Methods for predicting and managing readmission in patients with ischemic stroke, system for predicting and managing readmission in patients with ischemic stroke and a medium having computer readable program for executing the method

Abstract

The present invention relates to a method for predicting and managing the re-hospitalization of ischemic stroke patients, a system for predicting the re-hospitalization of ischemic stroke patients, and a medium storing a computer-readable program for executing the method for predicting and managing the re-hospitalization of ischemic stroke patients, which can predict the likelihood of re-hospitalization for a patient diagnosed with and treated for ischemic stroke based on specific factors of re-hospitalization factors for each patient, and thereby support the effective and efficient intensive post-hospitalization management of ischemic stroke patients. According to the present invention, a method for predicting and managing the re-hospitalization of ischemic stroke patients comprises: a raw data construction step of constructing personal information and treatment information of ischemic stroke patients in a raw data database; and a re-hospitalization patient history information construction step of constructing re-hospitalization information, including personal information and treatment information of re-hospitalized patients among ischemic stroke patients, in a history information database from the constructed raw data. A method for predicting and managing the re-hospitalization of a patient with ischemic stroke is provided, characterized by comprising: a step of extracting specific re-hospitalization factors based on established re-hospitalization information, wherein one or more specific re-hospitalization factors are extracted from said re-hospitalization information through a specific re-hospitalization factor extraction unit; and a step of predicting the re-hospitalization of a patient based on the extracted specific re-hospitalization factors, wherein the re-hospitalization status of the patient is analyzed and predicted through a re-hospitalization analysis prediction unit.

Description

Technology Field

[0001] The present invention relates to a method for predicting and managing the re-hospitalization of an ischemic stroke patient, a system for predicting the re-hospitalization of an ischemic stroke patient, and a medium storing a computer-readable program for executing the method for predicting and managing the re-hospitalization of an ischemic stroke patient. More specifically, the invention relates to a method for predicting and managing the re-hospitalization of an ischemic stroke patient, a system for predicting the re-hospitalization of an ischemic stroke patient, and a medium storing a computer-readable program for executing the method for predicting and managing the re-hospitalization of an ischemic stroke patient, which can predict the likelihood of re-hospitalization (or re-hospitalization rate) of a patient diagnosed with and treated for ischemic stroke based on specific factor(s) of re-hospitalization factors for each patient, and thereby support the effective and efficient intensive post-hospitalization management of the ischemic stroke patient. Background Technology

[0003] Stroke is the leading cause of death among single diseases in Korea, and the disease burden is gradually increasing as life expectancy extends and the elderly population grows.

[0004] Furthermore, since stroke is known to be the most significant cause of functional disability in adults, appropriate treatment during the acute phase is necessary, and the need for comprehensive rehabilitation is being emphasized.

[0005] Ultimately, in order to effectively implement comprehensive rehabilitation treatment for patients suffering from this disease, the setting of appropriate treatment goals must be prioritized; to achieve this, it is essential to predict the patient's level of recovery, and follow-up care is of the utmost importance.

[0006] Existing classical models for predicting the functional prognosis of stroke have many limitations, and standard criteria have not been established. Based on health checkup data from the National Health Insurance Service, the Korean Stroke Risk Prediction (KSRP) model was published in 2007 as a predictive model regarding the average risk of stroke occurring within 10 years, utilizing factors such as age, systolic blood pressure, diabetes, total cholesterol, smoking, exercise, body mass index, and alcohol consumption (Jee et al, 2007).

[0007] However, the above model did not consider major risk factors for stroke and evaluated the 10-year probability of stroke occurrence by scoring, so it could not predict the prognosis of patients after stroke occurrence. Furthermore, it had limitations in being applied to actual clinical practice and used for patient care because it did not consider stroke risk factors highly associated with prognosis, treatment in the event of acute stroke, or outcomes according to stroke subtypes, and it had the disadvantage of being difficult to predict the prognosis after stroke.

[0008] Meanwhile, readmission is an unintended consequence that occurs in discharged patients. In Korea, the 30-day readmission rate at tertiary general hospitals was approximately 30% in 2020, and it is increasing every year along with statistics on readmission costs.

[0009] According to a report by the U.S. Centers for Medicare and Medicaid Services (CMS), the annual readmission rate reaches 2 million patients, and the cost associated with readmission amounts to $26 billion. Despite increasing interest and literature on readmissions, the rate of unplanned readmissions still requires management. To reduce readmission costs and improve the quality of care, it is essential to accurately predict and screen high-risk discharged patients.

[0010] In this regard, predictive modeling is an efficient method for stratifying the risk of rehospitalization and optimizing the allocation of clinical resources by providing preventive interventions to high-risk patients. However, existing risk prediction models, such as regression analysis, have limited predictive power as seen in previous studies, and there is a problem in that the treatment of high-risk discharged patients, particularly those with a high likelihood of rehospitalization, is concentrated only at the time of discharge, such as by providing post-discharge management interventions.

[0011] Therefore, there is a need for research and the development of related technologies regarding personalized prediction and management treatment for individual ischemic stroke patients through stratified prediction models with high predictive power that analyze high-dimensional medical data. Prior art literature

[0013] Republic of Korea Published Patent Application 10-2025-0030287 (Published March 5, 2025) Republic of Korea Published Patent Application 10-2024-0157442 (Published Nov. 01, 2024) Republic of Korea Published Patent Application 10-2024-0157574 (Published Nov. 01, 2024) Republic of Korea Registered Patent Application 10-2216822 (Announced Feb. 18, 2021) Republic of Korea Registered Patent Application 10-2447401 (Announced Sep. 28, 2022) The problem to be solved

[0014] Accordingly, the present invention, which aims to solve the aforementioned conventional problems, provides a method for predicting and managing the re-hospitalization of ischemic stroke patients, a system for predicting the re-hospitalization of ischemic stroke patients, and a medium storing a computer-readable program for executing the method for predicting and managing the re-hospitalization of ischemic stroke patients, which can support effective and efficient intensive post-treatment care of ischemic stroke patients by predicting and managing re-hospitalization of ischemic stroke patients based on one or more specific factors for patient-specific re-hospitalization factors after treatment of ischemic stroke patients.

