KR102287968B1 - Method for predicting of bacteremia risk and device for predicting of bacteremia risk using the same - Google Patents

Abstract

In the present specification, as a method of predicting the risk of developing bacteremia implemented by a processor, receiving biological test data for an individual, using a bacteremia prediction model configured to predict the risk of developing bacteremia for an individual based on the biological test data estimating the individual's risk of developing bacteremia, and the biological test data are creatine (Creatinine) level, albumin (Albumin) level, CRP (Creactive protein) level, WBC (white blood cell) minimum level, WBC maximum level, Provided are a method for predicting the risk of developing bacteremia, including at least one of a platelet count, a prothrombin time, and an alkaline phosphatase (ALP) level, and a device for predicting the risk of developing bacteremia using the same.

Description

Method for predicting the risk of developing bacteremia and a device for predicting the risk of developing bacteremia using the same

The present invention relates to a method and a device for predicting the risk of developing bacteremia, and more particularly, to a method and a device using the same for predicting and providing the risk of developing bacteremia based on various clinical data obtained from an individual.

Bacteremia may refer to a clinical condition in which bacteria enter the blood and circulate throughout the body. Such bacteremia can occur when the bacteria multiply at a rate that exceeds the clearing ability of the reticuloendothelial system.

In this case, bacteremia may be divided into transient bacteremia, intermittent bacteremia, and persistent bacteremia according to clinical aspects. More specifically, transient bacteremia is bacteremia that occurs early in systemic or local infection, and may be accompanied by meningitis, pneumonia, suppurative arthritis, osteomyelitis, gonococcal or meningococcal infection. In addition, intermittent bacteremia can occur when there is an abscess in the abdominal cavity, pelvic cavity, perirenal, liver, prostate, etc., and persistent bacteremia can occur when bacterial endocarditis and intravascular catheter infection. It may accompany cellulosis.

On the other hand, as a treatment method for bacteremia, there may be the use of antibiotics, and the empirical use of appropriate antibiotics at an early stage can prevent bacteremia from progressing to sepsis and septic shock.

Therefore, predicting bacteremia before the onset of bacteremia may be of great clinical importance.

Accordingly, there is a continuous need for the development of a new bacteremia risk prediction system that can predict bacteremia before it occurs and take quick measures.

The description underlying the invention has been prepared to facilitate understanding of the invention. It should not be construed as an admission that the matters described in the background technology of the invention exist as prior art.
Prior literature: Bernard Hernanderz et al. 7, Supervised learning for infection risk inference using pathology data, BMC Medical Informatics and Decision Making (2017) 17:168

On the other hand, the inventors of the present invention noted that changes in biosignals will precede as a part of the physiological response of the human body in advance before a situation in which a morbidity of any disease, such as bacteremia, is clinically suspected.

More specifically, the inventors of the present invention noted that changes in various clinical data obtainable in a general ward or intensive care unit may be associated with a potential disease such as bacteremia.

In particular, the inventors of the present invention have noted that bacteremia is a disease that can be caused secondary by an external environment, and not only biological test data of analysis results of biological samples such as blood isolated from a patient, but also various biosignal data , it could be recognized that the data related to medical treatment were related to the development of bacteremia.

Furthermore, the inventors of the present invention tried to search for biological characteristics that increase the accuracy of prediction of bacteremia, which is distinguished from sepsis, which causes intoxication symptoms such as fever and low blood pressure by proliferation of bacteria in the blood.

As a result, the inventors of the present invention were able to discover that among the biological test data obtained from the subject, in particular, ALP (Alkaline phosphatase) and platelet levels were associated with the onset of bacteremia.

Accordingly, the inventors of the present invention can predict the risk of developing bacteremia with high reliability and accuracy by further considering biosignal data and clinical characteristic data together with biological test data including ALP (Alkaline phosphatase) level and/or platelet level. We have developed a system for predicting the risk of developing bacteremia.

Meanwhile, the inventors of the present invention were able to apply the learned predictive model to predict bacteremia based on biological test data, further biosignal data, and clinical feature data to a new bacteremia risk prediction system.

In this regard, the inventors of the present invention performed evaluation to determine clinical data that has a large influence on predicting bacteremia among various clinical data or clinical data that has a large influence on predicting normal rather than bacteremia, and It was intended to be applied to learning. Through this, the inventors of the present invention could expect to improve the diagnostic ability of the predictive model to predict bacteremia.

As a result, a new bacteremia risk prediction system based on a bacteremia prediction model could provide accurate and reliable bacteremia diagnosis information for an individual.

Furthermore, the inventors of the present invention expect that, through the development of such a bacteremia risk prediction system, the patient's condition can be quickly sensed and the risk of bacteremia development can be recognized in advance, and the treatment time for the patient can be advanced to improve the treatment performance of bacteremia. could Furthermore, the inventors of the present invention could further expect that the development of a system for predicting the risk of developing bacteremia could provide the effects of increasing the survival rate of patients, preventing complications, and reducing treatment costs.

In addition, the inventors of the present invention, if bacteremia is predicted by the predictive model for the individual, so that the guardian or medical staff can more easily recognize the high-risk group for the occurrence of bacteremia for the individual requiring continuous monitoring, such as a critically ill patient, alarm provided to inform the risk of developing bacteremia.

Accordingly, the problem to be solved by the present invention is to use a bacteremia prediction model configured to predict the risk of developing bacteremia based on the received biological test data, further biosignal data, and clinical characteristic data for the individual, To provide a method for predicting the risk of developing bacteremia, configured to predict the risk of developing bacteremia.

Another problem to be solved by the present invention is a receiving unit configured to receive biological test data for the subject, further biosignal data, clinical characteristic data, and the risk of developing bacteremia in the subject based on the predictive model and connected to communicate with the receiving unit It is to provide a device for predicting the risk of developing bacteremia, including a processor configured to predict.

The problems of the present invention are not limited to the problems mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the following description.

In order to solve the problems as described above, a method for predicting the risk of developing bacteremia according to an embodiment of the present invention is provided. At this time, the bacteremia prediction method of the present invention is configured to predict the risk of developing bacteremia implemented by the processor, receiving biological test data for the individual, and predicting the risk of developing bacteremia for the individual based on the biological test data predicting the risk of developing bacteremia in an individual using the constructed bacteremia prediction model.

In this case, the biological test data include Creatinine level, Albumin level, CRP (Creactive protein) level, WBC (white blood cell) minimum, WBC maximum, Platelet count, Prothrombin time and Includes alkaline phosphatase (ALP) levels.

According to a feature of the present invention, the method for predicting bacteremia of the present invention may further include receiving biosignal data for an individual, and the bacteremia prediction model predicts the risk of developing bacteremia based on biological test data and biosignal data It can be further configured to do so. In this case, the biosignal data may be at least one of systolic blood pressure (SBP), diastolic blood pressure (DBP), maximum body temperature, minimum body temperature, heart rate, and respiration rate.

According to another feature of the present invention, the bacteremia prediction method of the present invention further comprises the step of receiving clinical characteristic data for the subject, and the bacteremia prediction model determines the risk of developing bacteremia based on the biological test data and the clinical characteristic data. It may be further configured to predict. In this case, the clinical characteristic data may be at least one of gender, age, blood culture period, ICU (intensive care unit) treatment, central venous catheter insertion, steroid treatment, and antibiotic treatment.

According to another feature of the present invention, the biological test data may include creatine level, albumin level, CRP level, WBC minimum level, WBC maximum level, prothrombin time, and platelet count, ALP level.

According to another feature of the present invention, the biological test data are ALT (alanine aminotransferase) level, thromboplastin clotting time, AST (aspartate aminotransferase) level, CRP (C-reactive protein) level, erythrocyte sedimentation rate, FERR (ferritin) level, HMG (hemoglobin) level, PTINR (Prothrombin Time International Normalized Ratio), PTPER (Prothrombin Time (%)), PTSEC (Prothrombin Time (sec)), Red blood cell count, TBIL (Total bilirubin) level, TCO ₂ (total carbon dioxide) and at least one of a white blood cell count

According to another feature of the present invention, the bacteremia may be acute severe bacteraemia.

According to another feature of the present invention, the bacteremia prediction model is an ensemble model consisting of a combination of multi-layer perceptron (MLP), support vector machine (SVM), random forest (RF), and a plurality of models, pre-trained ( It may be at least one of a pre-trained model and an Xgboost model.

