermore, the first artificial intelligence model (10) may be configured based on a Transformer or Vision Transformer (ViT) structure. In this case, the first artificial intelligence model (10) can effectively extract inter-organ relationships or complex pathological patterns across the entire image by dividing the medical image into multiple patch units and then learning the global correlations between each patch using an attention mechanism. The first artificial intelligence model (10) may also include a Foundation Model pre-trained with large-scale medical image data. The Foundation Model may be pre-trained with data specific to abdominal CT images or medical image data of various body parts, and then fine-tuned to suit the purposes of this disclosure. This enables the first artificial intelligence model (10) to stably extract abnormal finding feature information even in a limited training data environment.

[0100] According to one embodiment, the first artificial intelligence model (10) may include an encoder, a classification module, and a computation module. Such a configuration reflects a functional separation for calculating the severity of abnormal findings from medical images and providing the results in a visually interpretable format.

[0101] As an example, an encoder can extract abnormal finding feature information (11) from medical images. The encoder can analyze the spatial structure of medical images and convert core features associated with abdominal diseases or abnormal findings into low-dimensional or high-dimensional feature vector forms, and in this process it may be configured to reduce image noise and enhance the expressiveness of bleeding-related information.

[0102] The classification module can calculate the severity (111) of abdominal disease based on extracted abnormal finding feature information (11). The classification module may be configured to take abnormal finding feature information (11) as input and output a severity that reflects the presence or absence, progression, or likelihood of deterioration to a clinically serious state of abdominal disease or abnormal findings in the patient corresponding to the medical image. The above severity may be expressed in the form of a probability value, a multi-level grade, or a binary classification result.

[0103] The calculation module can generate an abnormality estimation area map based on the calculated severity of abdominal disease (111). The calculation module may be configured to allow healthcare professionals to intuitively recognize the location of suspected abnormalities by highlighting specific areas on medical images with color, brightness, or contour lines, using the location information of image features that contribute highly to severity prediction.

[0104] According to one embodiment of the present disclosure, the first artificial intelligence model (10) may include a model trained using training medical images labeled with abnormal findings related to abdominal diseases. The training medical images may be abdominal CT images, and each medical image may be provided with label information relating to the presence, severity, or stage of progression of the patient's abdominal disease or abnormal findings corresponding to the image.

[0105] For example, the label information may include binary labels to distinguish between cases where abnormal findings are present and cases where they are not present in the medical image. Furthermore, it may also include multi-level labels that reflect the severity, extent, or clinical risk level of the abnormal findings. The first artificial intelligence model (10) may be configured to learn the correlation between image features related to bleeding contained in the medical image and the severity of the abdominal disease by learning the correspondence between the label information and the medical image.

[0106] Furthermore, the medical images used for training may include multiple patient data collected from actual clinical settings. By including a variety of medical images with different shooting conditions, patient positions, image resolution, or presence or absence of contrast agent use, the first artificial intelligence model (10) can learn to reliably predict the severity of abdominal diseases even under different clinical conditions.

[0107] The first artificial intelligence model (10) can iteratively update its learning parameters during the learning process to minimize the error between abnormal finding feature information and label information extracted from medical images. This allows it to be configured to accurately calculate the severity of abdominal diseases for actual medical image inputs.

[0108] However, the first artificial intelligence model (10) is not limited to this, and may include a model trained using training medical images labeled with various body parts. For example, the body parts may include at least one of the following: heart, abdomen, chest, head, pelvis, or limbs, thereby allowing the first artificial intelligence model (10) to be trained to be applicable to medical images of various anatomical regions, rather than being limited to a specific body part.

[0109] According to one embodiment of the present disclosure, a computing device (100) can segment the locations of anatomical structures from a medical image. The computing device (100) can identify the boundaries of anatomical structures such as organs or blood vessels within a medical image and spatially divide the area occupied by each structure. The anatomical structures may include at least one of the following: appendix, gallbladder, small intestine, large intestine, aorta, liver, spleen, kidney, pancreas, stomach, or vascular structure. The computing device (100) can segment multiple anatomical structures into independent regions. For example, in an abdominal CT image, the liver, spleen, and kidney can be generated as segmentation masks that separate them from each other.

[0110] Furthermore, the computing device (100) can weight and analyze abnormal findings observed in regions corresponding to divided anatomical structures based on the location information of those regions. The computing device (100) can combine abnormal finding feature information extracted from medical images with the division results to assign relatively higher weights to abnormal findings observed in specific anatomical structures or adjacent regions. For example, changes in free fluid or tissue density observed in the liver or spleen region can be more significantly reflected in the severity prediction.

[0111] The first artificial intelligence model (10) may be configured to apply an attention mechanism based on the results of segmenting anatomical structures. The first artificial intelligence model (10) can utilize the positional information of the segmented anatomical structures to generate an attention map, thereby assigning higher attention weights to feature maps corresponding to anatomical regions where abnormal findings frequently occur (appear). The attention mechanism may be implemented as spatial attention, channel attention, or a combination thereof, thereby configuring the first artificial intelligence model (10) to analyze abnormal findings intensively, focusing on anatomical regions of high clinical importance, rather than analyzing the entire medical image uniformly.

[0112] According to one embodiment of the present disclosure, a medical image may include three-dimensional volume data composed of multiple slices. A computing device (100) can spatially align the multiple slices to form a single continuous three-dimensional volume. For example, an abdominal CT image may include numerous cross-sectional images taken at regular intervals between slices, and these cross-sectional images can be reconstructed as three-dimensional volume data containing anterior-posterior and posterior-inferior information of the internal structure of the human body. A first artificial intelligence model (10) may be configured to learn the spatial context of the three-dimensional volume data to detect abnormal findings. The first artificial intelligence model (10) can extract abnormal finding feature information by considering not only local image features limited to a single slice, but also three-dimensional contextual information such as the continuity between adjacent slices, the relative positional relationship between organs, and the spatial expansion pattern of the lesion. This makes it possible to more accurately detect minute lesions and abnormal findings near inter-organ boundaries that are difficult to identify by analyzing only a single cross-sectional image.

