Apparatus, method, and system for diagnosing lower respiratory tract infections including juvenile pneumonia

Abstract

Disclosed are an apparatus, a method and a system for diagnosing lower respiratory tract infections including juvenile pneumonia, according to an embodiment. The apparatus for diagnosing lower respiratory tract infections including juvenile pneumonia, according to an embodiment, comprises: one or more processors; and a memory storing instructions executed by the one or more processors. The one or more processors inputs: at least one of medical questionnaire information and biometric index information of a patient to a pre-trained model; and outputs results of diagnosis of lower respiratory tract infections including pneumonia for the patient by using the pre-trained model.

Description

Device, method, and system for diagnosing lower respiratory tract infections, including pediatric pneumonia

[0001] The disclosed embodiments relate to devices, methods and systems for diagnosing lower respiratory tract infections, including pneumonia, in children and adolescents, and relate to techniques for diagnosing lower respiratory tract infections, including pneumonia, in children and adolescents.

[0002] [Cross-reference to related applications]

[0003] This application claims priority to Republic of Korea Patent Application No. 10-2023-0096739, filed July 25, 2023, the entire contents of which are incorporated herein by reference.

[0004] [Description of Nationally Supported Research and Development]

[0005] This study was supported by the Ministry of Science and ICT [MSRI Project Number: 08-2023-0006, Project Unique Number: 1711151151, Project Number: S0252-21-1001, Ministry: MSRI, Project Management (Specialized) Agency: National IT Industry Promotion Agency, Research Project Name: AI Precision Medical Solution (Doctor Answer 2.0) Development (R&D), Research Project Name: AI Precision Medical Solution (Doctor Answer 2.0) Development, Contribution Rate: 1 / 1, Project Implementing Agency: Seoul National University Bundang Hospital, Research Institution: 2021.04.01 ~ 2024.12.31].

[0006] Treatment for lower respiratory tract infections (LLIs), including pneumonia, varies depending on the causative agent. These agents include bacteria, viruses, and tuberculosis. Among these, lower respiratory tract infections caused by viruses do not require antibiotic treatment, while those caused by tuberculosis do not respond to standard antibiotics.

[0007] Thus, unnecessary antibiotic prescribing leads to resistance in patients. The resistance rate of Mycoplasma pneumoniae, the leading cause of community-acquired pneumonia in children and adolescents in Korea, to macrolide antibiotics has increased from 0% in 2000 to over 62.9% in 2011.

[0008] Moreover, although most lower respiratory tract infections in children and adolescents are caused by viruses, making antibiotics unnecessary, empirical antibiotics are often prescribed due to concerns about secondary bacterial infections. Furthermore, even with thorough examination and testing of patients in clinical settings, it is difficult to identify the causative microorganisms of lower respiratory tract infections in children and adolescents.

[0009] Therefore, for effective antibiotic prescription for patients with lower respiratory tract infections, including pediatric pneumonia with a high rate of viral infection, differentiation of the causative microorganism is of utmost importance.

[0010] The disclosed embodiments are directed to providing devices, methods and systems for diagnosing lower respiratory tract infections, including pneumonia, in children and adolescents.

[0011] According to one embodiment, a device for diagnosing lower respiratory tract infections, including pneumonia in children and adolescents, comprises one or more processors; and a memory for storing instructions executed by the one or more processors, wherein the one or more processors: input at least one of patient questionnaire information and biomarker information into a pre-learned model, and output a diagnosis result for lower respiratory tract infections, including pneumonia, for the patient using the pre-learned model.

[0012] The above questionnaire information may include personal information in text format, and the biomarker information may include basic test result information in text format performed at a primary hospital.

[0013] The above questionnaire information may include personal information values ​​for at least one of the patient's gender, age, height, weight, and time of hospital visit.

[0014] The biomarker information may include vital sign values ​​for at least one of the patient's respiratory rate, maximum body temperature, minimum body temperature, blood pressure, heart rate, body temperature, respiratory rate, heart rate, and systolic blood pressure.

[0015] The above biomarker information may include an underlying disease history for at least one of chronic pulmonary disease, cerebrovascular disease, neuromuscular disease, oncological disease, cardiovascular disease, nervous system disease, hematologic oncology disease, kidney disease, liver disease, other diseases, and immune disease.

[0016] The above biomarker information may include a vaccination history for at least one of a protein conjugate vaccine, a polysaccharide vaccine, a pneumococcal vaccine, and an influenza vaccine.

[0017] The above biomarker information may include lung sound results for at least one of rales, wheeze, breath sounds, and breath sound changes.

[0018] The biomarker information may include test results for at least one of hemoglobin, blood leukocytes, segmented neutrophils, C-reactive protein (CRP), blood urea nitrogen, serum sodium, glucose, aspartate aminotransferase (AST), and alanine aminotransferase (ALT).

[0019] The diagnosis result of the lower respiratory tract infection including the above pneumonia may include at least one of the presence or absence of a lower respiratory tract infection including pneumonia, probability, causative microorganism, predicted probability for the causative microorganism, infection finding factors, and normal finding factors.

[0020] The above model may include an XGBoost (Extreme Gradient Boost) decision tree or a random forest decision tree.