According to another feature of the present invention, the bacteremia prediction model may be an Xgboost model. In this case, the MLP model may include a single hidden layer having 128 nodes or 256 nodes, or a plurality of hidden layers in which a single hidden layer exists in plurality.

According to another feature of the present invention, the bacteremia prediction model comprises the steps of: receiving learning data consisting of biological experimental data, biosignal data, and clinical characteristic data for a bacteremia-infected sample object and a normal sample object, and the learning data As a basis, it can be a model trained through the steps to predict bacteremia or normality. In this case, the normal sample subject may be an individual evaluated as not clinically bacteremia and not bacteremia.

According to another feature of the present invention, the bacteremia prediction method of the present invention may further include the step of evaluating the learning data performed after the step of receiving the learning data. At this time, the step of evaluating the learning data includes calculating a relevance score with respect to the learning data, and determining the bacteremia-related learning data within a predetermined ranking based on the relevance score. may include

According to another feature of the present invention, the step of evaluating the learning data is OOS (one-out search), LRP (Layer-wise Relevance Propagation), GI (gradient input), IG (integrated gradients), and SM (saliency) map) based on at least one algorithm of, calculating a relevance score for the learning data with respect to bacteremia, and based on the relevance score, determining data for bacteremia-related learning within a predetermined ranking. can

According to another feature of the present invention, the method for predicting bacteremia of the present invention may further include providing a predicted risk of developing bacteremia for an individual.

According to another feature of the present invention, the step of providing the risk of developing bacteremia may include providing a risk notification when the risk of developing bacteremia for an individual is predicted by the bacteremia prediction model.

According to another feature of the present invention, the step of providing the risk of developing bacteremia may include calculating a risk probability of developing bacteremia for an individual by a bacteremia prediction model, and providing the risk probability of developing bacteremia.

In order to solve the above problems, there is provided a device for predicting the risk of developing bacteremia according to another embodiment of the present invention. In this case, the device for predicting bacteremia of the present invention includes a receiver configured to receive biological test data for an individual, and a processor connected to the receiver. In this case, the processor is configured to predict the risk of developing bacteremia in the subject by using the bacteremia prediction model configured to predict the risk of developing bacteremia in the subject based on the biological test data. Furthermore, biological test data include Creatinine level, Albumin level, CRP (Creactive protein) level, WBC (white blood cell) minimum, WBC maximum, Platelet count, Prothrombin time. and Alkaline phosphatase (ALP) levels.

According to a feature of the present invention, the receiving unit is further configured to receive biosignal data for the subject, and the bacteremia prediction model may be further configured to predict the risk of developing bacteremia based on the biological test data and the biosignal data. In this case, the biosignal data may be at least one of systolic blood pressure (SBP), diastolic blood pressure (DBP), maximum body temperature, minimum body temperature, heart rate, and respiration rate.

According to another feature of the present invention, the receiving unit is further configured to receive clinical characteristic data for the subject, and the bacteremia prediction model may be further configured to predict the risk of developing bacteremia based on the biological test data and the clinical characteristic data. . In this case, the clinical characteristic data may be at least one of gender, age, blood culture period, ICU (intensive care unit) treatment, central venous catheter insertion, steroid treatment, and antibiotic treatment.

According to another feature of the present invention, the biological test data may include creatine level, albumin level, CRP level, WBC minimum level, WBC maximum level, prothrombin time, and platelet count, ALP level.

According to another feature of the present invention, the biological test data are: ALT level, thrompoplastin clotting time, AST level, CRP level, erythrocyte sedimentation rate, ferritin level, hemoglobin level, PTINR, PTPER, PTSEC, red blood cell It may further include at least one of a level, a total bilirubin level, a TCO _{2 level, and a white blood cell level.}

According to another feature of the present invention, the bacteremia may be acute severe bacteraemia.

According to another feature of the present invention, the bacteremia prediction model is an ensemble model consisting of a combination of multi-layer perceptron (MLP), support vector machine (SVM), random forest (RF), and a plurality of models, pre-trained ( It may be at least one of a pre-trained model and an Xgboost model.

According to another feature of the present invention, the bacteremia prediction model may be an MLP model or an Xgboost model. In this case, the MLP model may include a single hidden layer having 128 nodes or 256 nodes, or a plurality of hidden layers in which a single hidden layer exists in plurality.

According to another feature of the present invention, the bacteremia prediction model receives learning data consisting of biological experimental data, biosignal data, and clinical characteristic data for a bacteremia-infected sample object and a normal sample object, and based on the learning data, , it can be a model trained through the steps to predict bacteremia or normality. In this case, the normal sample subject may be an individual evaluated as not clinically bacteremia and not bacteremia.

According to another feature of the present invention, it may further include an output unit configured to provide a predicted risk of developing bacteremia for the subject.

According to another feature of the present invention, the output unit may be configured to provide a risk notification when the risk of developing bacteremia for an individual is predicted by the bacteremia prediction model.

According to another feature of the present invention, the bacteremia prediction model may be further configured to calculate a risk probability of developing bacteremia for an individual. Furthermore, the output may be further configured to provide a probability of developing a bacteremia risk.

Hereinafter, the present invention will be described in more detail through examples. However, since these examples are merely for illustrative purposes of the present invention, the scope of the present invention should not be construed as being limited by these examples.

The present invention receives pharmacobiological test data, further biosignal data, and clinical characteristic data for an individual, and provides a system for predicting the risk of developing bacteremia based on the received data, thereby rapidly detecting the risk of developing bacteremia in an individual, Relevant information can be provided.

In addition, the present invention provides a prediction system to which a prediction model learned to predict bacteremia based on biological test data, further biosignal data, and clinical characteristic data is applied, thereby providing predictive information of the onset of bacteremia with high accuracy and reliability. can

More specifically, in the present invention, evaluation was performed to determine clinical data that has a large influence on predicting bacteremia among various clinical data or clinical data that has a large influence on predicting a normal state other than bacteremia, and By applying it to learning, it is possible to provide a predictive system with improved diagnostic ability of predicting bacteremia.

Accordingly, the present invention has the effect of predicting the risk of onset with high reliability and accuracy for bacteremia that is distinct from sepsis or different from sepsis.

In addition, according to the present invention, when bacteremia is predicted by a predictive model for an individual, it is configured to provide an alarm to inform the risk of developing bacteremia. It is possible to recognize high-risk groups for developing bacteremia.

Therefore, the present invention can provide a good prognosis for treatment by advancing the treatment time for bacteremia.

Furthermore, the present invention can provide the effects of increasing the survival rate of an individual with bacteremia, preventing complications, and reducing the cost of treatment.

The effect according to the present invention is not limited by the contents exemplified above, and more various effects are included in the present specification.

1A exemplarily illustrates a bacteremia risk prediction system based on a method and a device for predicting the risk of developing bacteremia according to an embodiment of the present invention.
1B exemplarily shows the configuration of a device for predicting the risk of developing bacteremia according to an embodiment of the present invention.
2 exemplarily shows a procedure of a method for predicting the risk of developing bacteremia according to an embodiment of the present invention.
3A and 3B exemplarily show training data of a bacteremia prediction model used in various embodiments of the present invention.
3C exemplarily shows the configuration of a bacteremia prediction model used in various embodiments of the present invention.
FIG. 3D exemplarily illustrates an evaluation method of a bacteremia prediction model used in various embodiments of the present invention.
4A and 4B show a first evaluation result for a bacteremia prediction model used in various embodiments of the present invention.
4c and 4d show clinical data according to the influence of bacteremia prediction on the bacteremia prediction model used in various embodiments of the present invention.
4e and 4f show performance changes according to changes in clinical data according to the influence of bacteremia prediction for the bacteremia prediction model used in various embodiments of the present invention.
5A and 5B show learning data and evaluation data for the second evaluation of the bacteremia prediction model used in various embodiments of the present invention.
5c to 5i show a second evaluation result for the bacteremia prediction model used in various embodiments of the present invention.

Advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but will be implemented in various different forms, and only these embodiments allow the disclosure of the present invention to be complete, and common knowledge in the art to which the present invention pertains It is provided to fully inform those who have the scope of the invention, and the present invention is only defined by the scope of the claims.