[0015] The problems solved by the present invention are not limited to those mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the description below. means of solving the problem

[0017] According to one aspect of the present invention for achieving the objectives and other features of the present invention, a method for predicting and managing the possibility of re-hospitalization of an ischemic stroke patient diagnosed with and treated for ischemic stroke is provided, comprising: a raw data construction step of constructing personal information and treatment information of the ischemic stroke patient in a raw data database; a re-hospitalization patient history information construction step of constructing re-hospitalization information including personal information and treatment information of a re-hospitalized patient among the ischemic stroke patients in the constructed raw data in a history information database; a re-hospitalization factor specific factor extraction step of extracting one or more re-hospitalization factor specific factors from the re-hospitalization information based on the constructed re-hospitalization information through a re-hospitalization factor specific factor extraction unit; and a re-hospitalization prediction step of analyzing and predicting whether the patient will be re-hospitalized based on the extracted re-hospitalization factor specific factors through a re-hospitalization analysis prediction unit.

[0018] In one aspect of the present invention, the data constructed in the raw data construction step may include personal information such as gender, age, smoking status, alcohol consumption, obesity level, and genetic relationship, and treatment information such as blood test results, comorbidities, cholesterol levels, embolism, thrombosis, blood pressure, arteriosclerosis, hypertension, diabetes, administered drugs and dosages, and nasogastric tube insertion.

[0019] In one aspect of the present invention, the blood test values ​​in the treatment information are preferably creatinine, blood urea nitrogen (BUN), calcium (Ca), sodium (Na), potassium (K), hemoglobin (Hb), albumin, mean corpuscular volume (MCV), leukocytes, chloride (Cl), and thrombin time (TT).

[0020] In one aspect of the present invention, the step of extracting specific factors for re-hospitalization factors may be comprised of extracting kidney and metabolism-related factors.

[0021] In one aspect of the present invention, the step of extracting specific rehospitalization factor factors may consist of extracting creatinine, blood urea nitrogen (BUN), calcium (Ca), sodium (Na), potassium (K), and the respective values ​​thereof as specific rehospitalization factor factors.

[0022] In one aspect of the present invention, the re-hospitalization prediction step may be configured to determine that there is a high probability of re-hospitalization when the value of at least one specific re-hospitalization factor deviates from a preset reference value or reference range.

[0023] In one aspect of the present invention, the re-hospitalization prediction step may be configured to visualize and provide the possibility of re-hospitalization by determining whether the numerical value of the re-hospitalization factor specific factor is higher or lower than a preset reference value or reference range.

[0024] In one aspect of the present invention, the re-hospitalization prediction step may be performed by analyzing and predicting using an artificial intelligence module of a deep learning model and a machine learning model.

[0025] According to another aspect of the present invention, a medium is provided that records a computer-readable program for executing a method for predicting and managing the re-hospitalization of an ischemic stroke patient according to the above-described aspect.

[0026] According to another aspect of the present invention, a system for predicting and managing the possibility of re-hospitalization of an ischemic stroke patient diagnosed with and treated for ischemic stroke is provided, comprising: a raw data database unit in which personal information and treatment information of the ischemic stroke patient are constructed; a history information database unit in which re-hospitalization information including personal information and treatment information of a re-hospitalized patient among the ischemic stroke patients is constructed from the constructed raw data; a re-hospitalization factor specific factor extraction unit configured to extract one or more re-hospitalization factor specific factors from the re-hospitalization information based on the constructed re-hospitalization information; and a re-hospitalization analysis prediction unit configured to analyze and predict whether the patient will be re-hospitalized based on the extracted re-hospitalization factor specific factors.

[0027] In another aspect of the present invention, the raw data database unit may include data as personal information such as gender, age, smoking status, alcohol consumption, obesity level, and genetic relationship, and as treatment information such as blood test results, comorbidities, cholesterol levels, embolism, thrombosis, blood pressure, arteriosclerosis, hypertension, diabetes, administered drugs and dosages, and nasogastric tube insertion status.

[0028] In another aspect of the present invention, the treatment information includes blood test results, and preferably, the blood test results are values ​​of creatinine, blood urea nitrogen (BUN), calcium (Ca), sodium (Na), potassium (K), hemoglobin (Hb), albumin, mean corpuscular volume (MCV), leukocytes, chloride (Cl), and thrombin time (TT).

[0029] In another aspect of the present invention, the re-hospitalization factor specific factor extraction unit may be configured to extract kidney and metabolism-related factors.

[0030] In another aspect of the present invention, the re-hospitalization factor specific factor extraction unit may be composed of extracting creatinine, blood urea nitrogen (BUN), calcium (Ca), sodium (Na), potassium (K), and the respective values ​​thereof as re-hospitalization factor specific factors.

[0031] In another aspect of the present invention, the re-hospitalization analysis prediction unit may be configured to determine that there is a high probability of re-hospitalization when the value of at least one specific re-hospitalization factor deviates from a preset reference value or reference range.

[0032] In another aspect of the present invention, the re-hospitalization analysis prediction unit may be configured to visualize and provide the possibility of re-hospitalization by determining whether the numerical value of the re-hospitalization factor specific factor is higher or lower than a preset reference value or reference range.

[0033] In another aspect of the present invention, the re-hospitalization analysis prediction unit is configured to analyze and predict using an artificial intelligence module of a deep learning model and a machine learning model, and may further include a post-monitoring unit configured to provide relevant information to the patient with ischemic stroke or a related guardian based on the results predicted by the re-hospitalization analysis prediction unit. Effects of the invention

[0035] According to the method for predicting and managing the re-hospitalization of an ischemic stroke patient, the system for predicting the re-hospitalization of an ischemic stroke patient, and a medium storing a computer-readable program for executing the method for predicting and managing the re-hospitalization of an ischemic stroke patient according to the present invention, the following effects are provided.

[0036] First, the present invention can provide a novel technology capable of predicting re-hospitalization for patients diagnosed with and treated for ischemic stroke, thereby contributing to the advancement of related technological fields.

[0037] Second, the present invention extends from predictive diagnosis of ischemic stroke to have the effect of promoting systematic management following the diagnosis and treatment of ischemic encephaloma.

[0038] Third, the present invention has the effect of enabling more effective and systematic post-operative care for ischemic stroke patients by predicting and managing re-hospitalization based on specific factors for patient-specific re-hospitalization.

[0039] Fourth, the present invention has the effect of supporting rapid and effective medical action by taking prompt measures for the patient based on specific factors of patient-specific re-hospitalization factors, such as creatinine levels, and visual results thereof.

[0040] Fifth, the present invention has the effect of supporting customized services such as rehabilitation plans, outpatient appointment schedules, home visits, and telephone monitoring by enabling the detailed establishment of post-discharge management plans for high-risk patients through patient-specific customized reports.

[0041] Sixth, the present invention has the effect of reducing unnecessary hospitalizations and enabling the efficient utilization of medical resources by selecting patients at high risk of re-hospitalization and providing intensive follow-up care.