[0113] The first artificial intelligence model (10) may include a foundation model pre-trained using self-supervised learning with large-scale medical image data. The foundation model may be configured to learn structural patterns in medical images even without explicit label information. For example, the foundation model may be trained on 3D volume data through self-supervised learning tasks such as predicting relationships between slices, reconstructing specific volume regions, or learning spatial consistency. The foundation model may also include a model that has intrinsically learned the 3D anatomical structure of the human body and the spatial relationships between organs. The foundation model may be fine-tuned for abdominal CT images based on the pre-training results to make it suitable for detecting abnormal findings. This allows the first artificial intelligence model (10) to effectively detect abnormal findings related to abdominal diseases while maintaining stable performance even in environments with a limited number of labeled training data.

[0114] According to another embodiment of the present disclosure, a computing device (100) can identify one or more abnormal findings (Findings) from a medical image (1) using a first artificial intelligence model. The medical image may be an abdominal CT image taken in an emergency setting, and the computing device (100) can analyze the image features contained in the medical image to identify at least one of the following abnormal findings: intra-abdominal hemorrhage, pneumoperitoneum, bowel obstruction, bowel wall thickening, bowel perforation, acute appendicitis, acute diverticulitis, acute pancreatitis, ascites, and fat stranding. The computing device (100) can individually identify abnormal findings and determine whether multiple abnormal findings are present simultaneously, thereby generating basic information for distinguishing between traumatic lesions and non-traumatic diseases.

[0115] As another example, the computing device (100) can use medical images to identify abnormal findings related to acute abdominal diseases. The medical images may be abdominal CT images taken in an emergency setting, and the computing device (100) can analyze the image features contained in the medical images to identify the possibility of at least one of the following acute abdominal diseases: appendicitis, diverticulitis, or cholecystitis. The computing device (100) can also identify general pathological signs that appear in association with acute abdominal diseases from the medical images. For example, general pathological signs may include at least one of the following: inflamed fat, bowel wall thickening, ileus, free fluid, or free air.

[0116] The computing device (100) can generate basic information to distinguish between clinically significant abnormal findings and relatively minor nonspecific acute abdominal diseases using a combination of identification results for acute abdominal diseases and general pathological signs. Such classification results can be used as auxiliary information to more precisely predict the severity of abdominal diseases. Furthermore, identification results related to acute abdominal diseases and general pathological signs can be included as part of the abnormal finding characteristic information (11). Identification results related to acute abdominal diseases and general pathological signs can be used as reference information in the abdominal disease severity calculation stage performed by the first artificial intelligence model and in the surgical necessity and urgency classification stage performed by the second artificial intelligence model. This allows the computing device (100) to support more reliable clinical judgments by comprehensively reflecting multiple pathological cues, rather than being limited to a single image finding.

[0117] According to one embodiment of the present disclosure, a computing device (100) can generate an estimated abnormality region map based on the predicted severity of abdominal disease using a first artificial intelligence model. The first artificial intelligence model (10) may be configured to receive a preprocessed CT image as input and output an estimated abnormality region map that shows spatial regions in the medical image where abnormalities are suspected, along with a severity index indicating the severity of abdominal disease. Here, the severity of abdominal disease can be an index that quantifies the likelihood of an abdominal disease or pathological abnormality occurring in the patient corresponding to the medical image, in the form of a probability value, a continuous numerical value, or one of several risk levels.

[0118] For example, an abnormality estimation area map can be a map that spatially displays areas in a medical image where there is a relatively high probability of abnormal findings being present. The abnormality estimation area map can be generated by considering both the severity of the abdominal disease (111) calculated by the first artificial intelligence model (10) and the location information of the abnormality feature information (11).

[0119] For example, the computing device (100) can analyze the internal feature map or output results of the first artificial intelligence model (10) to identify pixels or voxel regions that contribute highly to severity prediction. The computing device (100) can also generate an abnormality estimation region map by mapping this to a medical image coordinate system. The abnormality estimation region map may be configured to emphasize regions where the severity of abdominal disease is relatively high. As an example, the abnormality estimation region map may be represented on a medical image using at least one of the following visual elements: color, brightness, transparency, or contour lines. Regions with a high probability of abnormality in the abnormality estimation region map may be displayed with relatively darker colors or clear contour lines. On the other hand, regions with a low probability of abnormality in the abnormality estimation region map may be displayed with lighter colors or blurred contour lines. The computing device (100) can support healthcare professionals in intuitively recognizing the location of abnormal findings related to abdominal diseases during the medical image interpretation process by displaying an abnormal finding estimation area map overlaid on the original medical image. The abnormal finding estimation area map can be used as a component of the emergency medical interpretation report generated in a later stage.

[0120] According to one embodiment of the present disclosure, a computing device (100) can triage the degree of surgical necessity and urgency based on the severity of the abdominal disease (111) and the patient information (2) (S130). The computing device (100) can determine the urgency or priority of the treatment required for the patient by comprehensively considering the severity of the abdominal disease calculated by medical image analysis and patient information indicating the patient's clinical condition. For example, the degree of surgical necessity may mean an index indicating whether surgery or invasive treatment (intervention) is necessary for an acute abdominal disease or abnormal findings thereon, and the degree of urgency. The above-mentioned degree of surgical necessity can be calculated by comprehensively considering the severity of the abdominal disease predicted by medical image analysis and the patient's clinical condition. The degree of surgical necessity can be classified into one of the following: a condition requiring immediate surgery or treatment, a condition in which surgical intervention should be considered within a short period of time, or a condition in which non-surgical treatment or observation may be prioritized. For example, if abnormal findings that are likely to worsen the prognosis if delayed, such as intestinal perforation, extensive intestinal ischemia, progressive organ injury, or significant pathological changes in the abdominal cavity, are identified, the need for surgery may be considered high. Furthermore, the need for surgery can be expressed as a binary judgment value (surgery necessary / not necessary), a graded value divided into multiple stages, or a continuous numerical value.

[0121] For example, the computing device (100) can classify a patient into either a Surgical Emergency grade or a conservative treatment grade. For instance, the computing device (100) may classify a patient as requiring surgery if the severity of the abdominal disease is relatively high, or if the severity is moderate but the patient's vital signs or level of consciousness is unstable. Conversely, if the severity of the abdominal disease is low and the patient's clinical condition is stable, the patient may be classified as conservative treatment. As an example, the computing device (100) may classify a patient as requiring a higher level of urgency if the severity of the abdominal disease is relatively high, or if the severity is moderate but the patient's vital signs or level of consciousness is unstable. Conversely, the computing device (100) may classify a patient as requiring a relatively lower level of urgency if the severity of the abdominal disease is low and the patient's clinical condition is stable. Thus, the level of urgency may be determined by reflecting not only the severity of the abdominal disease alone, but also the patient's overall clinical condition.

