KR102601514B1 - System for prediction of early dropping out in outpatients with alcohol use disorders and method thereof - Google Patents

Abstract

Embodiments of the present invention relate to a system and method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder. According to embodiments of the present invention, data of a plurality of alcohol use disorder patients are collected, a plurality of independent variables to which one or more machine learning algorithms are to be applied are determined, and the data of the plurality of alcohol use disorder patients are processed to produce processed data. A preprocessor that generates; Receives processed data, sets whether multiple alcohol use disorder patients stop early outpatient treatment as a dependent variable, and generates a prediction model by applying one or more machine learning algorithms to all or part of the processed data based on the independent variables. Prediction model generation unit; Outpatient treatment for patients with alcohol use disorders, including a prediction unit that inputs all or part of the processed data into a prediction model to generate prediction results regarding early discontinuation of outpatient treatment for patients with multiple alcohol use disorders, and an output unit that outputs the prediction results. It can provide an early outage prediction system.

Description

System and method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder {SYSTEM FOR PREDICTION OF EARLY DROPPING OUT IN OUTPATIENTS WITH ALCOHOL USE DISORDERS AND METHOD THEREOF}

The present invention provides a system and method for predicting early cessation of outpatient treatment for patients with alcohol use disorder.

Previously, studies were conducted on factors affecting the maintenance of continuous outpatient treatment in order to increase the treatment rate of patients with alcohol use disorder. However, most of these studies were only prospective studies, and even in the case of retrospective studies, they focused on traditional methodologies such as regression analysis.

Machine learning is the study of computer algorithms that automatically improve through experience and the use of data. Machine learning is also considered a part of artificial intelligence. Machine learning algorithms do not explicitly program specific actions; instead, machine learning algorithms can be used to build models to make predictions or decisions based on samples, called training data. Machine learning can be used in a variety of applications, including medicine, speech recognition, and computer vision. For example, Publication Patent No. 10-2021-0047149 discloses a method and system for predicting the risk of myocardial infarction recurrence in cardiac rehabilitation patients using machine learning techniques. Specifically, clinical records and lifelog data of cardiac rehabilitation patients are acquired, a dataset to be input into a machine learning-based risk prediction model is created based on the data, and the data is generated using the machine learning-based risk prediction model. From the three, it includes the step of determining and providing a myocardial infarction recurrence risk index or myocardial infarction recurrence risk index for cardiac rehabilitation patients.

Prediction models based on machine learning can classify classes with high accuracy. Recently, in psychiatric research, prediction models based on machine learning have been usefully used in the process of developing decision support systems.

Alcohol use disorder can cause not only physical diseases such as physical complications and alcohol-related dementia caused by alcohol, but also social problems such as alcohol-related crimes and accidents, and enormous economic losses.

According to the 2016 Epidemiological Survey on Mental Illness in Korea, the lifetime prevalence rate of alcohol use disorder, which includes alcohol dependence and abuse, is 12.2% (18.1% for men, 6.4% for women), which is the highest prevalence rate compared to other mental disorders.

Alcohol use disorder is a disease with a higher relapse rate than other mental disorders. To prevent recurrence, it is necessary to manage it over a long period of time, rather than ending it with a one-time treatment. Additionally, consistent treatment can have a positive effect on treatment results. Therefore, continuous follow-up of patients becomes an important indicator for evaluating the prognosis of alcohol use disorder.

The reality is that the outpatient treatment retention rate for patients with alcohol use disorder is quite low. In foreign countries, the rate of loss of follow-up for patients receiving outpatient treatment for alcohol use disorder at the fourth treatment ranges from 52% to 75%. According to a domestic study, 91.7% of patients lost follow-up within 6 months after discharge. Among patients with alcohol use disorder, it is important to predict and manage patients whose follow-up is discontinued early.

Embodiments of the present invention can calculate the probability of early discontinuation of outpatient treatment for patients with alcohol use disorder by designing a prediction model using machine learning to predict whether outpatient treatment will be discontinued early.