[0042] The effects of the present invention are not limited to those mentioned above, and other unmentioned problems will be clearly understood by those skilled in the art from the description below. Brief explanation of the drawing

[0044] FIG. 1 is a flowchart schematically illustrating the process of predicting and managing the re-hospitalization of an ischemic stroke patient by the method for predicting and managing the re-hospitalization of an ischemic stroke patient according to the present invention. FIG. 2 is a diagram schematically illustrating the configuration of a system for predicting and managing the re-hospitalization of ischemic stroke patients according to the present invention. FIG. 3 is a block diagram schematically illustrating the configuration of a re-hospitalization prediction-management module included in the re-hospitalization prediction and management system for ischemic stroke patients according to the present invention. Figure 4 is a diagram of the AUROC (Area Under the Receiver Operating Characteristic Curve) of the model for the test set. Figure 5 is a diagram of the SHAP summary plot of the best model (Light GBM). Figure 6 is a graph showing patients with (A) the highest SHAP value and (B) the lowest SHAP value. Specific details for implementing the invention

[0045] Further objects, features, and advantages of the present invention can be more clearly understood from the following detailed description and the accompanying drawings.

[0046] Before providing a detailed description of the present invention, it should be understood that the present invention is capable of various modifications and may have various embodiments, and that the examples described below and illustrated in the drawings are not intended to limit the present invention to specific embodiments, but rather include all modifications, equivalents, and substitutions that fall within the spirit and scope of the present invention.

[0047] When it is stated that one component is "connected" or "connected" to another component, it should be understood that while it may be directly connected or connected to that other component, there may also be other components in between. On the other hand, when it is stated that one component is "directly connected" or "directly connected" to another component, it should be understood that there are no other components in between.

[0048] The terms used in this specification are used merely to describe specific embodiments and are not intended to limit the invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this specification, terms such as "comprising" or "having" are intended to indicate the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0049] Additionally, terms such as "...part," "...unit," and "...module" described in the specification refer to a unit that processes at least one function or operation, and this may be implemented in hardware, software, or a combination of hardware and software.

[0050] Furthermore, in the description referring to the attached drawings, identical components are assigned the same reference numeral regardless of drawing symbols, and redundant descriptions thereof are omitted. In describing the present invention, if it is determined that a detailed description of related prior art could unnecessarily obscure the essence of the present invention, such detailed description is omitted.

[0051] The method for predicting and managing the re-hospitalization of an ischemic stroke patient, the system for predicting the re-hospitalization of an ischemic stroke patient, and the medium storing a computer-readable program for executing the method for predicting and managing the re-hospitalization of an ischemic stroke patient according to the present invention are configured to construct various treatment-related data and data on re-hospitalized ischemic stroke patients regarding patients diagnosed with and treated for ischemic stroke, and based on the constructed related patient data, extract and analyze data on the possibility of re-hospitalization through an artificial intelligence (AI) module regarding specific factor(s) of re-hospitalization factors for each patient, and based on this, determine the possibility of re-hospitalization of an ischemic stroke patient and share this with medical professionals and patients (patient guardians) through various methods, thereby enabling more effective and systematic post-hospitalization management for each ischemic stroke patient.

[0052] Hereinafter, a method for predicting and managing the re-hospitalization of an ischemic stroke patient according to a preferred embodiment of the present invention, a system for predicting the re-hospitalization of an ischemic stroke patient, and a medium storing a computer-readable program for executing the method for predicting and managing the re-hospitalization of an ischemic stroke patient will be described in detail with reference to the attached drawings.

[0053] First, a method for predicting and managing the re-hospitalization of a patient with ischemic stroke according to the present invention will be described in detail with reference to FIG. 1.

[0054] FIG. 1 is a flowchart schematically illustrating the process of predicting and managing the re-hospitalization of an ischemic stroke patient by the method for predicting and managing the re-hospitalization of an ischemic stroke patient according to the present invention.

[0055] The method for predicting and managing the re-hospitalization of an ischemic stroke patient according to the present invention is a method for predicting and managing the possibility of re-hospitalization of an ischemic stroke patient diagnosed with and treated for ischemic stroke, and as shown in FIG. 1, it mainly comprises a raw data construction step (S100), a re-hospitalization patient history information construction step (S200), a re-hospitalization factor specific factor extraction step (S300), and a re-hospitalization prediction step (S400), and may additionally include a post-monitoring step (S500).

[0056] Specifically, the method for predicting and managing the re-hospitalization of an ischemic stroke patient according to the present invention is a method for predicting and managing the possibility of re-hospitalization of an ischemic stroke patient diagnosed with and treated for ischemic stroke, comprising: a step of constructing raw data of an ischemic stroke patient (hereinafter briefly referred to as the 'raw data construction step') (S100) configured to construct basic information (personal data) and treatment information (diagnosis and treatment data) of a patient diagnosed with and treated for ischemic stroke, as shown in FIG. 1; and a step of constructing re-hospitalization information of a re-hospitalized patient (hereinafter referred to as the 're-hospitalized patient') including basic information (personal data) and treatment information of a re-hospitalized patient among ischemic stroke patients, from the information constructed in the raw data construction step (S100), and constructing re-hospitalization information (re-hospitalization data). The method comprises: a step for extracting specific re-hospitalization factors (S300) for extracting one or more specific re-hospitalization factors from re-hospitalization information based on the re-hospitalization information established in the step for establishing re-hospitalization patient history information (S200); and a step for predicting re-hospitalization (S400) for analyzing and predicting whether the patient will be re-hospitalized based on the specific re-hospitalization factors extracted in the step for extracting specific re-hospitalization factors (S300). In addition, the method for predicting and managing re-hospitalization of an ischemic stroke patient according to the present invention may further include a post-monitoring step (S500) for providing relevant information to the ischemic stroke patient or a related guardian based on the result predicted in the re-hospitalization prediction step (S400).

[0057] The above raw data construction step (S100) is a process for constructing basic information (human data) and treatment information (diagnosis and treatment data) of a patient diagnosed with and treated for ischemic stroke.

[0058] Specifically, the above raw data construction step (S100) may be constructed by including gender, age, smoking status, alcohol consumption, obesity level, and genetic correlation as basic information (human data), and including blood test results, comorbidities, cholesterol levels, embolism, thrombosis, blood pressure, arteriosclerosis, hypertension, diabetes, etc. as treatment information (diagnosis and treatment data).