[0122] For example, the surgical necessity rating may include cases where immediate surgery or intervention is deemed necessary. The conservative treatment possibility rating may include cases where treatment is deemed possible through drug therapy such as antibiotics, observation, or other non-surgical treatment methods.

[0123] Furthermore, the degree of urgency can be subdivided and classified into multiple grades. For example, the degree of urgency may include Grade 1, which requires immediate surgery; Grade 2, which requires emergency treatment or aggressive treatment within a short period of time; and Grade 3, which may be treated conservatively or observed. However, the number of urgency grades and the classification criteria for each grade may vary depending on the medical environment, purpose of application, or clinical protocol, and this disclosure is not limited thereto.

[0124] The computing device (100) can support the determination of patient priorities and the formulation of treatment plans in emergency rooms and other settings by displaying the classified surgical necessity and urgency (211) on a screen in conjunction with medical images, the severity of abdominal diseases, and an estimated abnormality area map, or by including it as a component of an emergency image interpretation report and providing it to healthcare professionals. Such urgency classification results can be used as information to support the determination of patient priorities in the emergency room, the decision of whether or not to take additional images, the request for cooperation from specialists in diagnosis, or the formulation of treatment plans. The computing device (100) can display the classified surgical necessity and urgency as described above on a screen together with medical images, the severity of abdominal diseases, and an estimated abnormality area map. Furthermore, by including the surgical necessity and classified urgency as components of an emergency image interpretation report and providing it to healthcare professionals, the computing device (100) can support rapid diagnosis and treatment decision-making in emergency settings.

[0125] According to one embodiment, a computing device (100) can classify the need for surgery and urgency (211) based on the severity of abdominal disease (111) predicted using a second artificial intelligence model (20) and patient information (2). The second artificial intelligence model (20) may be configured to receive pre-processed patient clinical information (2) and abdominal disease severity feature information calculated by the first artificial intelligence model (10) as input and output a classification result indicating the urgency and need for surgery of the patient. The second artificial intelligence model (20) may be configured to analyze the coupling relationship between the severity of abdominal disease (111) and patient information (2) so that even if the severity is the same or similar, different levels of urgency (211) are classified depending on the patient's vital signs or level of consciousness. For example, even if the severity of abdominal disease is predicted to be moderate, if a decrease in the patient's blood pressure or a decrease in the Glasgow Coma Scale (GCS) is observed, the patient may be classified as an emergency with a high need for surgery.

[0126] In one embodiment, the second artificial intelligence model (20) may include a model trained on at least one of the following: a multi-layer perceptron (MLP), a transformer, a graph neural network (GNN), a large language model (LLM), or a combination thereof. The second artificial intelligence model (20) may be configured to non-linearly combine input data having different characteristics, such as the severity of abdominal disease (111) and patient information (2), and convert them into a representation suitable for emergency (211) classification. For this purpose, the second artificial intelligence model may include multiple input nodes, one or more hidden layers, and an output layer, the input nodes of which pre-processed patient clinical information and abdominal disease severity features calculated from the first artificial intelligence model (10) may be input, respectively. In the hidden layer, correlations and non-linear coupling relationships between the severity of abdominal disease (111) and patient information (2) may be learned.

[0127] The second artificial intelligence model (20) may be configured in its output layer to output either a patient urgency classification result or / or a surgical necessity classification result, the classification result of which may be expressed as one of several grades or as a continuous numerical value. However, the specific neural network structure, number of hidden layers, number of nodes, or learning parameters of the second artificial intelligence model (20) may be varied in any way that is sufficient to achieve the purposes of this disclosure, and this disclosure is not limited thereto.

[0128] According to one embodiment, the second artificial intelligence model (20) may include a model trained using training patient clinical information labeled with urgency grade and surgical necessity, and severity information of abdominal diseases extracted from the first artificial intelligence model (10). For example, training may be performed using supervised learning, for which the training data may be configured to include medical images, patient clinical information and corresponding severity labels and urgency and surgical necessity labels.

[0129] For example, each sample of the training data may be configured such that medical images collected from a specific patient correspond to the patient's clinical information, and the medical images may be associated with severity labels for the severity of abdominal disease. The patient's clinical information and severity information may be associated with correct labels related to urgency and the need for surgery. The second artificial intelligence model (20) may be configured to learn, using the training data configured as described above, a relationship in which appropriate urgency and the need for surgery are assigned depending on the combination of abdominal disease severity and patient's clinical information.

[0130] The second artificial intelligence model (20) can be implemented as a model trained to automatically classify the urgency and surgical necessity of a patient when actual patient data is input, by iteratively updating the learning parameters to minimize the error between the urgency classification result calculated as the output of the model during the learning process and the urgency grade label mentioned above.

[0131] According to another embodiment of the present disclosure, a computing device (100) can classify the urgency and surgical necessity, including the acuteness of the patient, based on the severity of the abdominal disease and patient information. The computing device (100) can calculate an urgency indicating the urgency of the procedure by comprehensively considering the severity of the abdominal disease predicted from medical images and the patient's clinical information. The acuteness can be used as an indicator to determine whether the patient's condition requires immediate treatment. The acuteness can be classified as low, not high, or high. For example, if the severity of the abdominal disease is low and the patient's vital signs are stable, it may be classified as low-acute, and if the severity of the abdominal disease is high or organ damage, organ dysfunction, or significant abnormal findings are identified, it may be classified as high-acute.

[0132] Furthermore, the computing device (100) can calculate the degree of acuteness by considering predictions regarding the presence and extent of damage to abdominal organs such as the liver, spleen, kidneys, and intestines, or the presence or absence of significant abnormal findings observed in the abdominal cavity. In this case, even the same type of injury may be classified into different degrees of acuteness depending on the extent of the injury, the type of organ affected, or the severity. The acuteness classification results may be used as detailed criteria for triage classification and may be included as information provided to healthcare professionals in the subsequent emergency radiology report generation stage. This allows healthcare professionals to intuitively grasp not only the type of injury or disease in the patient, but also the clinical severity of the condition and the need for immediate action.