[0021] A method for diagnosing lower respiratory tract infections, including pneumonia in children and adolescents, according to one embodiment is a method performed by a device for diagnosing lower respiratory tract infections, including pneumonia in children and adolescents, having one or more processors; and a memory for storing a program executed by the one or more processors, the method comprising: a step of inputting at least one of patient questionnaire information and biomarker information into a pre-learned model; and a step of outputting a diagnosis result for lower respiratory tract infections, including pneumonia, for the patient using the pre-learned model.

[0022] The above questionnaire information may include personal information in text format, and the biomarker information may include basic test result information in text format performed at a primary hospital.

[0023] The above questionnaire information may include personal information values ​​for at least one of the patient's gender, age, height, weight, and time of hospital visit.

[0024] The biomarker information may include vital sign values ​​for at least one of the patient's respiratory rate, maximum body temperature, minimum body temperature, blood pressure, heart rate, body temperature, respiratory rate, heart rate, and systolic blood pressure.

[0025] The above biomarker information may include an underlying disease history for at least one of chronic pulmonary disease, cerebrovascular disease, neuromuscular disease, oncological disease, cardiovascular disease, nervous system disease, hematologic oncology disease, kidney disease, liver disease, other diseases, and immune disease.

[0026] The above biomarker information may include a vaccination history for at least one of a protein conjugate vaccine, a polysaccharide vaccine, a pneumococcal vaccine, and an influenza vaccine.

[0027] The above biomarker information may include lung sound results for at least one of rales, wheeze, breath sounds, and breath sound changes.

[0028] The biomarker information may include test results for at least one of hemoglobin, blood leukocytes, segmented neutrophils, C-reactive protein (CRP), blood urea nitrogen, serum sodium, glucose, aspartate aminotransferase (AST), and alanine aminotransferase (ALT).

[0029] The diagnosis result of the lower respiratory tract infection including the above pneumonia may include at least one of the presence or absence of a lower respiratory tract infection including pneumonia, probability, causative microorganism, predicted probability for the causative microorganism, and infection finding factors.

[0030] The above model may include an XGBoost (Extreme Gradient Boost) decision tree or a random forest decision tree.

[0031] According to one embodiment, a lower respiratory tract infection diagnosis system including pediatric pneumonia includes the above-described pediatric pneumonia diagnosis device; a terminal that inputs the patient's questionnaire information and the biomarker information into the pediatric pneumonia diagnosis device through a server; and a server that provides a pediatric pneumonia diagnosis service including pediatric pneumonia to the terminal using the lower respiratory tract infection diagnosis results including pneumonia output by the pediatric pneumonia diagnosis device.

[0032] The disclosed embodiments can diagnose lower respiratory tract infections, including pneumonia, based on questionnaire information and biomarker results of pediatric and adolescent patients through a learned model.

[0033] The disclosed embodiments can prevent unnecessary antibiotic prescriptions by identifying causative microorganisms in pediatric and adolescent patients through a learned model.

[0034] Figure 1 is a block diagram illustrating a lower respiratory tract infection diagnosis system including pediatric and adolescent pneumonia according to one embodiment.

[0035] Figure 2 is a block diagram illustrating a device for diagnosing lower respiratory tract infections including pediatric and adolescent pneumonia according to one embodiment.

[0036] Figure 3 is a graph showing the performance of a device for diagnosing lower respiratory tract infections, including pediatric and adolescent pneumonia, according to one embodiment.

[0037] Figure 4a is an example diagram showing the initial screen displayed on the user terminal.

[0038] Figure 4b is an example of a landing page displayed on a user terminal.

[0039] Figure 4c is an example of the first input screen displayed on the user terminal.

[0040] Figure 4d is an example of a second input screen displayed on a user terminal.

[0041] Figure 4e is an example of a third input screen displayed on a user terminal.

[0042] Figure 4f is an example of the first result screen displayed on the user terminal.

[0043] Figure 4g is an example of a second result screen displayed on a user terminal.

[0044] Figure 5 is a flowchart for explaining a method for diagnosing pediatric pneumonia according to one embodiment.

[0045] Hereinafter, specific embodiments of one embodiment will be described with reference to the drawings. The following detailed description is provided to facilitate a comprehensive understanding of the methods, devices, and / or systems described herein. However, this is merely an example and the present invention is not limited thereto.

[0046] In describing embodiments, if a detailed description of a known technology related to the present invention is judged to unnecessarily obscure the gist of an embodiment, the detailed description will be omitted. In addition, the terms described below are terms defined in consideration of their functions in the present invention, and this may vary depending on the intention or custom of the user or operator. Therefore, the definitions should be made based on the contents throughout this specification. The terminology used in the detailed description is only for the purpose of describing an embodiment and should never be limited. Unless clearly used otherwise, the singular form includes the plural form. In this description, expressions such as "comprises" or "having" are intended to indicate certain components, numbers, steps, operations, elements, parts, or combinations thereof, and should not be construed to exclude the presence or possibility of one or more other components, numbers, steps, operations, elements, parts, or combinations thereof other than those described.

[0047] Additionally, the embodiments described herein may have aspects that are entirely hardware, partially hardware and partially software, or entirely software. As used herein, "unit," "module," "device," "server," or "system" refer to hardware, a combination of hardware and software, or a computer-related entity such as software. For example, a unit, module, device, server, or system may refer to hardware that constitutes part or all of a platform and / or software such as an application for operating the hardware.

[0048] FIG. 1 illustrates a lower respiratory tract infection diagnosis system (100) including pediatric and adolescent pneumonia according to one embodiment.