[0059] In the above treatment information, blood test values ​​may include creatinine, blood urea nitrogen (BUN), calcium (Ca), sodium (Na), potassium (K), hemoglobin (Hb), albumin, mean corpuscular volume (MCV), white blood cells, chloride (Cl), thrombin time (TT), etc., and all or part of these blood test values ​​may be used as one of the specific factors for re-hospitalization.

[0060] In addition, the treatment data in the above treatment information may include the administered drug and dosage (e.g., whether a thrombolytic agent is administered and the dosage), whether a nasogastric tube is inserted, etc.

[0061] In addition, the associated diseases in the above treatment information may include cerebral infarction (disease code I63), other cerebrovascular diseases (disease code I67), pneumonia (disease code I17), angina pectoris (disease code I20), dementia (disease code F03), etc.

[0062] Next, the above-mentioned re-hospitalization patient history information construction step (S200) is a process of extracting and constructing re-hospitalization information (re-hospitalization data) including basic information (personal data) and treatment information of a re-hospitalized patient (hereinafter referred to as 're-hospitalized patient') among ischemic stroke patients from the information constructed in the above-mentioned raw data construction step (S100).

[0063] Specifically, in the present invention, the step of constructing the history information of the re-hospitalized patient (S200) includes re-hospitalization treatment information, including treatment data prior to re-hospitalization and treatment data after re-hospitalization for a patient who has been re-hospitalized among patients with ischemic stroke, from the information constructed in the raw data construction step (S100), and additionally may further include the length of stay (date of completion of elapsed after re-hospitalization (e.g., date of discharge or death).

[0064] The treatment data after the above readmission is linked with the treatment data prior to readmission and includes not only the same treatment data as the prior treatment data but also newly added treatment data.

[0065] Next, the step of extracting specific factors for re-hospitalization factors (S300) is a process of extracting one or more specific factors for re-hospitalization factors from the re-hospitalization information (re-entered patient history information) based on the re-hospitalization information established in the step of establishing re-hospitalization patient history information (S200).

[0066] In the present invention, the step of extracting specific factors for re-hospitalization factors (S300) comprises extracting kidney and metabolism-related factors as the main specific factors for re-hospitalization factors.

[0067] Specifically, in the present invention, the rehospitalization factor specific factor extracted in the rehospitalization factor specific factor extraction step (S300) is preferably a main rehospitalization factor specific factor, comprising creatinine, blood urea nitrogen (BUN), calcium (Ca), sodium (Na), potassium (K), and the respective values ​​thereof.

[0068] Next, the re-hospitalization prediction step (S400) is a process for analyzing and predicting whether the patient will be re-hospitalized based on the specific re-hospitalization factor(s) extracted in the re-hospitalization factor specific factor extraction step (S300).

[0069] The above re-hospitalization prediction step (S400) is configured such that, among the re-hospitalization factor specific factor extracted in the above re-hospitalization factor specific factor extraction step (S300), at least one value among creatinine, blood urea nitrogen (BUN), calcium (Ca), sodium (Na), and potassium (K) is determined to have a high probability of re-hospitalization if it deviates from a predetermined standard value or standard range, for example, if it is higher than the standard value or standard range.

[0070] Alternatively, the above re-hospitalization prediction step (S400) may be configured to predict the possibility of re-hospitalization (re-hospitalization rate) and visualize and provide the result by determining whether the value of at least one specific re-hospitalization factor among creatinine, blood urea nitrogen (BUN), calcium (Ca), sodium (Na), and potassium (K) is higher or lower than a preset standard (value) or standard range.

[0071] In the above re-hospitalization prediction step (S400), the value of the re-hospitalization factor specific factor may be the value of the re-hospitalization factor specific factor linked to the administered drug and dosage administered not only to the patient in question but also to other ischemic stroke patients.

[0072] In the present invention, the re-hospitalization prediction step (S400) may be performed by learning and evaluating using an artificial intelligence module including a deep learning model and a machine learning model.

[0073] For example, artificial neural networks (ANN), convolutional neural networks (CNN), and recurrent neural networks (RNN) can be used as deep learning models, and Random Forest, Gradient Boosting Machine (GBM), XGBoost, Support Vector Machine (SVM), and k-Nearest Neighbors (KNN) can be used as machine learning models.

[0074] In other words, the rehospitalization prediction step (S400) of the present invention can be performed by using an artificial intelligence module such as SHAP (Shapley Additive Explanations) to analyze specific rehospitalization factors, which are rehospitalization prediction variables, based on whole patient group data (raw data of ischemic stroke patients and history data of rehospitalized patients), and to visualize and provide the influence of the risk of rehospitalization.

[0075] In addition, the method for predicting and managing the re-hospitalization of an ischemic stroke patient according to the present invention may further include a post-monitoring step (S500) for providing relevant information to the ischemic stroke patient or a related guardian based on the result predicted in the re-hospitalization prediction step (S400).

[0076] The above post-monitoring step (S500) may be provided to a terminal device of an ischemic stroke patient via a known wired or wireless communication network.

[0077] In particular, the post-monitoring step (S500) can be configured to receive real-time data from a bio-information detection device (200) equipped by a patient with ischemic stroke, for example, a bio-information detection wearable device capable of detecting bio-information in real time and transmitting and receiving related data through a communication module, and to transmit management information (e.g., messages such as '~ value is high', 'Please manage your blood pressure', 'Please visit a hospital within 00 days') to the bio-information detection wearable device based on the result predicted in the re-hospitalization prediction step (S400).

[0078] Meanwhile, the re-hospitalization prediction and management system for ischemic stroke patients according to the present invention will be described in detail with reference to FIG. 2.

[0079] FIG. 2 is a diagram schematically illustrating the configuration of a re-hospitalization prediction and management system for ischemic stroke patients according to the present invention, and FIG. 3 is a block diagram schematically illustrating the configuration of a re-hospitalization prediction-management module included in the re-hospitalization prediction and management system for ischemic stroke patients according to the present invention.

[0080] The re-hospitalization prediction and management system for an ischemic stroke patient according to the present invention is a system for predicting and managing the possibility of re-hospitalization of an ischemic stroke patient diagnosed with and treated for ischemic stroke, and as shown in FIGS. 2 and 3, is largely implemented by known hardware components, software components, and / or a combination of hardware components and software components to predict and manage the re-hospitalization of an ischemic stroke patient, and comprises a raw data database unit (100), a re-hospitalization patient history database unit (200), a re-hospitalization factor specific factor extraction unit (300), and a re-hospitalization analysis prediction unit (400), and may additionally include a post-monitoring unit (500).