[0133] Figure 5 is a schematic diagram showing the generation results of an emergency medical imaging report in one embodiment of the present disclosure.

[0134] According to one embodiment of the present disclosure, a computing device (100) can generate an emergency image interpretation report using medical images, patient information, severity of abdominal disease, surgical necessity, urgency, and an abnormality estimation area map. For example, the emergency image interpretation report may be configured to include, as a set of information to comprehensively provide the results of medical image analysis and the patient's clinical condition, the location of abnormal findings identified on the medical image, the severity of abdominal disease, the surgical necessity, and the urgency classified thereunder. The computing device (100) can display the abnormality estimation area map overlaid on the medical image and arrange the severity, surgical necessity, and urgency information linked to the image so that medical professionals can intuitively recognize the image interpretation results.

[0135] In one embodiment, the emergency imaging report may also display patient information (2) including at least one of the following: patient age, sex, body mass index, mechanism of injury, blood pressure, heart rate, respiratory rate, oxygen saturation, or Glasgow Coma Scale (GCS). This allows healthcare professionals to interpret the medical image analysis results in conjunction with the patient's clinical condition.

[0136] The computing device (100) can overlay an abnormality estimation area map onto a medical image as a heatmap that indicates information using color, brightness, or contour lines, and can be configured to visually emphasize areas where the severity of the abdominal disease contributes relatively more. The severity of the abdominal disease may also be expressed as a numerical value, probability value, or risk level, and the need for surgery and urgency may be included in the report as one of several levels or as a continuous value.

[0137] Furthermore, emergency radiology reports can be generated in at least one of the following formats: a report for screen display, a report in electronic document format, or a data format that can be transmitted to a medical information system. Emergency radiology reports can be used as reference material for assisting with radiology interpretation in the emergency room, determining patient priorities, and formulating additional tests or treatment plans. The computing device (100) can support diagnostic record management in the emergency medical environment and cooperation in medical treatment by storing the emergency radiology reports in a storage device or transmitting them to a hospital information system or a medical image storage or transmission system.

[0138] In one embodiment, referring to Figure 5, the computing device (100) can overlay an abnormality estimation area map (4) on the medical image (1) as a heat map indicating information with colors and outlines in the first area of ​​the emergency image interpretation report screen. For example, a medical image (1) such as an abdominal CT image may be displayed in the central area of ​​the screen. This allows healthcare professionals to intuitively recognize the location of suspected abnormalities during the medical image interpretation process.

[0139] Furthermore, the computing device (100) can display patient management information (3), including patient identification information and the date of diagnosis, in the second area of ​​the emergency radiology report screen. For example, the patient management information may include a patient identifier (ID), patient name, date of diagnosis, etc.

[0140] Furthermore, the computing device (100) can display patient information (2) in the third area of ​​the emergency medical imaging report screen. The area where patient information (2) is displayed may include and display emergency room vital signs information such as the patient's age, sex, body mass index (BMI), mechanism of injury, blood pressure, heart rate, respiratory rate, oxygen saturation, and Glasgow Coma Scale (GCS).

[0141] Furthermore, the computing device (100) can display the severity of the abdominal disease (111) predicted by the first artificial intelligence model in the fourth area of ​​the emergency medical imaging report screen. The severity of the abdominal disease (111) can be expressed as a numerical value, a probability value, or a risk indicator. The computing device (100) can also display a score (Acute Score) related to the severity of the abdominal disease in the fourth area.

[0142] Furthermore, the computing device (100) can display the surgical necessity and urgency (211) classified by the second artificial intelligence model in the fifth area of ​​the emergency radiology report screen. The urgency (211) may be expressed as one of several urgency levels or provided as a visual representation including color or icon, so that healthcare professionals can grasp at a glance the urgency of the treatment for the patient and whether surgery is required.

[0143] In this way, the computing device (100) can improve the efficiency of image interpretation by medical personnel in an emergency medical environment and support rapid and accurate decision-making by displaying medical images (1), abnormality estimation area maps (4), patient management information (3), patient information (2), and the severity (111) and urgency (211) of abdominal diseases in an integrated manner on a single screen.

[0144] According to one embodiment of the present disclosure, a computing device (100) can detect modifications to patient information while an emergency radiology report is displayed. Modifications to patient information may include changes to clinical information entered by healthcare professionals, such as changes to blood pressure, heart rate, respiratory rate, oxygen saturation, or Glasgow Coma Scale (GCS) values. The computing device (100) can detect modifications to patient information through an input interface on the emergency radiology report screen or through change events received from an external medical information system.

[0145] Furthermore, the computing device (100) can reclassify the need for surgery and urgency based on the predicted severity of the abdominal disease (111) and the corrected patient information. The computing device (100) can recalculate the need for surgery and urgency by reflecting the corrected patient clinical information using the second artificial intelligence model (20) as new input data, while maintaining the severity of the abdominal disease (111) predicted by the first artificial intelligence model (10). In this case, if the computing device (100) detects a change in the patient's clinical condition, such as a deterioration in vital signs or a change in the level of consciousness, it can reclassify the need for surgery or urgency as different from before, even if the severity remains the same.

[0146] Furthermore, the computing device (100) can dynamically reflect the reclassified surgical necessity and urgency information on the report in response to modifications in patient information. The computing device (100) can update and display the reclassified urgency on the emergency image interpretation report screen in real time. For example, the computing device (100) can enable healthcare professionals to immediately recognize the changed patient condition by changing the urgency grade, color display, or icon display. Such dynamic reflection can also be performed while maintaining the medical image and the abnormality estimation area map, allowing healthcare professionals to continuously check the urgency changes accompanying changes in patient clinical information within the existing image interpretation context.

[0147] On the other hand, the computing device (100) can effectively support the decision-making of healthcare professionals in an emergency medical environment, enabling them to respond quickly to changes in the patient's condition, by supporting the reclassification of surgical necessity and urgency in response to changes in patient information and the dynamic display of the results.

[0148] The steps described above may be divided into more steps or combined into fewer steps, depending on the example of the embodiment of this disclosure. Some steps may be omitted as needed, and the order of the steps may be changed.