[0049] Referring to FIG. 1, a lower respiratory tract infection diagnosis system (100) including pediatric and adolescent pneumonia according to one embodiment includes a user terminal (10), a lower respiratory tract infection diagnosis device (20) including pediatric and adolescent pneumonia, and a server (30).

[0050] The user terminal (10) is a commonly used device and may include all types of communication devices that can provide diagnosis services for lower respiratory tract infections, including pediatric and adolescent pneumonia, such as desktop computers, laptop computers, tablets, and smartphones.

[0051] Here, the user may be a medical professional who measured the test results at a primary hospital for auxiliary diagnosis.

[0052] The user terminal (10) can access a website or application provided by the server (30) and request user authentication. Specifically, the user terminal (10) can request user authentication from the server (30) using the user ID and password entered. The user terminal (10) can request membership registration from the server (30) based on data entered by the user. The user terminal (10) can request the server (30) to log out and invalidate a previously performed login. The user terminal (10) can request the server (30) to change its password. At this time, the password can be encrypted into a value that cannot be decrypted based on the bcrypt algorithm.

[0053] A user terminal (10) can request a lower respiratory tract infection diagnosis service, including pneumonia, from a server (30). Specifically, the user terminal (10) can input questionnaire information and biomarker information into a lower respiratory tract infection diagnosis device (20), including pediatric pneumonia, through the server (30) and receive a lower respiratory tract infection diagnosis result, including pneumonia, as an output value.

[0054] Meanwhile, in FIG. 1, each user terminal (10) is depicted as a single terminal, but this is exemplary and multiple user terminals (10) and administrator terminals (12) can be connected through a network, and are not necessarily limited thereto.

[0055] A device (20) for diagnosing lower respiratory tract infections, including pneumonia, in children and adolescents includes a pre-learned model (not shown) that diagnoses lower respiratory tract infections, including pneumonia, based on input values. Specifically, when questionnaire information and biomarker information are input from a terminal (10), the device (20) for diagnosing lower respiratory tract infections, including pneumonia, in children and adolescents uses a pre-learned model (not shown) to output a diagnosis result for lower respiratory tract infections, including pneumonia, for a patient.

[0056] Meanwhile, referring to FIG. 1, the lower respiratory tract infection diagnosis device (20) including pediatric and adolescent pneumonia is depicted as being physically separated from the server (30), but this is merely exemplary and, unlike the illustrated example, may be implemented by being included in the server (30), or, unlike the illustrated example, may not be clearly distinguished in specific operations.

[0057] The server (30) is a server (30) for providing a lower respiratory tract infection diagnosis service including pneumonia in children and adolescents. The server (30) is connected to a terminal (10) and a lower respiratory tract infection diagnosis device (20) including pneumonia in children and adolescents via a network to provide a lower respiratory tract infection diagnosis service including pneumonia. Here, the lower respiratory tract infection diagnosis service including pneumonia in children and adolescents refers to a service that provides diagnosis results for lower respiratory tract infections including pneumonia to pediatric patients in a non-face-to-face manner and assists medical professionals in diagnosing lower respiratory tract infections including pneumonia.

[0058] The server (30) can provide information corresponding to a request transmitted by the terminal (10). When the server (30) receives at least one of the patient questionnaire information and biomarker information entered by the terminal (10), it transmits the information to a lower respiratory tract infection diagnosis device (20) including pediatric pneumonia, and transmits the result value output by the lower respiratory tract infection diagnosis device (20) including pediatric pneumonia to the terminal (10) through a website or application.

[0059] Figure 2 is a block diagram illustrating a device (20) for diagnosing lower respiratory tract infections including pediatric and adolescent pneumonia.

[0060] Referring to Fig. 2, a device (20) for diagnosing lower respiratory tract infections including pediatric and adolescent pneumonia includes an input unit (21) and an output unit (22).

[0061] The input unit (21) and the output unit (22) may be implemented using one or more physically separated devices, or may be implemented by a combination of one or more processors or one or more processors and software, and may not be clearly distinguished in specific operations, unlike the illustrated example.

[0062] The input unit (21) inputs at least one of the patient's questionnaire information and biomarker information into a pre-learned model (not shown).

[0063] The questionnaire information may include personal information in text format. For example, the questionnaire information may include at least one of the following personal information values: the patient's name, gender, date of birth, height, weight (BMI), and the time of hospital visit.

[0064] Biomarker information may be text-based basic test results performed in a primary care hospital.

[0065] The biomarker information may include vital sign values ​​for at least one of current body temperature, respiratory rate, heart rate, systolic blood pressure, presence of cough, highest body temperature, lowest body temperature, coughing temperament, number of coughs, fever onset date, cough onset date, fever duration, cough duration, blood pressure, and heart rate.

[0066] The biomarker information may include a Review of System (ROS) value for at least one of fever, fever duration, cough, cough duration, sputum, rhinorrhea, nasal stuffiness, oral intake decrease, sore throat, chest pain, dyspnea, apnea, irritability, lethargy, headache, vomiting, diarrhea, joint pain, myalgia, urinary symptoms, rash, and

[0067] Biomarker information may include physical examination values ​​for at least one of chest retraction, rash, stridor, and cyanosis.

[0068] The biomarker information may include an underlying disease history of at least one of chronic pulmonary disease, cerebrovascular disease, neuromuscular disease, oncological disease, cardiovascular disease, nervous system disease, hematologic oncology disease, kidney disease, liver disease, other diseases, and immune disease.