[0081] Specifically, the re-hospitalization prediction and management system for ischemic stroke patients according to the present invention is a system for predicting and managing the possibility of re-hospitalization of ischemic stroke patients diagnosed with and treated for ischemic stroke, and as shown in FIGS. 2 and 3, comprises: a low data database section (low data DB section) (100) for ischemic stroke patients in which basic information (personal data) and treatment information (diagnosis and treatment data) of patients diagnosed with and treated for ischemic stroke are established; and a re-hospitalization information database section (200) in which re-hospitalization information (re-hospitalization data) including basic information (personal data) and treatment information (diagnosis and treatment data) of re-hospitalized patients (hereinafter referred to as 're-hospitalized patients') among ischemic stroke patients is established from the information established in the low data database section (100). The system comprises: a re-hospitalization factor specific factor extraction unit (300) configured to extract one or more re-hospitalization factor specific factor(s) from re-hospitalization information based on re-hospitalization information established in the re-hospitalization patient history information database unit (200); and a re-hospitalization analysis prediction unit (400) configured to analyze and predict whether the patient will be re-hospitalized based on the re-hospitalization factor specific factor(s) extracted by the re-hospitalization factor specific factor extraction unit (300). In addition, the re-hospitalization prediction and management system for ischemic stroke patients according to the present invention may further include a post-monitoring unit (500) configured to provide relevant information to the ischemic stroke patient or a related guardian based on the result predicted by the re-hospitalization analysis prediction unit (400).

[0082] The above raw data database section (100) is a component configured to build basic information (personal data) and treatment information (diagnosis and treatment data) of a patient diagnosed with and treated for ischemic stroke.

[0083] Specifically, the raw data database unit (100) may be constructed by including, as basic information (human data), gender, age, smoking status, amount of alcohol consumed, obesity level, genetic relationship, etc., and as treatment information (diagnosis and treatment data), data including blood test results, comorbidities, cholesterol levels, embolism, thrombosis, blood pressure, arteriosclerosis, hypertension, diabetes, etc.

[0084] In the above treatment information, blood test values ​​may include creatinine, blood urea nitrogen (BUN), calcium (Ca), sodium (Na), potassium (K), hemoglobin (Hb), albumin, mean corpuscular volume (MCV), white blood cells, chloride (Cl), thrombin time (TT), etc., and all or part of these blood test values ​​may be used as one of the specific factors for re-hospitalization.

[0085] In addition, the treatment data in the above treatment information may include the administered drug and dosage (e.g., whether a thrombolytic agent is administered and the dosage), whether a nasogastric tube is inserted, etc.

[0086] In addition, the associated diseases in the above treatment information may include cerebral infarction (disease code I63), other cerebrovascular diseases (disease code I67), pneumonia (disease code I17), angina pectoris (disease code I20), dementia (disease code F03), etc.

[0087] Next, the above-mentioned re-hospitalized patient history information database section (200) is a component section in which re-hospitalization information (re-hospitalization data) including basic information (personal data) and treatment information of a patient who has been re-hospitalized among patients with ischemic stroke (hereinafter referred to as 're-hospitalized patient') is constructed from the information constructed in the above-mentioned raw data database section (100).

[0088] Specifically, in the present invention, the re-hospitalized patient history information database unit (200) includes re-hospitalization treatment information, including treatment data before re-hospitalization and treatment data after re-hospitalization, for a patient who has been re-hospitalized among patients with ischemic stroke, in the information built in the raw data database unit (100), and additionally may include the length of stay (date of completion of elapsed after re-hospitalization (e.g., date of discharge or death)).

[0089] The treatment data after the above readmission is linked with the treatment data prior to readmission and includes not only the same treatment data as the prior treatment data but also newly added treatment data.

[0090] Next, the re-hospitalization factor specific factor extraction unit (300) is configured to extract one or more re-hospitalization factor specific factor(s) from the re-hospitalization information (re-entered patient history information) based on the re-hospitalization information built in the re-hospitalization patient history information database unit (200).

[0091] In the present invention, the re-hospitalization factor specific factor extraction unit (300) is configured to extract kidney and metabolism-related factors as the main re-hospitalization factor specific factor.

[0092] Specifically, in the present invention, the re-hospitalization factor specific factor extraction unit (300) preferably comprises extracting creatinine, blood urea nitrogen (BUN), calcium (Ca), sodium (Na), potassium (K), and the respective values ​​of these as main re-hospitalization factor specific factors.

[0093] Next, the re-hospitalization analysis prediction unit (400) is configured to analyze and predict whether the patient will be re-hospitalized based on the specific re-hospitalization factor(s) extracted from the re-hospitalization factor specific factor extraction unit (300).

[0094] The above re-hospitalization analysis prediction unit (400) is configured to determine that there is a high probability of re-hospitalization when, among the re-hospitalization factor specific factor extracted by the above re-hospitalization factor specific factor extraction unit (300), at least one value among creatinine, blood urea nitrogen (BUN), calcium (Ca), sodium (Na), and potassium (K) deviates from a preset standard (value) or standard range, for example, when it is higher than the standard value or standard range.

[0095] Alternatively, the above re-hospitalization analysis prediction unit (400) may be configured to predict the possibility of re-hospitalization (re-hospitalization rate) and provide it by visualizing it, by determining whether the value of at least one specific re-hospitalization factor among creatinine, blood urea nitrogen (BUN), calcium (Ca), sodium (Na), and potassium (K) is higher or lower than a preset standard (value) or standard range.

[0096] In the above re-hospitalization analysis prediction unit (400), the value of the re-hospitalization factor specific factor may be the value of the re-hospitalization factor specific factor linked to the administered drug and dosage administered not only to the patient in question but also to other ischemic stroke patients.

[0097] In the present invention, the re-hospitalization analysis prediction unit (400) may be configured to learn and evaluate analysis using an artificial intelligence module including a deep learning model and a machine learning model.

[0098] For example, artificial neural networks (ANN), convolutional neural networks (CNN), and recurrent neural networks (RNN) can be used as deep learning models, and Random Forest, Gradient Boosting Machine (GBM), XGBoost, Support Vector Machine (SVM), and k-Nearest Neighbors (KNN) can be used as machine learning models.

[0099] In other words, the rehospitalization analysis prediction unit (400) of the present invention can be configured to use an artificial intelligence module such as SHAP (Shapley Additive Explanations) to analyze specific rehospitalization factors, which are rehospitalization prediction variables, based on the entire patient group data (raw data of ischemic stroke patients and history data of rehospitalized patients), and to provide visualized influences on the risk of rehospitalization.