[0149] A computer-readable medium storing a data structure according to one embodiment of the present disclosure is disclosed.

[0150] A data structure can refer to the organization, management, and storage of data that enables efficient access to and modification of that data. A data structure can also refer to the organization of data to solve a specific problem (e.g., data retrieval, data storage, data modification in the shortest possible time). A data structure can also be defined as the physical or logical relationships between data elements designed to support specific data processing functions. Logical relationships between data elements can include linking relationships between user-defined data elements. Physical relationships between data elements can include actual relationships between data elements physically stored in a computer-readable medium (e.g., permanent storage). Specifically, a data structure can include a collection of data, relationships between data, and functions or instructions that can be applied to data. A well-designed data structure allows computing devices to perform operations with minimal use of their resources. Specifically, computing devices can improve the efficiency of operations such as arithmetic, reading, insertion, deletion, comparison, exchange, and retrieval through well-designed data structures.

[0151] Data structures can be classified into linear and nonlinear data structures depending on their form. A linear data structure may be one in which only one data item is linked after another. Linear data structures can include lists, stacks, queues, and deques. A list can represent a set of data items that have an internal order. A list can include linked lists. A linked list may be a data structure in which data is linked in a linear fashion, with each item having a pointer. Pointers in a linked list may contain linking information to preceding and succeeding data. A linked list can be represented as a single linked list, a double linked list, or a circular linked list, depending on its form. A stack may be a data list structure in which data can be accessed in a restricted manner. A stack may be a linear data structure in which data can only be processed (e.g., inserted or deleted) from one end of the data structure. Data stored in a stack may be a last-in, first-out (LIFO) data structure. A queue is a data structure that allows restricted access to data, and unlike a stack, it can be a first-in, first-out (FIFO) data structure. A deck can be a data structure that allows data to be processed from both ends of the data structure.

[0152] A nonlinear data structure can be a structure in which multiple data are concatenated after a single data. A nonlinear data structure can include a graph data structure. A graph data structure can be defined as a vertex and an edge, where an edge can contain a line connecting two distinct vertices, and can include a graph data structure tree. A tree data structure can be a data structure in which there is only one path connecting two distinct vertices among the multiple vertices contained in the tree. In other words, it can be a data structure that does not form a loop in a graph data structure.

[0153] Throughout this specification, the terms artificial intelligence-based model, computational model, neural network, network function, and neural network may be used interchangeably. Hereafter, the term neural network will be used consistently. A data structure may include a neural network, and such a data structure may be stored in a computer-readable medium. The data structure may also include pre-processed data for processing by the neural network, data input to the neural network, neural network weights, neural network hyperparameters, data obtained from the neural network, activation functions associated with each node or layer of the neural network, and loss functions for learning the neural network. The data structure may include any of the components in the disclosed configuration. That is, the data structure may include all or any combination of the following: pre-processed data for processing by the neural network, data input to the neural network, neural network weights, neural network hyperparameters, data obtained from the neural network, activation functions associated with each node or layer of the neural network, and loss functions for learning the neural network. Beyond the configurations described above, data structures including neural networks can contain any other information that determines the properties of the neural network. Furthermore, data structures can contain, and are not limited to, all forms of data used or generated during the computational processes of the neural network. Computer-readable media can include computer-readable recording media and / or computer-readable transmission media. A neural network can be structured as a collection of interconnected computational units, which may generally be referred to as nodes. These nodes may also be referred to as neurons.A neural network consists of at least one node.

[0154] A data structure may include data to be input to a neural network. A data structure containing data to be input to a neural network may be stored on a computer-readable medium. Data to be input to a neural network may include training data input during the training process of the neural network and / or input data to be input to a neural network after training is complete. Data to be input to a neural network may include pre-processed data and / or data to be pre-processed. Pre-processing may include data processing processes to prepare data for input to a neural network. Therefore, a data structure may include data to be pre-processed and data generated by pre-processing. The data structures described above are merely examples, and this disclosure is not limited thereto.

[0155] The data structure may include the weights of a neural network (in this specification, weights and parameters may be used interchangeably). The data structure containing the weights of a neural network may be stored in a computer-readable medium. A neural network may have multiple weights. The weights may be variable and can be varied by the user or algorithm to enable the neural network to perform desired functions. For example, if one or more input nodes are interconnected to an output node by their respective links, the output node can determine the data value output from the output node based on the values ​​input to the input nodes connected to the output node and the weights set on the links corresponding to each input node. The data structures described above are merely examples, and this disclosure is not limited thereto.

[0156] As an unrestricted example, weights may include weights that are variable during the learning process of a neural network and / or weights after the neural network has finished learning. Weights that are variable during the learning process of a neural network may include weights at the start of a learning cycle and / or weights that are variable during a learning cycle. Weights that have finished learning a neural network may include weights after a learning cycle has finished. Thus, a data structure containing the weights of a neural network may include a data structure containing weights that are variable during the learning process of a neural network and / or weights after the neural network has finished learning. Therefore, the aforementioned weights and / or each combination of weights shall be included in the data structure containing the weights of a neural network. The aforementioned data structures are illustrative only, and this disclosure is not limited thereto.

[0157] A data structure containing neural network weights can be stored in a computer-readable medium (e.g., memory, hard disk) after undergoing a serialization process. Serialization may be a process of storing the data structure in the same or other computing device and later reconstructing it into a usable form. The computing device can serialize the data structure and send and receive the data over a network. The serialized data structure containing neural network weights can be reconstructed in the same or other computing device by deserialization. The data structure containing neural network weights is not limited to serialization. Furthermore, the data structure containing neural network weights may include data structures that enhance computational efficiency while minimizing the use of computing device resources (e.g., B-trees, tries, m-way search trees, AVL trees, Red-Black Trees in nonlinear data structures). The foregoing is merely illustrative, and this disclosure is not limited thereto.

[0158] The data structure may include the hyperparameters of the neural network. Furthermore, the data structure containing the neural network's hyperparameters may be stored in a computer-readable medium. The hyperparameters may be variable variables that can be changed by the user. Examples of hyperparameters may include the learning rate, cost function, number of training cycles, weight initialization (e.g., setting the range of weight values ​​to be initialized), and the number of Hidden Units (e.g., the number of hidden layers, the number of nodes in the hidden layers). The data structures described above are merely examples, and this disclosure is not limited thereto.