[0069] The biomarker information may include a vaccination history for at least one of a protein conjugate vaccine, a polysaccharide vaccine, a pneumococcal vaccine, and an influenza vaccine.

[0070] Biomarker information may include lung sound results for at least one of crackles, wheeze, breath sounds, and breath sound changes (e.g., increase, decrease).

[0071] Biomarker information may include test results for at least one of hemoglobin (Hb), white blood cells (WBC), neutrophils, lymphocytes, fosinopril, platelets, C-reactive protein (CRP), procalcitonin (PCT), blood urea nitrogen (BUN), glucose, aspartate aminotransferase (AST), and alanine aminotransferase (ALT).

[0072] The output unit (22) outputs the diagnosis results for lower respiratory tract infection, including pneumonia, for the patient using a pre-learned model (not shown).

[0073] At this time, the diagnosis result of lower respiratory tract infection including pneumonia may include at least one of the following: presence or absence of pneumonia, probability of pneumonia, causative microorganism of lower respiratory tract infection including pneumonia, predicted probability of causative microorganism, lower respiratory tract infection including pneumonia finding factors, and normal finding factors.

[0074] Specifically, the output unit (22) can output the probability of lower respiratory tract infection, including pneumonia, as a continuous value using a pre-trained model (not shown). The output unit (22) can classify and output the causative microorganisms of lower respiratory tract infection, including pneumonia, using the pre-trained model (not shown). The output unit (22) can output the probability that lower respiratory tract infection, including pneumonia, is infected by a viral causative microorganism as a predicted probability for the causative microorganism using the pre-trained model (not shown). The output unit (22) can output the probability that lower respiratory tract infection, including pneumonia, is infected by a bacterial causative microorganism as a predicted probability for the causative microorganism using the pre-trained model (not shown). The output unit (22) can output the upper factors contributing to the findings of lower respiratory tract infection, including pneumonia, using the pre-trained model (not shown). The output unit (22) can output the upper normal findings factors contributing to the normal findings using the pre-trained model (not shown).

[0075] The output unit (22) can output the importance of factors diagnosed as lower respiratory tract infection including pneumonia or normal using a pre-learned model (not shown).

[0076] The pre-trained model (not shown) can be trained using a decision tree-based algorithm. For example, the pre-trained model (not shown) can be trained based on an XGBoost (Extreme Gradient Boost) decision tree or a random forest decision tree.

[0077] A pre-trained model (not shown) can output pneumonia diagnosis results by applying variable importance to input items. For example, the pre-trained model (not shown) can apply variable importance, as shown in [Table 1], to at least one feature included in each input item.

[0078] Ranking Characteristic Name Importance 1 Cough 0.58642 Rashes 0.11083 Reduced respiratory sounds 0.03274 Abnormal respiratory sounds 0.02165 Phlegm 0.01566 Wheezing 0.01307 Rash 0.01258 Stridor 0.00999 Immunity 0.009210 Abnormal respiratory rate 0.0089

[0079] The terminal (10) can input the characteristics of input items in the same manner as in [Table 2].

[0080] Characteristic name input method Cough Binary - When the symptom is present: 1, when it is absent: 0 Runny nose Binary - When discovered after observation: 1, when it is absent: 0 No chronic lung disease Binary - When none of the symptoms for asthma, interstitial lung disease, chronic obstructive pulmonary disease, or other chronic lung diseases are applicable, 1, when any of them are applicable, 0 Alanine aminotransferase (ALT) Alanine aminotransferase (ALT) measurement value (continuous value) Fever Binary - When the symptom is present: 1, when it is absent: 0 C-reactive protein plasma measurement value (continuous value) Runny nose Binary - When the symptom is present: 1, when it is absent: 0 Respiratory rate Respiration rate per minute (continuous value) Dyspnea Binary - When the symptom is present: 1, when it is absent: 0 Neutrophils Neutrophil measurement value (%) Other chronic lung disease Binary - When applicable: 1, when it does not apply: 0 Chronic Lung disease / Chronic obstructive pulmonary diseaseBinary - Applicable: 1, if not applicable: 0Reduced breath soundsBinary - Discovered after observation: 1, if not applicable: 0Gender0: Male 1: FemaleChest painBinary - With the symptom: 1, if not applicable: 0Neoplastic disease / CancerBinary - Applicable: 1, if not applicable: 0No immunosuppressionBinary - Not using steroids or other immunosuppressants: 1, Other: 0Blood - White blood cellsBlood white blood cell measurement (continuous value)WheezingBinary - Discovered after observation: 1, if not applicable: 0Other diseases (heart failure, cirrhosis, chronic kidney disease) NoneBinary - If none of cirrhosis, heart failure, and chronic kidney disease apply, 1, if any of them apply, 0Heart rateHeart rate value (continuous value)No neoplastic diseaseBinary - Applies to all solid tumors and blood cancers among neoplastic diseases If not: 1, Other: 0 BMI Continuous value, kg / m^2 Age Age on the date of visit (integer value) Pneumococcal vaccine (PCV / PPSV23) Binary - If vaccinated: 1, If ​​not: 0 Other diseases / heart failure Binary - If applicable: 1,If not applicable: 0Aspartate aminotransferase (AST)Aspartate aminotransferase (AST) measurement (continuous)HemoglobinHemoglobin measurement (continuous)Immunosuppression / ImmunosuppressionBinary - If applicable: 1, if not applicable: 0HeadacheBinary - If present: 1, if not present: 0Immunosuppression / SteroidsBinary - If present: 1, if not present: 0HerothermiaHighest body temperature within 48 hours (continuous)Systolic blood pressureSystolic blood pressure (continuous)Muscle painBinary - If present: 1, if not present: 0DiarrheaBinary - If present: 1, if not present: 0VomitingBinary - If present: 1, if not present: 0Blood urea nitrogenBlood urea nitrogen measurement (continuous)Month of visitMonth of visit (integer, (January - December) Hypothermia Lowest body temperature within 48 hours (continuous) Vaccination Binary - If vaccinated: 1, if not 0 Serum sodium Serum sodium measurement (continuous) Glucose Glucose measurement (continuous) Other diseases / chronic kidney disease Binary - If applicable: 1, if not applicable: 0 Cerebrovascular disease / history of previous stroke Binary - If applicable: 1, if not applicable: 0 Other diseases / liver cirrhosis Binary - If applicable: 1, if not applicable: 0 No neuromuscular disease Binary - nmd_als If not applicable: 1, if other: 0 No cerebrovascular disease Binary - cvd_pre If not applicable: 1, if other: 0 Sore throat Binary - If present: 1, if not present: 0 Chronic lung disease / interstitial lung disease Binary - If present: 1, if not: 0Chronic lung disease / asthmaBinary - Applies: 1, If ​​not applicable: 0Neoplastic disease / blood cancerBinary - Applies: 1, If ​​not applicable: 0Nasal congestionBinary - If present: 1, If ​​not present: 0Neuromuscular disease / amyotrophic lateral sclerosisBinary - Applies: 1, If ​​not applicable: 0