[0100] In addition, the re-hospitalization prediction and management system for an ischemic stroke patient according to the present invention may further include a post-monitoring unit (500) configured to provide relevant information to the ischemic stroke patient or a related guardian based on the result predicted by the re-hospitalization analysis prediction unit (400).

[0101] The above post-monitoring unit (500) can be provided to a terminal device (10) of an ischemic stroke patient through a known wired or wireless communication network.

[0102] In particular, the post-monitoring unit (500) may be configured to receive real-time data from a bio-information detection wearable device, such as a bio-information detection terminal device equipped by a patient with ischemic stroke, which can detect bio-information in real time and transmit and receive related data through a communication module, and to transmit management information (e.g., messages such as '00 value is high. Please manage your blood pressure', 'Please visit a hospital within 00 days') to the bio-information detection wearable device based on the result predicted by the re-hospitalization analysis prediction unit (400).

[0103] Meanwhile, although the above describes a method for predicting and managing the re-hospitalization of ischemic stroke patients, it goes without saying that a computer-readable recording medium storing a program for implementing the method for predicting and managing the re-hospitalization of ischemic stroke patients, and a program stored on the computer-readable recording medium for implementing the method for predicting and managing the re-hospitalization of ischemic stroke patients, can also be implemented.

[0104] In other words, those skilled in the art will readily understand that the above-described method for predicting and managing the re-hospitalization of ischemic stroke patients may be provided by being tangibly implemented as a program of instructions for implementing it, and included in a computer-readable recording medium. That is to say, it may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable recording medium. The computer-readable recording medium may include program instructions, data files, data structures, etc., either individually or in combination. The program instructions recorded on the computer-readable recording medium may be those specifically designed and configured for the present invention, or they may be those known and available to those skilled in computer software. Examples of the computer-readable recording medium include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical media such as CD-ROMs and DVDs; and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, flash memory, and USB memory. The above computer-readable recording medium may be a transmission medium, such as an optical or metallic wire or a waveguide, that includes a carrier wave transmitting a signal specifying program instructions, data structures, etc. Examples of program instructions include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc. The above hardware device may be configured to operate as one or more software modules to perform the operation of the present invention, and vice versa.

[0105] Meanwhile, the inventor of the present invention explains the technical background regarding the usefulness of the method for predicting and managing the re-hospitalization of ischemic stroke patients and the system for predicting and managing the re-hospitalization of ischemic stroke patients according to the present invention, as well as the derivation of specific factor(s) for re-hospitalization prediction.

[0106] The inventor(s) of the present invention extracted admission and readmission records of 4,577 patients who were primarily hospitalized for ischemic stroke (ICD-10 I63) at Kangwon National University Hospital from August 2003 to August 2021. The hospital is home to the Gangwon Regional Cardiovascular and Cerebrovascular Center and is a nationally designated hospital serving as a hub for cardiovascular treatment and rehabilitation for the entire Gangwon region. Among these patients, information on 1,136 patients was analyzed after excluding those who underwent blood tests during hospitalization and had missing values ​​in their blood test indicators.

[0107] To predict the likelihood of readmission for stroke and related diseases in hospitalized patients with ischemic stroke, Electronic Health Record (EHR) data at the time of admission and medical records after discharge were utilized. The EHR data at admission included demographic characteristics, blood test results, treatment, and underlying disease. For underlying disease, previous medical records from the relevant hospital were used to define the variable.

[0108] The target variable is rehospitalization due to stroke and stroke-related diseases. The selection of ischemic stroke-related diseases was performed by neurologists and family medicine specialists according to academic and clinical criteria, respectively (see Table 1). Table 1 is the rehospitalization index: stroke and related diseases.

[0109] Classification ICD Classification and Disease Names Stroke [I63] Cerebral infarction [I60] Subarachnoid hemorrhage [G45] Transient cerebral ischemic attacks and related syndromes Diseases related with Stroke [I20] Angina pectoris [G46] Vascular syndromes of brain in cerebrovascular diseases [J17] Pneumonia in diseases classified elsewhere [I65] Occlusion and stenosis of precerebral arteries, not resulting in cerebral infarction [I67] Other cerebrovascular diseases [I25] Chronic ischemic heart disease [I50] Heart failure [I21] Acute myocardial infarction [I48] Atrial fibrillation and flutter [L89] Pressure ulcer [I46] Cardiac arrest [G40] Epilepsy [R56] Convulsions, not elsewhere classified [R00] Abnormalities of heart beat [I74] Arterial embolism and thrombosis [I61] Intracerebral hemorrhage [G41] Status epilepticus [R57] Shock, not elsewhere classified [I44] Atrioventricular and left bundle-branch block [I47] Paroxysmal tachycardia [R13] Dysphagia

[0110] The inventors used four major categories extractable from Common Data Model (CDM) electronic medical record (EMR) data as features. The raw data was pre-processed for machine learning analysis. The selection of relevant variables was performed by a neurologist and a family medicine specialist according to academic and clinical criteria, respectively. Subsequently, variables with low feature importance were excluded using the backward elimination method. Finally, the top 22 features were used for analysis (see Table 2). Table 2 shows the top 22 features associated with re-hospitalization within 90 days of stroke.

[0111] Not readmitted(n=940) Readmitted(n=196) T-statistic / F-statistic P-value Variable Mean SD Mean SD age 79.551 13.150 80.117 12.826 0.560 0.576 Creatinine 1.042 0.694 1.171 0.892 1.910 0.057 Thrombin time 12.653 2.544 12.842 3.049 0.814 0.417 Potassium 4.191 0.518 4.283 0.656 1.830 0.069 Sodium 139.901 3.896 139.192 3.849 -2.340 0.020 Chloride 105.440 4.770 105.498 5.053 0.146 0.884 BUN 18.765 11.123 20.228 12.851 1.481 0.140 Albumin 4.001 0.482 3.917 0.524 -2.059 0.040 Calcium 9.116 0.544 9.006 0.504 -2.751 0.006 MCV 91.592 5.592 91.650 5.763 0.129 0.897 Hemoglobin 13.131 2.005 12.821 2.134 -1.871 0.062 Leukocytes 7.753 3.601 7.836 3.469 0.305 0.761 Number of underlying disease 1.003 1.497 1.337 1.897 2.316 0.021 Thrombolysis 3.885 5.485 3.628 4.307 -0.724 0.470 Nasogastric feeding 1.196 9.808 0.806 4.610 -0.849 0.396 Sex Male 542 57.7% 129 65.8% 4.473 0.035 Female 398 42.3% 67 34.2% I63:Cerebral infarction No 862 91.7% 181 92.3% 0.090 0.765 Yes 78 8.3% 15 7.7% I67: Other cerebrovascular diseases No 900 95.7% 189 96.4% 0.191 0.662 Yes 40 4.3% 7 3.6% J17: Pneumonia No 935 99.5% 192 98.0% 4.709 0.030 Yes 5 0.5% 4 2.0% I20: Angina pectoris No 929 98.8% 195 99.5% 0.675 0.411 Yes 11 1.2% 1 0.5% F03: Unspecified dementia No 930 98.9% 194 99.0% 0.003 0.957 Yes 10 1.1% 2 1.0% G54: Nerve root and plexus disorders No 931 99.0% 194 99.0% 0.007 0.935 Yes 9 1.0% 2 1.0%