The shapes, sizes, proportions, angles, numbers, etc. disclosed in the drawings for explaining the embodiments of the present invention are exemplary, and thus the present invention is not limited to the illustrated matters. In addition, in describing the present invention, if it is determined that a detailed description of a related known technology may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted. When 'including', 'having', 'consisting', etc. mentioned in this specification are used, other parts may be added unless 'only' is used. When a component is expressed in a singular, the case in which the plural is included is included unless otherwise explicitly stated.

In interpreting the components, it is interpreted as including an error range even if there is no separate explicit description.

Each feature of the various embodiments of the present invention may be partially or wholly combined or combined with each other, and technically various interlocking and driving are possible, as will be fully understood by those skilled in the art, and each embodiment may be independently implemented with respect to each other, It may be possible to implement together in a related relationship.

For clarity of interpretation of the present specification, terms used herein will be defined below.

As used herein, the term “bacteremia” refers to a clinical condition in which pathogens live and circulate in the bloodstream. At this time, bacteremia is different from 'sepsis', in which bacteria proliferate in the blood and cause poisoning symptoms such as fever and low blood pressure, and in the present specification, sepsis and bacteremia are distinguished.

As used herein, the term "individual" may refer to a subject for which the risk of developing bacteremia is to be predicted. On the other hand, in the present specification, an individual may be a 'patient' or a 'critical patient', but is not limited thereto and may include any target for predicting the risk of developing bacteremia.

At this time, the bacteremia-infected subject is Bacillus spp. may be different.

As used herein, the term “biological test data” may refer to clinical data on a biological sample isolated from a subject, for example, blood.

More specifically, the biological test data disclosed in the present specification are, creatine (Creatinine) level, albumin (Albumin) level, CRP (Creactive protein) level, WBC (white blood cell) minimum level, WBC maximum level, platelet (Platelet) level , prothrombin time, and alkaline phosphatase (ALP) levels.

Preferably, the biological test data may be an ALP level or a platelet level, but is not limited thereto.

As used herein, the term “biological signal data” refers to vital signs and vital signs, and may refer to data related to the condition of an individual. Such biosignal data may be correlated with the individual's risk of developing bacteremia.

In this case, the biosignal data may be body temperature, pulse, oxygen saturation, systolic blood pressure (SBP), diastolic blood pressure (DBP), mean blood pressure, and respiration rate of the individual measured by the biosignal measuring device. Preferably, the biosignal data may be at least one of SBP, maximum body temperature, minimum body temperature, heart rate, and respiration rate. However, the present invention is not limited thereto, and the biosignal data may include various measurement data related to a health state of an individual.

As used herein, the term “clinical characteristic data” may encompass data on whether or not a drug is administered along with an individual's characteristics, for example, gender and age, and whether or not a medical treatment is performed.

On the other hand, the clinical characteristic data disclosed in the present specification is at least one of sex, age, blood culture period, ICU (intensive care unit) treatment, central venous catheter insertion, steroid treatment, and antibiotic treatment. can be However, the clinical characteristic data is not limited thereto, and may include more diverse data about a drug administered to the subject and medical treatment for the subject.

For example, the clinical characteristic data is data on whether at least one of a catheter, Foley, mechanical ventilator, restraint, and drainage device is medically treated. , Ultracet, Midazolam, Ultiva, Pofol, Ativan, Fentanyl, Precedex, IR codon, Tagine 20/10 (TARGIN 20/10), peridol, risperdal, zyprexa, seroquel, abilify, pethidine, durogesic patch, It may further include dosing data for morphine and mypol.

As used herein, the term "bacteremia prediction model" may be a model trained to predict the risk of developing bacteremia based on clinical data of at least one of biological test data, biosignal data, and clinical characteristic data.

For example, the bacteremia prediction model may be a model trained to predict the risk of developing bacteremia based on clinical data acquired every predetermined time from a sample subject with or without bacteremia.

In this case, the bacteremia prediction model may be a model trained to predict the risk of developing bacteremia based on all clinical data of biological test data, biosignal data, and clinical characteristic data. However, the data used for learning is not limited thereto, and a combination of more various clinical data may be used for learning the predictive model.

On the other hand, the bacteremia prediction model of the present invention, MLP (Multi-Layer Perceptron), SVM (support vector machine), RF (random forest), and an ensemble model consisting of a combination of MLP, SVM, RF, pre-trained (pre- trained) model and Xgboost model.

Preferably, the bacteremia prediction model of the present invention may be a prediction model based on one MLP (Multi-Layer Perceptron). In this case, the bacteremia prediction model of the present invention based on the MLP model may include a hidden layer having 128 nodes or 256 nodes.

For example, the bacteremia prediction model of the present invention includes an input layer to which biological test data, biosignal data, and clinical characteristic data are input, and an output layer that predicts bacteremia or normal non-bacteremia, and 128 nodes or 256 nodes between these layers. It may be a predictive model of a multi-layered layer structure in which a hidden layer having nodes exists.

More specifically, the bacteremia prediction model of the present invention can be configured to use 'rmsprop' as an optimization function to update parameters in learning, and as a function to determine the strength of the input value of various clinical data transmitted to the output value. It can be configured to use the 'relu' function.

However, the configuration of the bacteremia prediction model of the present invention is not limited thereto.

According to another embodiment of the present invention, the bacteremia prediction model of the present invention may be an MLP model having a single node of 128 nodes. Furthermore, the bacteremia prediction model of the present invention may be an MLP model having two hidden layers in which a hidden layer having 128 nodes has .

According to another embodiment of the present invention, the bacteremia prediction model of the present invention may be an Xgbooxt model having a gbtree layer.

However, the types of the bacteremia prediction model of the present invention are not limited to those described above. For example, the prediction model is a more diverse ANN (Artificial Neural Network) model, or DNN (Deep Neural Network), CNN Convolutional Neural Network), RNN (Recurrent Neural Network), DCNN (Deep Convolutional Neural Network), RBM ( Restricted Boltzmann Machine), DBN (Deep Belief Network), SSD (Single Shot Detector) model, it can also be a predictive model based on U-net.

On the other hand, in the learning of the bacteremia prediction model of the present invention, evaluation may be performed in which clinical data having a large influence on predicting normal, not bacteremia, is determined. At this time, the influence evaluation is to be performed based on at least one algorithm of OOS (one-out search), LRP (Layer-wise Relevance Propagation), GI (gradient input), IG (integrated gradients), and SM (saliency map). can

For example, by the at least one algorithm, among various clinical data of biological test data, biosignal data, and clinical characteristic data, clinical data having a large influence on predicting bacteremia or predicting normality other than bacteremia has an influence. Large clinical data can be determined.

As the clinical data having a large influence can be applied as input data to the learning of the predictive model, the predictive model may be a model with improved bacteremia predictive ability than other models.

Hereinafter, a device for predicting the risk of developing bacteremia and a system for predicting the risk of developing bacteremia using the device according to an embodiment of the present invention will be described in detail with reference to FIGS. 1A and 1B .

1A exemplarily illustrates a bacteremia risk prediction system based on a method and a device for predicting the risk of developing bacteremia according to an embodiment of the present invention. 1B exemplarily shows the configuration of a device for predicting the risk of developing bacteremia according to an embodiment of the present invention.

Referring to FIG. 1A , the bacteremia risk prediction system 1000 according to an embodiment of the present invention includes a device 100 for predicting the risk of developing bacteremia, and biological test data 310, obtained for an individual 200, It consists of biosignal data 320 , clinical data 300 including clinical characteristic data 330 , medical treatment instrument 400 , and medical staff device 500 .

In this case, the clinical data 300 may be data evaluated from the subject 200 at an arbitrary time point. However, it is not limited thereto.

More specifically, the device 100 for predicting the risk of developing bacteremia in the system 1000 for predicting the risk of developing bacteremia receives various clinical data 300 measured or evaluated for the subject 200, and based on this, the risk of developing bacteremia can be configured to predict.

In this case, the medical treatment device 400 may provide the subject 200 with at least one biosignal data 320 from the group consisting of temperature, pulse, oxygen saturation, systolic blood pressure, diastolic blood pressure, mean blood pressure, and respiration rate. It may be a biosignal measuring device.