[0081] The preset features of [Table 2] above may fall into the categories of [Table 3].

[0082] Name of feature Name of category Cough Review of system (ROS) Bullous sound Lung sound No chronic lung disease Underlying disease Alanine aminotransferase (ALT) Lab test Fever Review of system (ROS) C-reactive protein Lab test Runny nose Review of system (ROS) Respiratory rate Vital signs Respiratory disorder Review of system (ROS) Segmented neutrophils Lab test Other chronic lung disease Underlying disease Chronic lung disease / chronic obstructive pulmonary disease Underlying disease Reduced breath sounds Lung sound Demographics by gender (patient information) Chest pain Review of system (ROS) Neoplastic disease Underlying disease No immunosuppression Underlying disease Blood - white blood cell Lab test Wheezing Lung sound Other diseases (heart failure, cirrhosis, chronic kidney disease) NoneUnderlying diseaseHeart rateVital signsNeoplastic disease NoneUnderlying diseaseBMI Demographic (patient information)Age Demographic (patient information)Pneumococcal vaccination (PCV / PPSV23)Vaccination statusOther diseases / heart failureUnderlying diseaseAspartate aminotransferase (AST)Lab testHemoglobinLab testImmunosuppression status / ImmunosuppressantUnderlying diseaseHeadphonesReview of system (ROS)Immunosuppression status / SteroidsUnderlying diseaseHyperthermiaVital signsSystolic blood pressureVital signsMusclesReview of system (ROS)DiarrheaReview of system (ROS)VomitingReview of system (ROS)Blood urea nitrogenLab testVisit month Demographic (patient information)HypothermiaVital signsVaccination (influenza)Vaccination statusSerum Sodium Lab Test Glucose Lab Test Other Diseases / Chronic Kidney Disease Underlying Disease Cerebrovascular Disease / History of Previous Stroke Underlying Disease Other Diseases / Liver Cirrhosis Underlying Disease No Neuromuscular Disease Underlying Disease No Cerebrovascular Disease Underlying Disease Throat Communication System Investigation (ROS) Chronic Lung Disease / Interstitial Lung Disease Underlying Disease Chronic Lung Disease / Asthma Underlying Disease Neoplastic Disease / Blood Cancer Underlying Disease Nasal Congestion System Investigation (ROS)ROS) Neuromuscular disease / amyotrophic lateral sclerosis (ALS) Basal disease;

[0083] The output section (22) classifies the patient as a pneumonia patient based on the input information using a pre-learned model (not shown), and uses beta-lactam as an empirical antibiotic. -lactam) can be output to recommend the use of. Alternatively, the pre-trained model (not shown) can output a recommendation to use respiratory fluoroquinolones as empirical antibiotics. The output unit (22) can output a recommendation not to use respiratory fluoroquinolones as empirical antibiotics for patients diagnosed with tuberculosis using the pre-trained model (not shown). The output unit (22) can output a recommendation to use beta-lactam antibiotics or respiratory fluoroquinolones as empirical treatment for patients diagnosed with mild or moderate disease using the pre-trained model (not shown).

[0084] The output unit (22) can output a recommendation for limited combined administration of beta-lactam antibiotics and macrolide antibiotics for patients diagnosed with atypical bacterial infection or severe pneumonia using a pre-trained model (not shown).

[0085] When the output unit (22) outputs an antibiotic prescription recommendation using a pre-learned model (not shown), it can output a treatment period of at least 4 days.

[0086] Figure 4 is a graph showing the performance of a lower respiratory tract infection diagnosis device (20) including pediatric and adolescent pneumonia according to one embodiment.

[0087] Referring to FIG. 4, a Receiver Operating Characteristic Curve (ROC curve) for a lower respiratory tract infection diagnosis device (20) including pediatric and adolescent pneumonia is shown.