Embodiments of the present invention help manage patients so that they can consistently maintain treatment for patients with alcohol use disorder who are at high risk of discontinuing outpatient treatment at an early stage, and ultimately contribute to preventing patient recurrence and increasing the success rate of treatment. .

However, the technical challenge that this embodiment aims to achieve is not limited to the technical challenges described above, and other technical challenges may exist.

In one aspect, embodiments of the present invention collect data from multiple alcohol use disorder patients and apply one or more machine learning algorithms to generate a prediction model for early discontinuation of outpatient treatment for multiple alcohol use disorder patients. A pre-processing unit that determines a plurality of independent variables and processes data of a plurality of alcohol use disorder patients to generate processed data; Receives processed data, sets whether multiple alcohol use disorder patients stop early outpatient treatment as a dependent variable, and generates a prediction model by applying one or more machine learning algorithms to all or part of the processed data based on the independent variables. Prediction model generation unit; a prediction unit that inputs all or part of the processed data into a prediction model to generate prediction results regarding whether to discontinue early outpatient treatment for patients with multiple alcohol use disorders; and an output unit for outputting the prediction results. A system for predicting early cessation of outpatient treatment for patients with alcohol use disorder is provided.

In another aspect, embodiments of the present invention include a data collection step of collecting data on a plurality of alcohol use disorder patients; An independent variable determination step of determining a plurality of independent variables to which one or more machine learning algorithms will be applied to create a prediction model for early discontinuation of outpatient treatment for multiple alcohol use disorder patients; A pre-processing step of processing data of a plurality of alcohol use disorder patients to generate processed data; Receives processed data, sets whether multiple alcohol use disorder patients stop early outpatient treatment as a dependent variable, and generates a prediction model by applying one or more machine learning algorithms to all or part of the processed data based on the independent variables. Predictive model creation step; A prediction step for inputting all or part of the processed data into a prediction model to generate a prediction result regarding whether to initially discontinue outpatient treatment for multiple alcohol use disorder patients, and an output step for outputting the prediction result. Provides a method for predicting early discontinuation of treatment.

According to embodiments of the present invention, it is possible to predict whether early discontinuation of outpatient treatment will occur by calculating the probability of early discontinuation of outpatient treatment for patients with alcohol use disorder through designing a prediction model using machine learning.

According to embodiments of the present invention, it is possible to help manage patients so that they can consistently maintain treatment for patients with alcohol use disorder who are at high risk of early outpatient treatment discontinuation, and ultimately prevent patients from relapse and increase the success rate of treatment. can contribute to raising

Figure 1 is a schematic diagram of a system for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder according to embodiments of the present invention.
Figure 2 is a flowchart showing a variable determination operation performed by the preprocessor according to embodiments of the present invention.
Figure 3 is a diagram showing the operation of the pre-processing unit classifying data of an alcohol use disorder patient into a learning data group and a test data group according to embodiments of the present invention.
Figure 4 is a diagram showing an operation of a preprocessor performing sampling on a specific class according to embodiments of the present invention.
Figure 5 is a diagram illustrating an example of an operation in which a prediction model generator according to embodiments of the present invention generates a prediction model by applying one or more machine learning algorithms to a group of learning data.
Figure 6 is a diagram showing the operation of the prediction model generator according to embodiments of the present invention to determine a prediction model according to the performance evaluation index.
Figure 7 is a diagram showing AUC, one of the performance evaluation indicators according to embodiments of the present invention.
Figure 8 is a diagram showing a method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder according to embodiments of the present invention.

Below, with reference to the attached drawings, embodiments of the present invention will be described in detail so that those skilled in the art can easily implement it. However, the present invention may be implemented in many different forms and is not limited to the embodiments described herein. In order to clearly explain the present invention in the drawings, parts that are not related to the description are omitted, and similar parts are given reference numerals throughout the specification.