Meanwhile, in the bacteremia risk prediction system 1000 , the biological test data 310 , and furthermore the clinical characteristic data 330 may be acquired from an external system such as an EMR (Electronic Medical Record) system.

More specifically, referring to FIG. 1B , the device 100 for predicting the risk of developing bacteremia includes a receiving unit 110 , an input unit 120 , an output unit 130 , a storage unit 140 , and a processor 150 .

Specifically, the receiver 110 may be configured to receive the clinical data 300 including the biological test data 310 , the biosignal data 320 , and the clinical characteristic data 330 for the subject 200 .

According to a feature of the present invention, the receiving unit 110 may be further configured to transmit, to the medical staff device 500 , a result predicted for the entity 200 determined by the processor 150 , which will be described later.

The input unit 120 is not limited to a keyboard, a mouse, a touch screen panel, and the like. The input unit 120 may set the device 100 for predicting the risk of developing bacteremia and instruct the operation of the device 100 for predicting the risk of developing bacteremia.

Meanwhile, the output unit 130 may display various clinical data 300 received by the receiving unit 110 . Furthermore, the output unit 130 may display information related to the diagnosis of bacteremia predicted by the processor 150 . Furthermore, the output unit 130 may be further configured to output a notification sound when it is determined by the processor 150 that the risk of developing bacteremia is high.

The storage unit 140 stores various clinical data 300 for the subject 200 received through the receiving unit 110 , and stores the instruction of the device 100 for predicting the risk of bacteremia onset through the input unit 120 . can be configured to Furthermore, the storage unit 140 is configured to store the bacteremia prediction information in the entity 200 generated by the processor 150, which will be described later. However, it is not limited to the above, and the storage unit 140 may store various pieces of information determined by the processor 150 for predicting the risk of developing bacteremia.

The processor 150 may be a component for providing an accurate prediction result of the device 110 for predicting the risk of developing bacteremia. In this case, in order to accurately predict the risk of developing bacteremia, the processor 150 may be based on a predictive model configured to predict the risk of developing bacteremia based on various clinical data 300 .

Meanwhile, the processor 150 may be configured to provide a notification to the medical staff device 500 when the subject 200 is predicted as a high-risk group for developing bacteremia by the bacteremia prediction model. Accordingly, the medical staff can take quick action on the subject 200 .

As described above, the system 1000 for predicting the risk of developing bacteremia of the present invention can provide early diagnosis of bacteremia and early treatment of bacteremia to the individual 200 , and thus can be applied to various medical systems.

Hereinafter, a method of predicting the risk of developing bacteremia according to an embodiment of the present invention will be described in detail with reference to FIG. 2 . 2 exemplarily shows a procedure of a method for predicting the risk of developing bacteremia according to an embodiment of the present invention.

Referring to FIG. 2 , in the method of predicting the risk of developing bacteremia according to an embodiment of the present invention, first receiving biological test data from an individual (S210), and predicting bacteremia configured to predict the risk of developing bacteremia based on the biological test data Using the model, the risk of developing bacteremia for an individual is predicted (S220). Finally, it provides the predicted risk of developing bacteremia for the individual (S230).

More specifically, in the step of receiving data (S210), creatine (Creatinine) level, albumin (Albumin) level, CRP (Creactive protein) level, WBC (white blood cell) minimum level, WBC maximum level, platelet (Platelet) Biological test data of at least one of water, Prothrombin time, and Alkaline phosphatase (ALP) levels may be received.

According to a feature of the present invention, in the step of receiving the data ( S210 ), biosignal data of at least one of a systolic blood pressure (SBP), a maximum body temperature, a minimum body temperature, a heart rate, and a respiration rate may be further received.

According to another feature of the present invention, in the step of receiving data (S210), gender, age, blood culture period, intensive care unit (ICU) treatment, central venous catheter insertion, steroid treatment, and At least one clinical characteristic data of whether antibiotic treatment is performed may be further received.

According to a feature of the present invention, the biosignal data received in the step of receiving the data ( S210 ) may be an ALP level or a platelet count, but is not limited thereto.

Next, in the step of predicting the risk of developing bacteremia ( S220 ), the subject may be determined to be normal, not bacteremia or bacteremia, by the bacteremia prediction model configured to predict the risk of developing bacteremia.

At this time, the bacteremia prediction model is trained to predict the risk of developing bacteremia based on at least one clinical data of biological test data, biosignal data, and clinical characteristic data, received in the step of receiving data ( S210 ). can be a model.

For example, the bacteremia prediction model receives training data of at least one of biological test data, biosignal data, and clinical evaluation data for a sample subject having bacteremia and a sample subject to normal, and based on the training data, bacteremia or It may be a model trained through the step of predicting or classifying normal.

Furthermore, the bacteremia prediction model calculates a relevance score with bacteremia for the learning data, and, based on the relevance score, determines the bacteremia-related learning data within a predetermined rank for bacteremia-related learning determined through the steps Based on the data, it may be a model configured to predict the risk of developing bacteremia.

According to another feature of the present invention, the bacteremia prediction model of the present invention is composed of a multi-layer perceptron (MLP), a support vector machine (SVM), a random forest (RF), and a combination of the MLP, the SVM, and the RF. It may be at least one model among ensemble models.

According to another feature of the present invention, the bacteremia prediction model may be an MLP model consisting of a hidden layer having 128 nodes or 256 nodes, but is not limited thereto.

Finally, in the step of providing the risk of developing bacteremia ( S230 ), information on the diagnosis of bacteremia predicted in the step of predicting the risk of developing bacteremia ( S220 ) may be provided.

According to another feature of the present invention, in the step of providing the risk of developing bacteremia (S230), the risk of developing bacteremia for the individual is predicted by the bacteremia prediction model in the step (S220) of predicting the risk of developing bacteremia. A warning of the risk of developing bacteremia may be provided.

On the other hand, in the step (S230) of providing the risk of developing bacteremia, a variety of information other than those described above may be provided. For example, in the step of providing the risk of developing bacteremia ( S230 ), additional treatment information that may be effective in correcting bacteremia may be further provided.

According to the above procedure, the method for predicting the risk of developing bacteremia according to an embodiment of the present invention determines whether or not the onset of bacteremia for an individual in real time or predicts the risk of onset of bacteremia in advance, information on the risk of developing bacteremia, furthermore, bacteremia It can provide an alarm to notify the high-risk group. Accordingly, the medical staff may be able to diagnose bacteremia early in the subject. Furthermore, medical staff can take quick action for high-risk groups of bacteremia.

Hereinafter, a bacteremia prediction model used in various embodiments of the present invention will be described in detail with reference to FIGS. 3A to 3D .

3A and 3B exemplarily show training data of a bacteremia prediction model used in various embodiments of the present invention.

Referring to FIG. 3A , 22332 blood culture episodes obtained from 13402 patients obtained from Gangnam Severance Hospital were collected for learning the bacteremia prediction model used in various embodiments of the present invention, and 1260 of them were collected. Episodes for bacteremia subjects and episodes for 1260 normal subjects were set as training data. Accordingly, the bacteremia prediction model of the present invention was trained to predict bacteremia based on positive and negative result data in a 1:1 ratio.

Referring to FIG. 3B , learning data obtained from specimens of bacteremia-infected individuals and normal individuals are composed of biosignal data, biological test data, and clinical characteristic data. The values of each of these data appear to differ at a statistically significant level for bacteremia-affected individuals and normal individuals.

More specifically, the biosignal data for learning may include SBP, maximum body temperature, minimum body temperature, heart rate, and respiration rate. The learning biological test data may consist of creatine level, albumin level, CRP level, WBC minimum level, WBC maximum level, platelet count, prothrombin time and ALP level. Furthermore, the clinical characteristic data may consist of blood culture period, ICU treatment, central venous catheter, steroid treatment, antibiotic treatment, age and gender.

At this time, among the clinical characteristic data, whether ICU treatment and central venous catheter treatment may be data evaluated at the time of blood culture, antibiotic treatment may be data evaluated 3 weeks before blood culture, and whether steroid treatment is blood It may be data evaluated before 2 weeks of culture. In this case, the biosignal data and the biological test data may be data evaluated before 72 hours of blood culture.

The bacteremia prediction model of the present invention may be trained to determine whether or not the occurrence of bacteremia in a sample object is based on the learning data as described above. FIG. 3C illustrates the configuration of a bacteremia prediction model used in various embodiments of the present invention It is shown hostilely.