[0088] At this time, the ROC curve is a graph drawn based on the threshold value calculated based on the following mathematical formula 1.

[0089] [Mathematical Formula 1]

[0090]

[0091] Here, optimal threshold is an optimal threshold value, t is an arbitrary real number that can be used as a threshold value, TPR is a true positive rate of a pediatric pneumonia diagnosis device (20) according to an embodiment based on t, and FPR is a false positive rate of a lower respiratory tract infection diagnosis device (20) including pediatric and adolescent pneumonia according to an embodiment based on t.

[0092] Specifically, the confusion matrix of the binary classification of the lower respiratory tract infection diagnosis device (20) including pediatric and adolescent pneumonia to derive the optimal threshold value is as shown in [Table 4] below.

[0093] Predicted value: Negative, Positive, Actual value: Positive 60,736, Negative 1,175,152

[0094] A device (20) for diagnosing lower respiratory tract infections including pneumonia in children and adolescents according to one embodiment produces a confusion matrix of [Table 4] below. At this time, the test dataset includes 796 data cases diagnosed as lower respiratory tract infections including pneumonia and 1327 data cases diagnosed as not having lower respiratory tract infections including pneumonia. Specifically, the device (20) for diagnosing lower respiratory tract infections including pneumonia in children and adolescents according to one embodiment classified the test dataset into 736 true positives (TP), 60 false negatives (FN), 152 false positives (FP), and 1175 true negatives (TN). In summary, the performance of the diagnosis of lower respiratory tract infections including pneumonia in children and adolescents according to one embodiment, which is produced based on the confusion matrix of [Table 4], is as follows.

[0095] Accuracy: 0.9025

[0096] F-1 score: 0.8741

[0097] ROC-AUC: 0.9046

[0098] Figures 4a to 4g are examples of screens of a user terminal (10).

[0099] Fig. 4a is an example of a login screen displayed on a user terminal (10). Referring to Fig. 4a, the login screen may include an input tab that allows the user terminal (10) to input an authentication key to the server (30).

[0100] Specifically, the user terminal (10) can access a website or application and input the ID and password registered in the input tab. The user terminal (10) can request user authentication by transmitting the entered authentication key to the server (30) through the login button displayed on the login screen. At this time, the password is preferably composed of at least 10 characters by combining at least two types of English letters, numbers, and special characters, or at least 8 characters by combining at least three types. Thereafter, the server (30) can perform user authentication by comparing the entered ID and password with data previously stored in the database of the server (30).

[0101] The user terminal (10) can request account creation from the server (30) through the member registration button displayed on the login screen. The user terminal (10) can request a change of a password previously registered with the server (30) through the password search button displayed on the login screen.

[0102] Figure 4b is a landing page displayed on a user terminal (10). Here, the landing page refers to a page related to service provision and account management of a logged-in account.

[0103] For example, a user terminal (10) may request the provision of a lower respiratory tract infection diagnosis service, including pediatric pneumonia, to the server (30) through a diagnosis button on a landing page. As another example, a user terminal may request access to a page for changing the settings of a lower respiratory tract infection diagnosis device (20) including pediatric pneumonia or the server (30), through an administrator screen button on a landing page.

[0104] Figure 4c is an example of a first input screen displayed on a user terminal (10).

[0105] Referring to FIG. 4c, the first input screen may include a plurality of input tabs for entering questionnaire information.

[0106] Specifically, the user terminal (10) can input at least one of the patient's gender, age, height, weight, BMI (Body Mass Index), and time of hospital visit through the input tab of the first input screen.

[0107] At this time, BMI may be a value automatically calculated based on height and weight.

[0108] Referring again to FIG. 4c, the first input screen may include multiple input tabs for entering values ​​for vital signs.

[0109] Specifically, the user terminal (10) can input at least one of the current body temperature, whether or not there is a cough, the patient's respiratory rate, the highest body temperature, the lowest body temperature, the coughing temperament, the number of coughs, the start date of fever, the start date of cough, the duration of fever, the duration of cough, the respiratory rate, the heart rate, and the systolic blood pressure.

[0110] At this time, vital sign information may be biomarker information measured through a primary hospital.

[0111] Figure 4d is an example diagram illustrating a second input screen displayed on a user terminal. For example, the second input screen may be a screen converted from the first input screen.

[0112] Referring to FIG. 4d, the second input screen may include multiple input tabs for entering values ​​of ROS.

[0113] The user terminal (10) can input an ROS value for at least one of the patient's fever, sputum, rhinorrhea, nasal stuffiness, oral intake decrease, sore throat, chest pain, dyspnea, apnea, irritability, lethargy, headache, vomiting, diarrhea, joint pain, myalgia, urinary symptoms, and rash through multiple input tabs.

[0114] As a specific example, the user terminal (10) can input the value of ROS through binary input. For example, if a patient exhibits fever symptoms, the user terminal (10) can input the value of the fever item by checking 1 (YES).

[0115] At this time, the value of ROS may be biomarker information measured through a primary hospital.

[0116] Figure 4e is an example diagram illustrating a third input screen displayed on a user terminal. For example, the third input screen may be a screen converted from the second input screen.

[0117] Referring to FIG. 4e, the third input screen may include multiple input tabs for entering Lab test results.