[0159] Figure 6 is a simplified and general schematic diagram of an exemplary computing environment in which embodiments of the present disclosure can be realized.

[0160] While it has been stated that this disclosure can generally be embodied by computing devices, those skilled in the art will understand that it can also be embodied in combination with computer executable instructions and / or other program modules that can be run on one or more computers, and / or as a combination of hardware and software. Generally, modules as used herein include routines, programs, components, data structures, etc., that perform a particular task or implement a particular abstract type of data. Furthermore, those skilled in the art will understand that the methods of this disclosure can be implemented by configurations of other computer systems, including single-processor or multi-processor computer systems, minicomputers, mainframe computers, as well as personal computers, handheld computing devices, microprocessor boards, or programmable consumer electronics, etc. (each of which can operate in conjunction with one or more associated devices).

[0161] The embodiments described herein can further be implemented in a distributed computing environment in which a task is performed by remote processing units connected via a communication network. In a distributed computing environment, program modules can reside in both local and remote memory storage devices.

[0162] Computers include a variety of computer-readable media. Any media accessible by a computer can be computer-readable, but such computer-readable media include volatile and non-volatile media, transient and non-transitory media, and portable and non-portable media. By example, rather than by limitation, computer-readable media may include computer-readable storage media and computer-readable transmission media. Computer-readable storage media include volatile and non-volatile media, transient and non-transitory media, portable and non-portable media implemented by any method or technique for storing information such as computer-readable instructions, data structures, program modules, or other data. Computer-readable storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, DVD (digital videodisk) or other optical disc storage devices, magnetic cassettes, magnetic tapes, magnetic disk storage devices or other magnetic storage devices, or any other media that can be accessed by a computer and used to store information.

[0163] Computer-readable transmission media typically include all information transmission media that implement computer-readable instructions, data structures, program modules, or other data on a modulated data signal, such as a carrier wave or other transmission mechanism. The term modulated data signal means a signal in which one or more of its characteristics are set or modified in order to encode information within the signal. By example, rather than by limitation, computer-readable transmission media include wired media such as wired networks or direct-wired connections, and wireless media such as acoustic, RF, infrared, and other wireless media. Any combination of any of the aforementioned media is also included in the scope of computer-readable transmission media.

[0164] An exemplary environment (1100) is shown that realizes various aspects of this disclosure, including a computer (1102), the computer (1102) including a processor (1104), system memory (1106), and a system bus (1108). The system bus (1108) connects system components, including (but not limited to) system memory (1106), to the processor (1104). The processor (1104) can be any processor from a variety of commercial processors. Dual processors and other multiprocessor architectures can also be used as the processor (1104).

[0165] The system bus (1108) can be one of several types of bus structures that can be further interconnected to a local bus using any of the following: a memory bus, a peripheral bus, and various commercial bus architectures. System memory (1106) includes read-only memory (ROM) (1110) and random access memory (RAM) (1112). The basic input / output system (BIOS) is stored in non-volatile memory (1110) such as ROM, EPROM, or EEPROM, and this BIOS includes basic routines that support the exchange of information between multiple components within the computer (1102) during startup, etc. RAM (1112) may also include high-speed RAM such as static RAM for caching data.

[0166] The computer (1102) also includes an internal hard disk drive (HDD) (1114) (e.g., EIDE, SATA)—this internal hard disk drive (1114) can also be configured for external use in a suitable chassis (not shown), a magnetic floppy disk drive (FDD) (1116) (e.g., for reading from and writing to a portable diskette (1118)), and an optical disk drive (1120) (e.g., for reading CD-ROM disks (1122) or for reading from and writing to other high-capacity optical media such as DVDs). The hard disk drive (1114), magnetic disk drive (1116), and optical disk drive (1120) can be connected to the system bus (1108) by a hard disk drive interface (1124), a magnetic disk drive interface (1126), and an optical drive interface (1128), respectively. The interface (1124) for implementing external drives includes, for example, at least one or both of the following: USB (Universal Serial Bus) or IEEE 1394 interface technology.

[0167] These drives and computer-readable media provide non-volatile storage of data, data structures, computer-executable instructions, and so on. In the case of a computer (1102), the drives and media correspond to storing any data in a suitable digital format. While the above description of computer-readable storage media refers to HDDs, portable magnetic disks, and portable optical media such as CDs or DVDs, those skilled in the art will understand that other types of computer-readable storage media, such as zip drives, magnetic cassettes, flash memory cards, cartridges, and so on, can also be used in exemplary operating environments, and that any of these media may contain computer-executable instructions for performing the methods of the present disclosure.

[0168] Numerous program modules, including an operating system (1130), one or more application programs (1132), other program modules (1134), and program data (1136), can be stored in the drive and RAM (1112). All or part of the operating system, applications, modules, and / or data can also be cached in RAM (1112). It will be understood that this disclosure can be implemented by various commercially available operating systems or combinations of operating systems.

[0169] The user can input commands and information to the computer (1102) through one or more wired or wireless input devices, such as a keyboard (1138) and a pointing device such as a mouse (1140). Other input devices (not shown in the diagram) may include a microphone, IR remote control, joystick, gamepad, stylus pen, touchscreen, and so on. These and other input devices are often connected to the processing unit (1104) through an input device interface (1142) connected to the system bus (1108), but they can also be connected through other interfaces such as parallel ports, IEEE1394 serial ports, game ports, USB ports, IR interfaces, and so on.

[0170] A monitor (1144) or other type of display device is also connected to the system bus (1108) through an interface such as a video adapter (1146). In addition to the monitor (1144), the computer generally includes other peripheral output devices such as speakers, printers, and so on (not shown in the illustration).

[0171] A computer (1102) can operate in a networked environment by utilizing logical connections to one or more remote computers (1148), such as multiple remote computers (1148), via wired and / or wireless communication. The multiple remote computers (1148) can be workstations, server computers, routers, personal computers, portable computers, microprocessor-based entertainment devices, peer devices, or other typical network nodes, and generally include many or all of the components described for a computer (1102), although for simplification only a memory storage device (1150) is illustrated. The illustrated logical connections include wired and wireless connections in a short-range network (LAN) (1152) and / or a larger network, such as a long-range network (WAN) (1154). Such LAN and WAN networking environments are common in offices and companies, facilitating enterprise-wide computer networks such as intranets, all of which can connect to global computer networks, such as the Internet.