Referring to FIG. 3C , the bacteremia prediction model of the present invention may be a prediction model based on an MLP algorithm, which is a multilayer artificial neural network.

More specifically, the bacteremia prediction model of the present invention is an input layer to which biological test data such as ALP level or platelet level, biosignal data such as maximum body temperature, and clinical characteristic data are input, and predicts bacteremia or normal other than bacteremia. It may be a prediction model of a multi-layer structure in which an output layer and one hidden layer exist between these layers.

In this case, the bacteremia prediction model of the present invention may be configured to include one hidden layer composed of 128 or 256 nodes. Furthermore, in the predictive model, a learning rate value that may be a parameter for finding a weight that minimizes a prediction error in predictive learning of bacteremia may be set to 0.1. In addition, a momentum value, which is a parameter value for minimizing prediction error and increasing the learning rate, may be set to 0.95. In addition, the bacteremia prediction model of the present invention can be configured to use 'adadelta' as an optimization function to update parameters in learning, and ' It can be configured to use the tanh' function.

However, the type of the bacteremia prediction model of the present invention is not limited thereto. For example, the prediction model may be at least one of a support vector machine (SVM), a random forest (RF), and an ensemble model consisting of a combination of MLP, SVM, and RF, and DNN, CNN, RNN, DCNN , RBM, DBN, SSD model, or a predictive model based on U-net.

On the other hand, in the learning of the bacteremia prediction model of the present invention, evaluation may be performed in which clinical data having a large influence on predicting normal, not bacteremia, is determined. At this time, the influence evaluation is to be performed based on at least one algorithm of OOS (one-out search), LRP (Layer-wise Relevance Propagation), GI (gradient input), IG (integrated gradients), and SM (saliency map). can

FIG. 3D exemplarily illustrates an evaluation method of a bacteremia prediction model used in various embodiments of the present invention.

At this time, the evaluation of clinical data based on the LRP algorithm will be described as an example, but the evaluation algorithm may be performed based on more various algorithms as described above.

Referring to FIG. 3D , when a bacteremia prediction result is output by a bacteremia prediction model based on various clinical data, an evaluation based on the LRP algorithm may be performed.

More specifically, in this evaluation, the input value of various clinical data of ALP level, biological test data such as platelet count, biosignal data such as maximum body temperature, and further clinical characteristic data by the LRP algorithm and normal values other than bacteremia or bacteremia The relevance of the output value to .

More specifically, the relationship between the input value of various clinical data and the output value of bacteremia or normal non-bacteremia can be calculated by the following [Equation 1].

[Equation 1]

). 나아가, b _j 는 j번째 노드의 편향값을 의미할 수 있다.Here, L may mean an output layer value, and Z _ji may mean the product of the weight of the l +1th layer and the lth layer, and the input value of the lth layer (

). Further, b _j may mean a bias value of the j-th node.

According to the calculated relevance score, among various clinical data, clinical data having a high relevance to predicting bacteremia different from sepsis or clinical data having a large influence in predicting normality other than bacteremia may be determined.

As clinical data having such a large influence can be applied as input data to the learning of the predictive model, the predictive model of the present invention may have superior bacteremia predictive ability than other models in predicting bacteremia.

Example 1: First evaluation of a bacteremia prediction model according to various embodiments of the present invention

Hereinafter, the first evaluation result of the bacteremia prediction model used in various embodiments of the present invention will be described in detail with reference to 4a to 4f. 4A and 4B show a first evaluation result for a bacteremia prediction model used in various embodiments of the present invention.

First, referring to FIG. 4A , evaluation results of a bacteremia prediction model configured to predict bacteremia or normal other than bacteremia based on biological test data, biosignal data, and data are shown.

ve Bayesian이 본 평가에 함께 이용되었다.At this time, the bacteremia prediction model of the present invention is an MLP model composed of a hidden layer having 128 nodes (bacteremia prediction model (A) of the present invention), an MLP model composed of a hidden layer having 256 nodes (bacteremia prediction of the present invention) Model (B)), classification model of random forest (bacteremia prediction model of the present invention (C)), classification model of SVM (bacteremia prediction model of the present invention (D)), and MLP model consisting of a hidden layer with 128 nodes , it is composed of an ensemble model (bacteremia prediction model (E) of the present invention) in which an MLP model consisting of a hidden layer having 256 nodes and a random forest classification model are combined. Furthermore, a conventional predictive model Na, configured to analyze the risk for bacteremia after each variable is stratified and divided into sections

ve Bayesian was also used in this evaluation.

Referring to FIG. 4A , the sensitivity of the bacteremia prediction model (A) of the present invention and the bacteremia prediction model (B) of the present invention based on MLP is 0.81, which is the highest compared to the rest of the models. Furthermore, the specificity of the bacteremia prediction model (C) of the present invention is 0.655, which is the highest compared to the rest of the models.

ve Bayesian에 비하여 높은 것으로 나타난다. 이때, AUC 값은 진단 능력과 연관이 있을 수 있다. 따라서, 본 발명의 다양한 실시예에 따른 균혈증 예측 모델은, 균혈증 진단 시스템에 적용될 경우, 개체에 대하여 균혈증의 발병 위험도를 높은 정확도로 미리 예측할 수 있다.In particular, referring together with FIG. 4b, the AUC value, which can mean the diagnostic ability of the bacteremia prediction models (A) to (E) of the present invention, is an average of about 0.729, and Na, which is a conventional predictive model

ve appears to be higher than that of Bayesian. In this case, the AUC value may be related to the diagnostic ability. Therefore, when the bacteremia prediction model according to various embodiments of the present invention is applied to a bacteremia diagnosis system, it is possible to predict in advance the risk of developing bacteremia for an individual with high accuracy.

According to the above results, it appears that the bacteremia prediction model according to various embodiments of the present invention predicts the risk of developing bacteremia with higher reliability than the conventional prediction model. In particular, the MLP-based bacteremia prediction model appears to have a high AUC value while maintaining high sensitivity than other models. Accordingly, the bacteremia prediction model of the present invention may be an MLP-based prediction model, but is not limited thereto.

4c and 4d show clinical data according to the influence of bacteremia prediction on the bacteremia prediction model used in various embodiments of the present invention.

Referring to FIG. 4C , the evaluation results of influence (relevance) of clinical data (biological test data, biosignal data, and clinical characteristic data) used in various embodiments of the present invention are shown. In this evaluation, the ranking of variables having a high influence on the prediction of bacteremia was evaluated based on 'One-out search' and 'Gini Importance'. More specifically, according to the ranking evaluation result of One-out search that can be applied to the MLP-based prediction model, the ALP level and platelet count in the biological test data, and the highest body temperature in the biosignal data predict the risk of bacteremia. appears to have a high influence. On the other hand, according to the ranking evaluation result of Gini Importance that can be used in the random forest-based prediction model, although it is somewhat different from the ranking evaluation result of one-out search, the ALP level and platelets evaluated as having high influence by the one-out search Number and peak body temperature appear to have a high influence on classifying bacteremia.

Referring further to FIG. 4D , a variable with a high influence on the prediction of bacteremia based on the gradient-based method LRP (Layer-wise Relevance Propagation), GI (gradient input), IG (integrated gradients), and SM (saliency map) The results of evaluating their rankings are shown. More specifically, according to the ranking evaluation results of LRP, GI, and IG, it appears that the ALP level has a high influence in predicting the risk of bacteremia inventing, similar to the ranking evaluation result of the one-out search described above. In addition, platelet count appears to have a high influence on predicting the risk of developing bacteremia. In the ranking evaluation results according to SM, platelet counts and ALP levels appear to have a high influence on predicting the risk of bacteremia.

These results suggest that, in predicting bacteremia based on clinical data for an individual, the ALP level and platelet count among biological test data, and furthermore, the highest body temperature among biosignal data provides a great influence in predicting bacteremia. there is. That is, the bacteremia prediction model of the present invention can be learned to predict whether or not bacteremia will develop based on the middle ALP level and platelet count, and furthermore, the highest body temperature, and can predict the risk of developing bacteremia for an individual with excellent diagnostic ability.

4e and 4f show performance changes according to changes in clinical data according to the influence of bacteremia prediction for the bacteremia prediction model used in various embodiments of the present invention.