[0118] Specifically, the user terminal (10) can input values ​​of lab test results including at least one of hemoglobin (Hb), white blood cell (WBC), neutrophil, lymphocyte, fosinopril, platelet, C-reactive protein (CRP), procalcitonin (PCT), blood urea nitrogen (BUN), glucose, aspartate aminotransferase (AST), and alanine aminotransferase (ALT) through a plurality of input tabs.

[0119] At this time, the value of the lab test result may be biomarker information measured through a primary hospital.

[0120] FIG. 4f is an example of a first result screen showing the diagnosis results of lower respiratory tract infection, including pneumonia, displayed on a user terminal (10).

[0121] Referring to FIG. 4f, the first result screen includes the diagnosis results for lower respiratory tract infections including pneumonia.

[0122] Specifically, the first result screen may include at least one of the probability of lower respiratory tract infection including pneumonia, the causative microorganism, the predicted probability for the causative microorganism, the diagnostic finding factor, and the importance of the diagnostic finding factor.

[0123] At this time, the first result screen exemplarily displays the output values ​​of the pediatric pneumonia diagnosis device (20) based on the items input through the server (30), and the arrangement of the UIs is not limited to the illustrated example.

[0124] The first results screen can display the presence or probability of a lower respiratory tract infection, including pneumonia, as a lower respiratory tract infection.

[0125] The first result screen can output the probability of whether the microorganism causing the lower respiratory tract infection is a viral genus. The first result screen can output the probability of whether the microorganism causing the lower respiratory tract infection is a bacterial genus.

[0126] Figure 4g is an example of a second result screen displayed on a user terminal (10). At this time, the second result screen may be provided by switching from or overlapping the first result screen.

[0127] Referring to FIG. 4g, the second result screen may provide at least one of treatment guidelines, diagnostic basis, detailed explanation, and reevaluation information related to the first result screen.

[0128] At this time, the second result screen exemplarily displays the output values ​​of the lower respiratory tract infection diagnosis device (20) including pediatric and adolescent pneumonia based on the items input through the server (30), and the arrangement of the UIs is not limited to the illustrated example.

[0129] For example, the second result screen may provide treatment guidelines recommending antibiotics if the probability of bacterial infection as the cause of the lower respiratory tract infection is higher than the preset probability.

[0130] For example, the second results screen may provide treatment guidelines that include at least one of the oral medication name, indication, recommended single dose, and maximum daily dose.

[0131] Referring to FIG. 5, the method of FIG. 5 can be performed by a lower respiratory tract infection diagnosis device (20) including pediatric and adolescent pneumonia according to one embodiment of FIG. 2.

[0132] First, the lower respiratory tract infection diagnosis device (20) including pediatric and adolescent pneumonia according to one embodiment inputs (510) the patient's questionnaire information into a pre-learned model (not shown).

[0133] Thereafter, the lower respiratory tract infection diagnosis device (20) including pediatric and adolescent pneumonia according to one embodiment outputs (520) the diagnosis results of lower respiratory tract infection including pneumonia for the patient using a pre-learned model (not shown).

[0134] The above-described Figure 5 has been described with reference to the sequence presented in the drawings. For purposes of illustration, the method has been depicted and described as a series of blocks; however, the present invention is not limited to the sequence of the blocks, and some blocks may occur in a different order or concurrently with other blocks than depicted and described herein, and various other branches, flow paths, and sequences of blocks that achieve the same or similar results may be implemented. Furthermore, not all depicted blocks may be required to implement the method described herein.

[0135] Meanwhile, embodiments of the present invention may include a program for performing the methods described herein on a computer, and a computer-readable recording medium including the program. The computer-readable recording medium may include program commands, local data files, local data structures, etc., alone or in combination. The medium may be specially designed and configured for the present invention, or may be one commonly used in the field of computer software. Examples of the computer-readable recording medium include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and hardware devices specially configured to store and execute program commands such as ROMs, RAMs, and flash memories. Examples of the program may include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter, etc.

[0136] While representative embodiments of the present invention have been described in detail above, those skilled in the art will appreciate that various modifications to the above-described embodiments are possible without departing from the scope of the present invention. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be determined not only by the claims set forth below but also by equivalents thereof.

[0137] A device, method, and system for diagnosing lower respiratory tract infections, including pediatric pneumonia, according to one embodiment can be used in the digital medical industry by predicting whether a patient has pneumonia using a learning model.

Claims

1. One or more processors; and A device for diagnosing lower respiratory tract infections, including pediatric pneumonia, having a memory for storing instructions executed by one or more processors, The one or more processors: input at least one of the patient's medical history information and biomarker information into the pre-trained model, A device for diagnosing lower respiratory tract infections, including pneumonia, in children and adolescents, which outputs a diagnosis result for lower respiratory tract infections, including pneumonia, for the patient using the above-mentioned learned model.

2. In paragraph 1, The above questionnaire information includes personal information in text format, The above biomarker information is a diagnostic device for lower respiratory tract infections, including pediatric pneumonia, which includes basic test result information in text format performed at a primary hospital.

3. In paragraph 1, A device for diagnosing lower respiratory tract infections, including pediatric and adolescent pneumonia, wherein the above questionnaire information includes personal information values ​​for at least one of the patient's gender, age, height, weight, and time of hospital visit.

4. In paragraph 1, A device for diagnosing lower respiratory tract infections including pneumonia in children and adolescents, wherein the biomarker information includes vital sign values ​​for at least one of the patient's respiratory rate, maximum body temperature, minimum body temperature, blood pressure, heart rate, body temperature, respiratory rate, heart rate, and systolic blood pressure.