[0172] When used in a LAN networking environment, the computer (1102) connects to the local network (1152) via a wired and / or wireless network interface, or via an adapter (1156). The adapter (1156) facilitates wired or wireless communication to the LAN (1152), which also includes a wireless access point installed therein to communicate with the wireless adapter (1156). When used in a WAN networking environment, the computer (1102) may include a modem (1158), connect to a communication server on the WAN (1154), or have other means of establishing communication through the WAN (1154), such as via the Internet. The modem (1158), which can be internal or external, and wired or wireless, connects to the system bus (1108) via a serial port interface (1142). In a networked environment, a program module or part thereof described for a computer (1102) can be stored in a remote memory / storage device (1150). While the illustrated network connection is illustrative, it is readily apparent that other means of establishing communication links between multiple computers may be used.

[0173] The computer (1102) operates to communicate with any wireless device or unit that is arranged and operates wirelessly, such as a printer, scanner, desktop and / or portable computer, PDA (portable data assistant), communication satellite, any equipment or location relating to a wirelessly discoverable tag, and a telephone. This includes at least Wi-Fi and Bluetooth® wireless technologies. Thus, the communication may be a predefined structure like a conventional network, or simply ad-hoc communication between at least two devices.

[0174] Wi-Fi (Wireless Fidelity) enables internet access and other connectivity without a wired connection. Wi-Fi is a wireless technology, similar to cell phones, that allows devices like computers to send and receive data indoors and outdoors, i.e., anywhere within the range of a base station. Wi-Fi networks use IEEE 802.11 (a, b, g, etc.) wireless technology to provide secure, reliable, and high-speed wireless connectivity. Wi-Fi can be used to connect computers to each other, to the internet, and to wired networks (using IEEE 802.3 or Ethernet). Wi-Fi networks can operate in unlicensed 2.4 and 5 GHz wireless bands at data rates such as 11 Mbps (802.11a) or 54 Mbps (802.11b), or in products that include both bands (dual band).

[0175] A person with ordinary skill in the art of this disclosure will understand that information and signals can be represented using any variety of different techniques and methods. For example, the data, instructions, commands, information, signals, bits, symbols and chips referenced in the foregoing description can be represented by voltage, current, electromagnetic waves, magnetic fields, etc. or particles, optical fields, etc. or particles, or any combination thereof.

[0176] A person with ordinary skill in the art of this disclosure will understand that the various exemplary logic blocks, modules, processors, means, circuits, and algorithmic stages described herein can be implemented by electronic hardware, various forms of programs or design code (referred to herein for convenience as “software”), or a combination of all of these. To illustrate this interoperability of hardware and software, various exemplary components, blocks, modules, circuits, and stages have been generally described above with respect to their functions. Whether such functions are implemented in hardware or software depends on the design constraints imposed on a particular application and the overall system. A person with ordinary skill in the art of this disclosure can implement the functions described in various ways for individual specific applications, but such decisions should not be construed as departing from the scope of this disclosure.

[0177] The various embodiments described herein can be realized by methods, apparatus, or manufactured articles using standard programming and / or engineering techniques. The term “manufactured article” includes computer programs, carriers, or media accessible from any computer-readable device. For example, computer-readable storage media include, but are not limited to, magnetic storage devices (e.g., hard disks, floppy disks, magnetic strips, etc.), optical disks (e.g., CDs, DVDs, etc.), smart cards, and flash memory devices (e.g., EEPROMs, cards, sticks, key drives, etc.). The various storage media described herein also include one or more devices and / or other machine-readable media for storing information.

[0178] It should be understood that the specific order or hierarchical structure of the multiple stages in the presented process is an example of an exemplary approach. It should be understood that, based on design priorities, the specific order or hierarchical structure of the stages in the process can be rearranged within the scope of this disclosure. The appended method claims provide a variety of stage elements in sample order, but are not limited to the specific order or hierarchical structure shown.

[0179] The descriptions relating to the embodiments provided herein are provided so that any person with ordinary skill in the art of the present disclosure may utilize or implement the present disclosure. Various variations of such embodiments are readily apparent to a person with ordinary skill in the art of the present disclosure, and the general principles defined herein can be applied to other embodiments without departing from the scope of the present disclosure. Accordingly, the present disclosure is not limited by the embodiments provided herein and should be interpreted in the broadest sense consistent with the principles and novel features provided herein.

Claims

1. A method for classifying the urgency of abdominal diseases using an artificial intelligence model executed by a computing device, The stage of acquiring medical images and patient information, A step of predicting the severity of the abdominal disease using the aforementioned medical images, This includes a step of classifying the severity of the emergency based on the predicted severity and the patient information, method.

2. In the method according to claim 1, The patient information includes at least one of the following: age, sex, body mass index, mechanism of injury, blood pressure, heart rate, respiratory rate, oxygen saturation, or Glasgow Coma Scale (GCS). method.

3. In the method according to claim 1, The step of predicting the severity of the abdominal disease using the medical images includes the step of using a first artificial intelligence model to extract abnormal finding feature information from the medical images and predicting the severity based on the extracted abnormal finding feature information. method.

4. In the method according to claim 3, The step of classifying the urgency based on the predicted severity and patient information includes the step of classifying the urgency based on the predicted severity and patient information using a second artificial intelligence model. method.

5. In the method according to claim 4, The first artificial intelligence model includes a deep learning model trained to extract feature information from medical images, The second artificial intelligence model includes a model trained to classify the urgency by integrating the predicted severity and the patient information. method.

6. In the method according to claim 4, The first artificial intelligence model includes a model trained using training medical images on which abnormal findings related to the abdominal disease are labeled, The second artificial intelligence model includes a model trained using patient clinical information for training, which is labeled with surgical necessity and urgency grades, and severity levels extracted from the first artificial intelligence model. method.

7. In the method according to claim 6, The abnormal findings related to the aforementioned abdominal disease include at least one of the following: intraperitoneal hemorrhage, bowel ischemia, bowel perforation, organ damage, free air, free fluid, vascular abnormalities, inflammatory changes, or mass. method.