[0112] Sociodemographic characteristics included sex and age. Blood test results performed during hospitalization included Creatinine, Thrombin time, Potassium, Sodium, Chloride, BUN (Blood Urea Nitrogen), Albumin [ref], Calcium, MCV (mean corpuscular volume), Hemoglobin, and Leukocytes; the highest value among multiple measurements taken during hospitalization was used. Treatment details during hospitalization included the presence of thrombolysis (an acute treatment for ischemic stroke) and the insertion of a nasogastric tube. Underlying diseases included in the analysis were I63: Cerebral infarction; I67: Other cerebrovascular diseases; J17: Pneumonia; I20: Angina pectoris; F03: Unspecified dementia; G54: Nerve root and plexus disorders. Chi-square tests and ANOVA were performed between each characteristic and the label (90-day readmission).

[0113] The generalization performance of the models was evaluated by performing cross-validation on the dataset divided into three folds. Through this process, the average performance of each model was derived. Model performance was comprehensively evaluated using metrics such as AUC (Area Under the ROC Curve), Precision, Recall, F1 Score, and Accuracy. SHAP values ​​were used to interpret the model's prediction results and assess the importance of each variable. This allowed for the visualization and analysis of the impact of each variable on model predictions. The SHAP Summary Plot indicates how much a specific characteristic contributes positively or negatively to the model's predictions. The results derived from the Shapley force plot of the Light GBM model allow for the identification of which variables influence readmission predictions for each patient.

[0114] From the above results, it was confirmed that it provides higher predictive performance than Lineback et al. (0.64, 95% CI 0.62-0.66), Darabi et al. (0.74, 95% CI 0.64-0.78), and Ji Lv et al. (0.80, 95% CI 0.68-0.80), which predicted 30-day readmission of stroke patients discharged.

[0115] Figure 4 is a plot of the AUROC (Area Under the Receiver Operating Characteristic Curve) of the model for the test set, Figure 5 is a plot of the SHAP summary plot of the best model (Light GBM), and Figure 6 is a graph showing patients with (A) the highest and (B) the lowest SHAP values.

[0116] Risk factors for re-hospitalization within 90 days of stroke admission were analyzed at the cohort level. As shown in Figure 4, it was confirmed that renal function (Creatinine), number of underlying diseases (count), sodium levels, thrombolysis, and blood urea nitrogen (BUN) levels were the top five risk factors contributing to re-hospitalization 90 days after stroke.

[0117] Patient A's SHAP value was 0.98, with hemoglobin and chloride levels lower than the normal range, mean erythrocyte volume (MCV) within the normal range, and blood creatinine levels exceeding the normal range. On the other hand, Patient B's SHAP value was -5.53, with BUN and sodium levels higher than the normal range, while calcium was within the normal range.

[0118] Pre-discharge blood test results were found to be variables that play an important role in predicting readmission. In particular, renal and metabolic variables such as creatinine, urea nitrogen (BUN), calcium, sodium, and potassium showed a high correlation with readmission, and the number of underlying diseases was also identified as a major variable closely associated with readmission.

[0119] According to the method for predicting and managing the re-hospitalization of an ischemic stroke patient, the system for predicting and managing the re-hospitalization of an ischemic stroke patient, and the medium storing a computer-readable program for executing the method for predicting and managing the re-hospitalization of an ischemic stroke patient according to the present invention as described above, it is possible to provide a novel technology capable of predicting the re-hospitalization of a patient diagnosed with and treated for ischemic stroke, thereby contributing to the advancement of related technological fields. In particular, there is an advantage in that it can be extended from risk diagnosis for ischemic stroke to promote systematic management following the diagnosis and treatment of ischemic encephaloma.

[0120] In addition, according to the present invention, post-treatment care for ischemic stroke patients can be performed more effectively and systematically through the prediction and management of re-hospitalization based on specific factors for re-hospitalization for each patient, and there is an advantage of supporting rapid and effective medical treatment by taking prompt action for the patient based on specific factor(s) for re-hospitalization for each patient, such as creatinine levels, and through visual results.

[0121] In addition, according to the present invention, detailed post-discharge management plans for high-risk patients can be established through patient-specific customized reports, thereby supporting customized services such as rehabilitation plans, outpatient appointment schedules, home visits, and telephone monitoring. Furthermore, by identifying patients at high risk of readmission and providing intensive follow-up care, there is an advantage of reducing unnecessary hospitalizations and efficiently utilizing medical resources.

[0122] Meanwhile, the method for predicting and managing the re-hospitalization of an ischemic stroke patient, the system for predicting the re-hospitalization of an ischemic stroke patient, and the medium storing a computer-readable program for executing the method for predicting and managing the re-hospitalization of an ischemic stroke patient according to the present invention described above can be linked with Customer Data Management (CDM)-based data, so it can be applied to predicting the re-hospitalization of other chronic diseases such as kidney disease and cardiovascular disease, and can be utilized in various clinical fields by adding input variables suitable for the characteristics of each disease. Furthermore, through CDM-based data standardization, data can be integrated and analyzed across multiple institutions, thereby enhancing the generalization performance of the prediction model and enabling application to various patient groups.