5. In paragraph 1, A diagnostic device for lower respiratory tract infections, including pneumonia in children and adolescents, wherein the biomarker information includes an underlying disease history for at least one of chronic pulmonary disease, cerebrovascular disease, neuromuscular disease, oncological disease, cardiovascular disease, nervous system disease, hematologic oncology, renal disease, liver disease, other diseases, and immune disease.

6. In paragraph 1, A diagnostic device for lower respiratory tract infections, including pediatric pneumonia, wherein the biomarker information includes a vaccination history for at least one of a protein conjugate vaccine, a polysaccharide vaccine, a pneumococcal vaccine, and an influenza vaccine.

7. In paragraph 1, A device for diagnosing lower respiratory tract infections, including pneumonia in children and adolescents, wherein the biomarker information includes lung sound results for at least one of rales, wheezing, breath sounds, and breath sound changes.

8. In paragraph 1, A device for diagnosing lower respiratory tract infections, including pneumonia in children and adolescents, wherein the biomarker information includes test results for at least one of hemoglobin, blood leukocytes, segmented neutrophils, C-reactive protein (CRP), blood urea nitrogen, serum sodium, glucose, aspartate aminotransferase (AST), and alanine aminotransferase (ALT).

9. In paragraph 1, A device for diagnosing lower respiratory tract infections including pneumonia in children and adolescents, wherein the diagnosis result of lower respiratory tract infections including pneumonia includes at least one of the presence or absence of lower respiratory tract infections including pneumonia, probability, causative microorganism, predicted probability for causative microorganism, infection finding factors, and normal finding factors.

10. In paragraph 1, The above model is a device for diagnosing lower respiratory tract infections including pediatric pneumonia, including an XGBoost (Extreme Gradient Boost) decision tree or a random forest decision tree.

11. One or more processors; and A method performed by a device for diagnosing lower respiratory tract infections including pediatric pneumonia, having a memory for storing a program executed by one or more processors, A step of inputting at least one of the patient's medical history information and biomarker information into a pre-trained model; and A method for diagnosing lower respiratory tract infections, including pneumonia, in children and adolescents, comprising a step of outputting a diagnosis result for lower respiratory tract infections, including pneumonia, for the patient using the above-mentioned learned model.

12. In paragraph 11, The above questionnaire information includes personal information in text format, A method for diagnosing lower respiratory tract infections, including pediatric pneumonia, wherein the above biomarker information includes basic test result information in text format performed at a primary hospital.

13. In paragraph 11, A method for diagnosing lower respiratory tract infections, including pediatric pneumonia, wherein the above questionnaire information includes personal information values ​​for at least one of the patient's gender, age, height, weight, and time of hospital visit.

14. In paragraph 11, A method for diagnosing lower respiratory tract infection, including pediatric pneumonia, wherein the biomarker information includes vital sign values ​​for at least one of the patient's respiratory rate, maximum body temperature, minimum body temperature, blood pressure, heart rate, body temperature, respiratory rate, heart rate, and systolic blood pressure.

15. In paragraph 11, A method for diagnosing lower respiratory tract infection, including pneumonia in children and adolescents, wherein the biomarker information includes an underlying disease history of at least one of chronic pulmonary disease, cerebrovascular disease, neuromuscular disease, oncological disease, cardiovascular disease, nervous system disease, hematologic oncology, renal disease, liver disease, other diseases, and immune disease.

16. In paragraph 11, A method for diagnosing lower respiratory tract infections, including pneumonia in children and adolescents, wherein the biomarker information includes a vaccination history for at least one of a protein conjugate vaccine, a polysaccharide vaccine, a pneumococcal vaccine, and an influenza vaccine.

17. In paragraph 11, A method for diagnosing lower respiratory tract infections, including pneumonia in children and adolescents, wherein the biomarker information includes lung sound results for at least one of rales, wheezing, breath sounds, and breath sound changes.

18. In paragraph 11, A method for diagnosing lower respiratory tract infection, including pediatric pneumonia, wherein the biomarker information includes test results for at least one of hemoglobin, blood leukocytes, segmented neutrophils, C-reactive protein (CRP), blood urea nitrogen, serum sodium, glucose, AST (Aspartate Aminotransferase), and ALT (Alanine Aminotransferase).

19. In paragraph 11, A method for diagnosing lower respiratory tract infections, including pneumonia, in children and adolescents, wherein the diagnosis result of lower respiratory tract infections, including pneumonia, includes at least one of the presence or absence of lower respiratory tract infections, probability, causative microorganism, predicted probability for causative microorganism, and infection finding factors.

20. In paragraph 11, The above model is a method for diagnosing lower respiratory tract infections, including pediatric pneumonia, including an XGBoost (Extreme Gradient Boost) decision tree or a random forest decision tree.

21. A device for diagnosing lower respiratory tract infections, including pediatric and adolescent pneumonia, according to any one of Articles 1 to 10; A terminal that inputs the patient's questionnaire information and the biomarker information into the lower respiratory tract infection diagnosis device including pediatric and adolescent pneumonia through a server; and A lower respiratory tract infection diagnosis system including pediatric pneumonia, which includes a server that provides a lower respiratory tract infection diagnosis service including pediatric pneumonia to the terminal using the lower respiratory tract infection diagnosis results including the pneumonia output by the pediatric pneumonia diagnosis device.

Images