8. In the method according to claim 3, The aforementioned method, The steps include generating an abnormality estimation area map based on the predicted severity using the first artificial intelligence model, The step of generating an emergency medical imaging report using the medical images, patient information, urgency level, and the abnormal finding estimation area map further includes: method.

9. In the method described in claim 8, The first artificial intelligence model described above is: An encoder for extracting abnormal finding feature information from the aforementioned medical image, A classification module that calculates the severity based on the extracted abnormal finding characteristic information, A computation module that generates an abnormality finding estimation area map based on the calculated severity, method.

10. In the method described in claim 8, The step of generating an emergency medical imaging report using the aforementioned medical images, patient information, and abnormal finding estimation area map is as follows: The steps include: overlaying the map of estimated abnormal areas on the medical image as a heat map that shows information using colors and outlines; The stage includes further indicating the severity and urgency of the above, method.

11. In the method described in claim 8, The aforementioned method, With the aforementioned emergency radiology report displayed, the process involves detecting corrections to the patient information, A step of reclassifying the urgency based on the severity and the corrected patient information, The step further includes dynamically reflecting the reclassified urgency level in the report in response to the correction in the patient information, method.

12. In the method according to claim 3, The step of predicting the severity of the abdominal disease using the aforementioned medical images is: The steps include segmenting anatomical structures in the aforementioned medical image, The process further includes the step of using the first artificial intelligence model to weight and analyze the abnormal finding feature information in the regions corresponding to the divided anatomical structures, method.

13. In the method according to claim 12, The first artificial intelligence model is configured to apply an attention mechanism to the segmentation results of the anatomical structures and to analyze the regions corresponding to the anatomical structures associated with the abnormal findings by assigning weights to them. method.

14. In the method according to claim 3, The aforementioned medical image includes three-dimensional volume data composed of multiple slices, The first artificial intelligence model is configured to learn the spatial context of the three-dimensional volume data and detect the anomaly. method.

15. In the method according to claim 1, The step of classifying the urgency based on the predicted severity and patient information includes the step of classifying it into either a Surgical Emergency grade or a grade that can be treated conservatively. method.

16. A computer program stored on a computer-readable storage medium, When the computer program is executed on one or more processors, it causes the one or more processors to perform the following operations for classifying the urgency of abdominal diseases using an artificial intelligence model: The aforementioned operation is, The operation of acquiring medical images and patient information, The operation of predicting the severity of the abdominal disease using the aforementioned medical images, This includes the operation of classifying the severity of the emergency based on the predicted severity and the patient information, A computer program stored on a computer-readable storage medium.

17. In the computer program described in claim 16, The operation of predicting the severity of the abdominal disease using the medical image includes the operation of extracting abnormal finding feature information from the medical image using a first artificial intelligence model, and predicting the severity based on the extracted abnormal finding feature information. A computer program stored on a computer-readable storage medium.

18. In the computer program described in claim 17, The operation of classifying the urgency based on the predicted severity and patient information includes the operation of classifying the urgency based on the predicted severity and patient information using a second artificial intelligence model. A computer program stored on a computer-readable storage medium.

19. In the computer program described in claim 18, The first artificial intelligence model includes a deep learning model trained to extract feature information from medical images, The second artificial intelligence model includes a model trained to classify the urgency by integrating the predicted severity and the patient information. A computer program stored on a computer-readable storage medium.

20. In the computer program described in claim 18, The first artificial intelligence model includes a model trained using training medical images on which abnormal findings related to the abdominal disease are labeled, The second artificial intelligence model includes a model trained using patient clinical information for training, which is labeled with surgical necessity and urgency grades, and severity levels extracted from the first artificial intelligence model. A computer program stored on a computer-readable storage medium.

21. In the computer program described in claim 17, The aforementioned operation is, Using the first artificial intelligence model, the operation of generating an abnormality finding estimation area map based on the predicted severity, The operation further includes generating an emergency medical imaging report using the medical images, patient information, urgency level, and the abnormal finding estimation area map, A computer program stored on a computer-readable storage medium.

22. In the computer program described in claim 21, The operation of generating an emergency medical imaging report using the aforementioned medical images, patient information, and abnormal finding estimation area map is as follows: The operation of overlaying the map of estimated abnormal findings onto the medical image as a heat map that shows information using colors and outlines, Includes an operation to further display the severity and urgency, A computer program stored on a computer-readable storage medium.

23. A computing device, It includes at least one processor and memory, The aforementioned at least one processor is We acquire medical images and patient information, Using the aforementioned medical images, the severity of abdominal disease is predicted. The system is configured to classify the severity of the emergency based on the predicted severity and the patient information. Device.

24. In the apparatus according to claim 23, The at least one processor is configured to use a first artificial intelligence model to extract abnormal finding feature information from the medical image and to predict the severity based on the extracted abnormal finding feature information. Device.

25. In the apparatus according to claim 24, The at least one processor is configured to classify the urgency based on the predicted severity and patient information using a second artificial intelligence model. Device.

26. In the apparatus described in claim 25, The first artificial intelligence model includes a deep learning model trained to extract feature information from medical images, The second artificial intelligence model includes a model trained to classify the urgency by integrating the predicted severity and the patient information. Device.

27. In the apparatus described in claim 25, The first artificial intelligence model includes a model trained using training medical images on which abnormal findings related to the abdominal disease are labeled, The second artificial intelligence model includes a model trained using patient clinical information for training, which is labeled with surgical necessity and urgency grades, and severity levels extracted from the first artificial intelligence model. Device.

28. In the apparatus according to claim 24, The aforementioned at least one processor is Using the first artificial intelligence model, an abnormality finding estimation area map is generated based on the predicted severity. The system is further configured to generate an emergency medical imaging report using the aforementioned medical images, patient information, urgency level, and abnormal finding estimation area map. Device.

29. In the apparatus according to claim 28, The aforementioned at least one processor is The abnormality estimation area map is overlaid on the medical image as a heat map that shows information using colors and outlines. The system is configured to further display the severity and urgency of the aforementioned conditions. Device.

Citation Information

Patent Citations

  • Emergency medical system and method for vehicles that guide emergency patient triage, treatment and hospital transfer

    KR1020250083115A