Referring to (a) of Figure 4e, after applying Occlusion of LRP, SM, one-out search (OOS), and perturbation-based method for the evaluation of variables for the ANN machine learning algorithm of MLP, the influence is evaluated as high The AUC values when accumulating and excluding variables in the order of influence are shown. Furthermore, for the evaluation of variables for the RF-based prediction model, the AUC values when variables evaluated as having high influence by the Gini Importance method are accumulated and excluded in the order of influence are shown together.

More specifically, as the top variables evaluated as having a high influence in predicting bacteremia (eg, ALP level, platelet count, maximum body temperature, etc.) are excluded, the AUC value for predicting bacteremia appears to decrease. . This may mean a decrease in the performance of predicting bacteremia due to the exclusion of high-impact clinical data.

Referring to (b) of Figure 4e, after applying the occlusion of LRP, SM, one-out search (OOS), and disturbance-based method for the evaluation of variables for the ANN machine learning algorithm of MLP, the influence is evaluated as high. The AUC values when added variables are accumulated in the order of influence are shown. Furthermore, for the evaluation of variables in the RF-based prediction model, the AUC values when variables evaluated as having high influence by the Gini Importance method are accumulated and added in the order of influence are shown together.

More specifically, as the top variables evaluated to have high influence in predicting bacteremia (eg, ALP level, platelet count, maximum body temperature, etc.) are added, the AUC value for predicting bacteremia appears to increase. . In this case, the change in the AUC value after adding the top 10 variables appears to be insignificant. These results may mean that the top 10 clinical data have a high influence in predicting bacteremia. On the other hand, in the case of OOS, there is a significant increase in the AUC value according to the addition of a variable compared to other evaluation models. Accordingly, in the method for predicting the risk of developing bacteremia according to various embodiments of the present invention, the OOS model may be applied to the evaluation of various clinical variables, that is, biological test data, biosignal data, and clinical characteristic data, but is limited thereto. it is not

Referring to (a) of Figure 4f, after applying the Gini Importance method or OOS for the evaluation of variables to the non-ANN machine learning algorithm of RF and SVM, variables evaluated as having high influence are accumulated and excluded in the order of influence. The AUC value at the time is shown.

More specifically, as the top variables evaluated as having a high influence in predicting bacteremia (eg, ALP level, platelet count, maximum body temperature, etc.) are excluded, the AUC value for predicting bacteremia appears to decrease. . This may mean a decrease in the performance of predicting bacteremia due to the exclusion of high-impact clinical data.

Referring to (b) of FIG. 4f, after applying the Gini Importance method or OOS for the evaluation of variables for non-ANN machine learning algorithms of RF and SVM, variables evaluated as having high influence are accumulated and added in the order of influence. The AUC value at the time is shown.

More specifically, as the top variables evaluated to have high influence in predicting bacteremia (eg, ALP level, platelet count, maximum body temperature, etc.) are added, the AUC value for predicting bacteremia appears to increase. . In this case, the change in the AUC value after adding the top 10 variables appears to be insignificant. These results may mean that the top 10 clinical data have a high influence in predicting bacteremia. On the other hand, in the case of OOS, there is a significant increase in the AUC value according to the addition of a variable compared to other evaluation models. Accordingly, in the method for predicting the risk of developing bacteremia according to various embodiments of the present invention, the OOS model may be applied to the evaluation of various clinical variables, that is, biological test data, biosignal data, and clinical characteristic data, but is limited thereto. it is not

On the other hand, the bacteremia prediction model used in various embodiments of the present invention may be configured to use clinical data having a large influence in predicting bacteremia as learning data. Accordingly, the bacteremia prediction model may provide a bacteremia prediction result with improved diagnostic ability than other predictive models in predicting bacteremia. As a result of Example 1 above, it was confirmed that the bacteremia prediction model used in various Examples of the present invention predicts the risk of developing bacteremia with high accuracy. Furthermore, according to the results shown to have a high AUC value, it was confirmed that the predictive model of the present invention has excellent diagnostic ability.

Accordingly, the present invention can provide a good prognosis for treatment by advancing the treatment time for bacteremia. Furthermore, the present invention has the effect of predicting the risk of onset with high reliability and accuracy for bacteremia different from sepsis.

In particular, the present invention can provide an early diagnosis of the onset of bacteremia in a subject by providing a system for predicting the risk of developing bacteremia based on clinical data that can be quickly obtained from the subject.

Accordingly, the present invention can provide the effects of increasing the survival rate of an individual with bacteremia, preventing complications, and reducing the cost of treatment.

In addition, the present invention, when bacteremia is predicted by the predictive model for the subject, as it is configured to provide an alarm to inform the risk of developing bacteremia, caregivers or medical staff are easier for subjects requiring continuous monitoring, such as critically ill patients High-risk groups for developing bacteremia can be recognized.

Example 2: Second evaluation of bacteremia prediction model according to various embodiments of the present invention

Hereinafter, the second evaluation result of the bacteremia prediction model used in various embodiments of the present invention will be described in detail with reference to 5a to 5i. 5A and 5B show learning data and evaluation data for the second evaluation of the bacteremia prediction model used in various embodiments of the present invention. 5c to 5i show a second evaluation result for the bacteremia prediction model used in various embodiments of the present invention.

At this time, referring to FIG. 5A , for bacteremia prediction in this evaluation, age and gender were set as characteristic parameters as clinical characteristic data, and SBP (systolic blood pressure), DBP (diastolic blood pressure), and highest as biosignal data. Body temperature, minimum body temperature, heart rate, and respiration rate were set as characteristic parameters. Furthermore, as biological test data, albumin level, ALT (alanine aminotransferase) level, thromboplastin clotting time, AST (aspartate aminotransferase) level, CRP (C-reactive protein) level, erythrocyte sedimentation rate, ferritin (FERR) Level, HMG (hemoglobin) level, PTINR (Prothrombin Time International Normalized Ratio), PTPER (Prothrombin Time (%)), PTSEC (Prothrombin Time (sec)), Red blood cell count, TBIL (Total bilirubin) level, TCO ₂ (total carbon) dioxide) and white blood cell count were set as characteristic parameters.

In this case, the biosignal data and the biological test data may include a minimum value, a maximum value, or a measured value measured within 3 days.

More specifically, referring to FIG. 5B , the configuration of data sets A, B, and C used for the second evaluation of the bacteremia prediction model used in various embodiments of the present invention is shown.

A data set is configured such that bacteremia positive and bacteremia negative data have a ratio of 1:1 for training and validation, and bacteremia positive and bacteremia negative data have a ratio of about 1:10 for testing. At this time, data set A is bacteremia-positive and bacteremia-negative data sampled between 2007 and 2018 at Gangnam Severance Hospital and Sinchon Severance Hospital, with data from 2007 to 2015 for training and validation, and 2016 for testing Data for years to 2018 are available.

The B data set is configured such that bacteremia positive and bacteremia negative data have a ratio of 1:1 for training and validation, and bacteremia positive and bacteremia negative data have a ratio of about 1:7 for testing. In this case, the C data set may be randomly divided data after merging bacteremia positive and bacteremia negative data sampled between 2007 and 2018 at Gangnam Severance Hospital and Sinchon Severance Hospital by year.

The C data set is configured such that bacteremia positive and bacteremia negative data have a ratio of 1:1 for training and validation, and bacteremia positive and bacteremia negative data have a ratio of about 1:7 for testing. At this time, the C data set is randomly divided after merging the bacteremia-positive and bacteremia-negative data sampled between 2007 and 2018 at Gangnam Severance Hospital and Sinchon Severance Hospital by year, and the same patient sample is not duplicated for each set. It may be data from which a negative sample is removed if it is separated and positive/negative samples of the same patient exist at the same time.

Referring to FIG. 5C , an evaluation result for the bacteremia prediction model learned based on the aforementioned data A is shown. At this time, for the bacteremia prediction model consisting of a plurality of layers of 4 layers consisting of 128 nodes, 64 nodes, 25 nodes, and 128 nodes, evaluation of the bacteremia prediction diagnosis performance according to whether prior learning was performed was performed. On the other hand, in the pre-trained bacteremia prediction model, weight initializing was applied to an AutoEncoder that learns a label value by itself.