[0123] The embodiments described in this specification and the accompanying drawings are merely illustrative of a part of the technical concept included in the present invention. Accordingly, since the embodiments disclosed in this specification are intended to explain, not limit, the technical concept of the present invention, it is obvious that the scope of the technical concept of the present invention is not limited by these embodiments. All variations and specific embodiments that can be easily deduced by a person skilled in the art within the scope of the technical concept included in the specification and drawings of the present invention should be interpreted as being included within the scope of the rights of the present invention. Explanation of the symbols

[0125] S100: Raw data construction phase S200: Step to establish readmission patient history information S300: Step for extracting specific factors for re-hospitalization. S400: Re-hospitalization prediction step S500: Post-monitoring stage 100: Raw data database section 200: Readmission Patient History Database 300: Extraction unit for specific factors of re-hospitalization causes 400: Re-admission Analysis Prediction Unit 500: Post-monitoring department

Claims

Claim 1 A method for predicting and managing the possibility of re-hospitalization of an ischemic stroke patient diagnosed with and treated for ischemic stroke, comprising: a raw data construction step of constructing personal information and treatment information of the ischemic stroke patient in a raw data database; a re-hospitalization patient history information construction step of constructing re-hospitalization information including personal information and treatment information of a re-hospitalized patient among the ischemic stroke patients in the constructed raw data in a history information database; a re-hospitalization factor specific factor extraction step of extracting one or more re-hospitalization factor specific factors from the re-hospitalization information based on the constructed re-hospitalization information through a re-hospitalization factor specific factor extraction unit; and a re-hospitalization prediction step of analyzing and predicting whether the patient will be re-hospitalized based on the extracted re-hospitalization factor specific factors through a re-hospitalization analysis prediction unit. Claim 2 A method for predicting and managing re-hospitalization of a patient with ischemic stroke according to claim 1, wherein the data constructed in the raw data construction step includes personal information such as gender, age, smoking status, alcohol consumption, obesity level, and genetic correlation, and treatment information such as blood test results, comorbidities, cholesterol levels, embolism, thrombosis, blood pressure, arteriosclerosis, hypertension, diabetes, administered drugs and dosages, and nasogastric tube insertion. Claim 3 A method for predicting and managing re-hospitalization of a patient with ischemic stroke according to claim 1 or 2, wherein the blood test values ​​in the treatment information are values ​​of creatinine, blood urea nitrogen (BUN), calcium (Ca), sodium (Na), potassium (K), hemoglobin (Hb), albumin, mean corpuscular volume (MCV), leukocytes, chloride (Cl), and thrombin time (TT). Claim 4 A method for predicting and managing rehospitalization of a patient with ischemic stroke, characterized in that, in claim 1 or 2, the step of extracting specific factors for rehospitalization factors comprises extracting kidney and metabolism-related factors. Claim 5 A method for predicting and managing rehospitalization of a patient with ischemic stroke according to claim 1, wherein the step of extracting specific rehospitalization factor factors comprises extracting creatinine, blood urea nitrogen (BUN), calcium (Ca), sodium (Na), potassium (K), and their respective values ​​as specific rehospitalization factor factors. Claim 6 A method for predicting and managing re-hospitalization of a patient with ischemic stroke according to claim 5, wherein the re-hospitalization prediction step is characterized by determining that there is a high probability of re-hospitalization when the value of at least one specific re-hospitalization factor deviates from a preset reference value or reference range. Claim 7 A method for predicting and managing re-hospitalization of an ischemic stroke patient according to claim 6, wherein the re-hospitalization prediction step is configured to visualize and provide the possibility of re-hospitalization by determining whether the numerical value of the re-hospitalization factor specific factor is higher or lower than a preset reference value or reference range. Claim 8 A method for predicting and managing re-hospitalization of an ischemic stroke patient, characterized in that, in any one of claims 5 to 7, the re-hospitalization prediction step is performed by analyzing and predicting using an artificial intelligence module of a deep learning model and a machine learning model. Claim 9 A medium storing a computer-readable program for executing a method for predicting and managing the re-hospitalization of a patient with ischemic stroke according to claim 1 or 2. Claim 10 A system for predicting and managing the possibility of re-hospitalization of an ischemic stroke patient diagnosed with and treated for ischemic stroke, comprising: a raw data database unit in which personal information and treatment information of the ischemic stroke patient are constructed; a history information database unit in which re-hospitalization information including personal information and treatment information of a re-hospitalized patient among the ischemic stroke patients is constructed from the constructed raw data; a re-hospitalization factor specific factor extraction unit configured to extract one or more re-hospitalization factor specific factors from the re-hospitalization information based on the constructed re-hospitalization information; and a re-hospitalization analysis prediction unit configured to analyze and predict whether the patient will be re-hospitalized based on the extracted re-hospitalization factor specific factors. Claim 11 A system for predicting and managing the re-hospitalization of ischemic stroke patients according to claim 10, wherein the raw data database unit includes data as personal information such as gender, age, smoking status, alcohol consumption, obesity level, and genetic correlation, and includes data as treatment information such as blood test results, comorbidities, cholesterol levels, embolism, thrombosis, blood pressure, arteriosclerosis, hypertension, diabetes, administered drugs and dosages, and nasogastric tube insertion status. Claim 12 A system for predicting and managing the re-hospitalization of patients with ischemic stroke according to claim 10, wherein the above treatment information includes blood test values, and the above blood test values ​​are values ​​of creatinine, blood urea nitrogen (BUN), calcium (Ca), sodium (Na), potassium (K), hemoglobin (Hb), albumin, mean corpuscular volume (MCV), leukocytes, chloride (Cl), and thrombin time (TT). Claim 13 A system for predicting and managing the re-hospitalization of ischemic stroke patients, characterized in that, in item 10, the re-hospitalization factor specific factor extraction unit comprises extracting kidney and metabolism-related factors. Claim 14 In claim 10, the re-hospitalization factor specific factor extraction unit is characterized by extracting creatinine, blood urea nitrogen (BUN), calcium (Ca), sodium (Na), potassium (K), and their respective values ​​as re-hospitalization factor specific factors, in a re-hospitalization prediction and management system for patients with ischemic stroke. Claim 15 A system for predicting and managing re-hospitalization of ischemic stroke patients according to claim 14, wherein the re-hospitalization analysis prediction unit is configured to determine that there is a high probability of re-hospitalization when the value of at least one specific re-hospitalization factor deviates from a preset reference value or reference range. Claim 16 In claim 15, the re-hospitalization analysis prediction unit is characterized by determining whether the numerical value of the re-hospitalization factor specific factor is higher or lower than a preset reference value or reference range, and visualizing and providing the possibility of re-hospitalization, thereby providing a re-hospitalization prediction and management system for patients with ischemic stroke. Claim 17 A system for predicting and managing the re-hospitalization of an ischemic stroke patient, characterized in that, in any one of claims 14 to 16, the re-hospitalization analysis prediction unit is configured to analyze and predict using an artificial intelligence module of a deep learning model and a machine learning model, and further comprises a post-monitoring unit configured to provide relevant information to the ischemic stroke patient or a related guardian based on the results predicted by the re-hospitalization analysis prediction unit.

Images