More specifically, in the case of the pre-trained bacteremia prediction model (pret) of the present invention, the accuracy of predicting positive or negative bacteremia is 0.759 and the sensitivity is 0.762, and the accuracy (0.729) and sensitivity (0.75) of the models that are not are higher than the accuracy (0.729) and sensitivity (0.75). appear. In particular, the AUC value, which can mean the diagnostic ability of the pre-trained bacteremia prediction model (pret) of the present invention, is 0.840, which is higher than the AUC value of the model that is not, 0.818.

According to the above results, the bacteremia prediction model according to various embodiments of the present invention may be a model to which weight initialization is applied to the auto encoder, which has a high AUC value while maintaining high sensitivity. However, the present invention is not limited thereto.

Referring to FIGS. 5D and 5E , evaluation results according to the layer size of the bacteremia prediction model learned based on the above-described B data and C data are shown.

First, referring to FIG. 5D , a prediction model consisting of 64 nodes, 128 nodes, and a plurality of layers (64-128-64) of 3 layers each consisting of 64 nodes, 64 nodes, 128 nodes, 128 nodes, and A performance evaluation result using the above-described B data is shown for a prediction model composed of a plurality of layers 64-128-128-64 of 4 layers each composed of 64 nodes.

In this case, do (drop-out) or bn (batch normalization) with a probability of 0.5 is applied to each of the 64-128-64 prediction model and the 64-128-128-64 prediction model.

More specifically, the bacteremia prediction accuracy for each of the 64-128-64 prediction model and the 64-128-128-64 prediction model to which bn is applied is 0.735 and 0.738, indicating that they have similar values even when the layer size changes. In addition, the AUC values for bacteremia prediction for the 64-128-64 prediction model and the 64-128-128-64 prediction model to which bn is applied are 0.847 and 0.859, indicating that they have similar values even when the layer size changes.

Meanwhile, further referring to FIG. 5E , a predictive model consisting of a single layer 128 consisting of 128 nodes, a plurality of layers 64-128-64 of three layers each consisting of 64 nodes, 128 nodes, and 64 nodes. Performance using the C data described above for a prediction model composed of 64 nodes, 128 nodes, 128 nodes, and a plurality of layers (64-128-128-64) of 4 layers each composed of 64 nodes. The evaluation results are shown.

At this time, a do of 0.5 probability was applied to the 128 prediction model, and do and/or bn of 0.5 probability was applied to each of the 64-128-64 prediction model and the 64-128-128-64 prediction model.

More specifically, the AUC value of the single-layer 128 prediction model is 0.751, which is higher than the AUC values of the 64-128-64 prediction model and the 64-128-128-64 prediction model having a plurality of layers.

At this time, in the case of the 64-128-64 predictive model and the 64-128-128-64 predictive model, the accuracy, sensitivity, and AUC value of bacteremia prediction when bn is applied similarly to the result of FIG. 5d described above are high. .

According to the above results, it appears that the bacteremia prediction model according to various embodiments of the present invention has high sensitivity, accuracy, and AUC value even with a change in layer size. Accordingly, the bacteremia prediction model according to various embodiments of the present invention has a single layer composed of 128 nodes, has a plurality of three layers of 64-128-64, or has a 64-128-128-64 It may be a model to which bn has been applied and has a plurality of layers of 4 layers. However, the structure of the bacteremia prediction model of the present invention is not limited thereto.

Referring to FIGS. 5F and 5G , evaluation results for a stacking ensemble model learned based on the above-described B data and C data are shown.

First, referring to FIG. 5F , an ensemble having a plurality of layers 64-128-64 of 3 layers consisting of 64 nodes, 128 nodes and 64 nodes, respectively, and having 10, 15 and 20 stacking models. The performance evaluation results using the above-described B data for the model are shown.

More specifically, when data set B was applied, the AUC value for the ensemble model with 15 stacking models was 0.887, the AUC value for the ensemble model with 10 stacking models (0.865), and the ensemble with 20 stacking models. It appears to be higher than the AUC value of the model (0.873).

Meanwhile, referring to FIG. 5G , performance evaluation results using the aforementioned C data are shown for an ensemble model having a single layer composed of 128 nodes and having 10, 15, and 20 stacking models.

More specifically, when the C data set is applied, the AUC value for the ensemble model with 15 and 20 stacking models is 0.752, which is higher than the AUC value for the ensemble model with 10 stacking models (0.750).

According to the above results, the bacteremia prediction model according to various embodiments of the present invention may be an ensemble model having 15 to 20 stacking models. However, the structure of the bacteremia prediction model of the present invention is not limited thereto.

5H and 5I , evaluation results of the MLP-based predictive model and the Xgboost-based predictive model learned based on the above-described B data and C data are shown.

First, referring to FIG. 5H , it has a plurality of layers 64-128-64 of 3 layers consisting of 64 nodes, 128 nodes and 64 nodes, respectively, and is composed of an MLP model with 15 stacking models and a gbtree layer. Performance evaluation results using the above-described B data for the Xgboost model are shown.

More specifically, when the B data set was applied, the AUC value of the predictive model based on the Xgboost model was 0.842 and the sensitivity was 0.745, which was similar to the AUC value (0.887) and sensitivity (0.801) of the MLP-based predictive model. appears to be

Meanwhile, referring together with FIG. 5I, the performance evaluation results using the above-described C data are shown for the Xgboost model, which has a single layer composed of 128 nodes, and is composed of a 20-person MLP model of a stacking model and a gbtree layer.

More specifically, when the C data set was applied, the AUC value of the predictive model based on the Xgboost model was 0.745 and the sensitivity was 0.635, and the AUC value (0.752) and sensitivity (0.664) of the MLP-based predictive model had similar levels. appears to be

According to the above results, the bacteremia prediction model according to various embodiments of the present invention may be an MLP-based prediction model or an Xgboost-based prediction model. However, the bacteremia prediction model of the present invention is not limited thereto, and may be a model configured to predict bacteremia positive or negative based on more various predictive models.

As a result of Example 2 above, it was confirmed that the bacteremia prediction model used in various Examples of the present invention predicted the risk of developing bacteremia with high accuracy. Furthermore, according to the results shown to have a high AUC value, it was confirmed that the predictive model of the present invention has excellent diagnostic ability.

Accordingly, the present invention can provide a good prognosis for treatment by advancing the treatment time for bacteremia. Furthermore, the present invention has the effect of predicting the risk of onset with high reliability and accuracy for bacteremia different from sepsis.

In particular, the present invention can provide an early diagnosis of the onset of bacteremia in a subject by providing a system for predicting the risk of developing bacteremia based on clinical data that can be quickly obtained from the subject.

Accordingly, the present invention can provide the effects of increasing the survival rate of an individual with bacteremia, preventing complications, and reducing the cost of treatment.

In addition, according to the present invention, when bacteremia is predicted by a predictive model for an individual, it is configured to provide an alarm to inform the risk of developing bacteremia. High-risk groups for developing bacteremia can be recognized.

Each feature of the various embodiments of the present invention may be partially or wholly combined or combined with each other, and as those skilled in the art will fully understand, technically various interlocking and driving are possible, and each embodiment may be implemented independently of each other. and may be implemented together in a related relationship.

Although the embodiments of the present invention have been described in more detail with reference to the accompanying drawings, the present invention is not necessarily limited to these embodiments, and various modifications may be made within the scope without departing from the technical spirit of the present invention. . Therefore, the embodiments disclosed in the present invention are not intended to limit the technical spirit of the present invention, but to explain, and the scope of the technical spirit of the present invention is not limited by these embodiments. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive. The protection scope of the present invention should be construed by the following claims, and all technical ideas within the scope equivalent thereto should be construed as being included in the scope of the present invention.

100: Device for predicting the risk of developing bacteremia
110: receiver
120: input unit
130: output unit
140: storage
150: processor
200: object
300: clinical data
310: biological test data
320: biosignal data
330: clinical characteristic data
400: medical treatment device
500: medical staff device
1000: risk prediction system

Claims (26)

In the method for predicting the risk of developing bacteremia implemented by the processor and the receiver,
receiving, through the receiver, biological test data for the subject;
Using the processor, using a bacteremia prediction model configured to predict the risk of developing bacteremia for the subject based on the biological test data, predicting the risk of developing bacteremia of the subject,
The biological test data,
At least one of Creatinine Level, Albumin Level, C-reactive Protein (CRP) Level, White Blood Cell (WBC) Minimum, WBC Maximum, Prothrombin Time, Platelet Count and ALP (Alkaline phosphatase) including levels,
Further comprising the step of receiving the biosignal data for the object through the receiving unit,
The bacteremia prediction model is,
Further configured to predict the risk of developing bacteremia based on the biological test data and the biosignal data,
The biosignal data is
A method of predicting a risk of developing bacteremia, comprising at least one of systolic blood pressure (SBP), diastolic blood pressure (DBP), minimum body temperature, heart rate and respiration rate, and maximum body temperature. delete According to claim 1,
Further comprising the step of receiving, through the receiving unit, clinical characteristic data for the subject,
The bacteremia prediction model is,
Further configured to predict the risk of developing bacteremia based on the biological test data and the clinical characteristic data,
The clinical characteristic data is
A method of predicting the risk of developing bacteremia, which is at least one of gender, age, blood culture period, intensive care unit (ICU) treatment, central venous catheter insertion, steroid treatment, and antibiotic treatment. According to claim 1,
The biological test data,
At least one of the creatine level, the albumin level, the CRP level, the WBC minimum level, the WBC maximum level, the prothrombin time, and the platelet count and the ALP level, the method of predicting the risk of developing bacteremia. According to claim 1,
The biological test data,
ALT (alanine aminotransferase) level, thromboplastin clotting time, AST (aspartate aminotransferase) level, CRP (C-reactive protein) level, erythrocyte sedimentation rate, FERR (ferritin) level, HMG (hemoglobin) level, PTINR At least one of (Prothrombin Time International Normalized Ratio), PTPER (Prothrombin time (%)), PTSEC (Prothrombin time (sec)), red blood cell count, TBIL (Total bilirubin) level, TCO ₂ (total carbon dioxide) level, and white blood cell count Further comprising, a method of predicting the risk of developing bacteremia. According to claim 1,
The bacteremia is, acute severe bacteraemia, a method of predicting the risk of developing bacteremia. According to claim 1,
The bacteremia prediction model is,
MLP (multi-layer perceptron), SVM (support vector machine), RF (random forest), and at least one model of an ensemble model consisting of a combination of a plurality of models, a pre-trained model, and an Xgboost model , a method of predicting the risk of developing bacteremia. 8. The method of claim 7,
The bacteremia prediction model is,
The MLP model or the Xgboost model,
The MLP model includes a single hidden layer having 128 nodes or 256 nodes, or
A method for predicting the risk of developing bacteremia, including a plurality of hidden layers in which the single hidden layer is present in plurality. According to claim 1,
The bacteremia prediction model is,
Receiving, through the receiving unit, learning data consisting of biological experimental data, biosignal data, and clinical characteristic data for a bacteremia-infected sample object and a normal sample object, and
Through the processor, based on the learning data, the model learned through the step of predicting bacteremia or normal, a method of predicting the risk of bacteremia. 10. The method of claim 9,
Performed after the step of receiving the learning data,
Further comprising the step of evaluating the learning data,
The step of evaluating the learning data is,
Calculating, through the processor, a relevance score with respect to the learning data with respect to bacteremia; and
Based on the relevance score, the method of predicting the risk of developing bacteremia, comprising the step of determining data for bacteremia-related learning within a predetermined rank. 11. The method of claim 10,
The step of evaluating the learning data is,
Through the processor, based on at least one algorithm of one-out search (OOS), Layer-wise Relevance Propagation (LRP), gradient input (GI), integrated gradients (IG), and saliency map (SM),
calculating a relevance score with respect to the learning data for bacteremia, and
Based on the relevance score, the method of predicting the risk of developing bacteremia, comprising the step of determining data for bacteremia-related learning within a predetermined rank. According to claim 1,
Through the processor, the method of predicting the risk of developing bacteremia further comprising the step of providing the predicted risk of developing bacteremia for the individual. 13. The method of claim 12,
The step of providing the risk of developing bacteremia comprises:
When the risk of developing bacteremia for an individual is predicted by the bacteremia prediction model,
Through the processor, the method of predicting the risk of developing bacteremia, comprising the step of providing a risk notification. 13. The method of claim 12,
The step of providing the risk of developing bacteremia comprises:
Calculating a risk probability of developing bacteremia for an individual by the bacteremia prediction model through the processor, and
A method of predicting the risk of developing bacteremia, comprising the step of providing the probability of the risk of developing bacteremia. a receiver configured to receive biological test data for the subject; and
and a processor connected to the receiver;
The processor is configured to predict the risk of developing bacteremia in the subject by using a bacteremia prediction model configured to predict the risk of developing bacteremia in the subject based on the biological test data,
The biological test data,
At least one of Creatinine Level, Albumin Level, C-reactive Protein (CRP) Level, White Blood Cell (WBC) Minimum, WBC Maximum, Prothrombin Time, Platelet Count and ALP (Alkaline phosphatase) including levels,
The receiving unit is further configured to receive biosignal data for an object,
The bacteremia prediction model is,
Further configured to predict the risk of developing bacteremia based on the biological test data and the biosignal data,
The biosignal data is
A device for predicting the risk of developing bacteremia, including at least one of systolic blood pressure (SBP), diastolic blood pressure (DBP), minimum body temperature, heart rate and respiration rate, and maximum body temperature. delete 16. The method of claim 15,
The receiving unit is further configured to receive clinical characteristic data for the subject,
The bacteremia prediction model is,
Further configured to predict the risk of developing bacteremia based on the biological test data and the clinical characteristic data,
The clinical characteristic data is
A device for predicting the risk of developing bacteremia, which is at least one of gender, age, blood culture period, intensive care unit (ICU) treatment, central venous catheter insertion, steroid treatment, and antibiotic treatment. 16. The method of claim 15,
The biological test data,
At least one of the creatine level, the albumin level, the CRP level, the WBC minimum level, the WBC maximum level, the prothrombin time, and the platelet count, the ALP level, the device for predicting the risk of developing bacteremia. 16. The method of claim 15,
The biological test data,
ALT (alanine aminotransferase) level, thromboplastin clotting time, AST (aspartate aminotransferase) level, CRP (C-reactive protein) level, erythrocyte sedimentation rate, FERR (ferritin) level, HMG (hemoglobin) level, PTINR At least one of (Prothrombin Time International Normalized Ratio), PTPER (Prothrombin time (%)), PTSEC (Prothrombin time (sec)), red blood cell count, TBIL (Total bilirubin) level, TCO ₂ (total carbon dioxide) level, and white blood cell count Further comprising, a device for predicting the risk of developing bacteremia. 16. The method of claim 15,
The bacteremia is, acute severe bacteraemia, a device for predicting the risk of developing bacteremia. 16. The method of claim 15,
The bacteremia prediction model is,
MLP (multi-layer perceptron), SVM (support vector machine), RF (random forest), and at least one model of an ensemble model consisting of a combination of a plurality of models, a pre-trained model, and an Xgboost model , A device for predicting the risk of developing bacteremia. 22. The method of claim 21,
The bacteremia prediction model is,
The MLP model or the Xgboost model,
The MLP model includes a single hidden layer having 128 nodes or 256 nodes, or
A device for predicting the risk of developing bacteremia, comprising a plurality of hidden layers in which the single hidden layer is present in plurality. 16. The method of claim 15,
The bacteremia prediction model is,
A model trained through the steps of receiving training data consisting of biological experimental data, biosignal data, and clinical characteristic data for a bacteremia-infected sample object and a normal sample object, and predicting bacteremia or normality based on the training data A device for predicting the risk of developing phosphorus and bacteremia. 16. The method of claim 15,
The device for predicting the risk of developing bacteremia, further comprising an output unit configured to provide the predicted risk of developing bacteremia for the individual. 25. The method of claim 24,
the output unit,
When the risk of developing bacteremia for an individual is predicted by the bacteremia prediction model,
A device for predicting the risk of developing bacteremia, configured to provide a risk alert. 25. The method of claim 24,
The bacteremia prediction model is,
further configured to calculate a risk probability of developing bacteremia for the subject,
the output unit,
A device for predicting the risk of developing bacteremia, further configured to provide the probability of developing the bacteremia.

Images