Throughout the specification, when a part is said to be "connected" to another part, this includes not only the case where it is "directly connected," but also the case where it is "electrically connected" with another element in between. . In addition, when a part is said to "include" a certain component, this does not mean excluding other components unless specifically stated to the contrary, but may further include other components, and one or more other features. It should be understood that this does not exclude in advance the presence or addition of numbers, steps, operations, components, parts, or combinations thereof.

The terms "about", "substantially", etc. used throughout the specification are used to mean at or close to that value when manufacturing and material tolerances inherent in the stated meaning are presented, and are used to enhance the understanding of the present invention. Precise or absolute figures are used to assist in preventing unscrupulous infringers from taking unfair advantage of stated disclosures. The term “step of” or “step of” as used throughout the specification of the present invention does not mean “step for.”

In this specification, 'part' includes a unit realized by hardware, a unit realized by software, and a unit realized using both. Additionally, one unit may be realized using two or more pieces of hardware, and two or more units may be realized using one piece of hardware.

Figure 1 is a schematic diagram of a system for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder according to embodiments of the present invention.

Referring to FIG. 1, the system 100 for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder according to embodiments of the present invention includes a preprocessor 110, a prediction model generation unit 120, a prediction unit 130, and It may include an output unit 140.

The preprocessor 110 may collect data from multiple alcohol use disorder patients. In addition, the preprocessor 110 may determine a plurality of independent variables to which one or more machine learning algorithms will be applied to generate a prediction model for whether or not outpatient treatment for multiple alcohol use disorder patients will stop early from outpatient treatment. In addition, the pre-processing unit 110 may process data of a plurality of alcohol use disorder patients to apply one or more machine learning algorithms described above to generate processed data.

When collecting data on patients with alcohol use disorder, the preprocessor 110 may use wired or wireless communication with a server or terminal that stores the data. As an example, the preprocessor 110 may receive medical data of patients with alcohol use disorder from one or more medical institutions. At this time, the data of multiple alcohol use disorder patients may be standardized in a common data model (CDM, Common Data Model).

As an example, data on multiple alcohol use disorder patients may be collected from a clinical data warehouse (CDW). The clinical data warehouse (CDW) can deliver data extracted according to study characteristics to the preprocessor 110 through de-identification.

Meanwhile, to increase the sensitivity of the prediction model, patients with multiple alcohol use disorders can be selected from among patients with a hospitalization period of more than 2 weeks.

At this time, among patients with multiple alcohol use disorders, the date of hospitalization for patients who have been hospitalized more than twice for more than 2 weeks can be defined based on the date of the first hospitalization.

Continuous outpatient visits for patients with alcohol use disorder can be defined as whether the patient visits outpatients at least once a month for 6 months after discharge.

Additionally, patients who die within 6 months after discharge may be excluded.

The preprocessor 110 may determine a plurality of independent variables to which one or more machine learning algorithms will be applied to generate a prediction model for whether or not outpatient treatment for patients with alcohol use disorder will be discontinued early.

As an example, the preprocessing unit 110 includes 1) patient's age, 2) gender, 3) hospitalization period, 4) address, 5) medical department, 6) diabetes, liver disease, depressive disorder, and anxiety diagnosed within 1 year before hospitalization. The independent variable can be determined among variables including whether there are comorbidities such as disabilities, 7) whether alcohol use disorder is treated outpatiently before hospitalization, and 8) whether naltrexone is prescribed.

Meanwhile, among patients with multiple alcohol use disorders, t-test, chi-square test, etc. were performed on the determined independent variables to confirm the difference between the outpatient treatment maintenance group and the outpatient treatment early discontinuation group. Statistical analysis can be performed.

The t-test is a statistical method that verifies whether the difference in means between two groups is significant. The chi-square test is a statistical method based on the chi-square distribution and is a test method used to test whether the observed frequency is significantly different from the expected frequency.

As an example, the results of the statistical analysis described above may be as shown in Table 1.

Retention in outpatient treatment (n=126) Early discontinuation of outpatient treatment (n=713) P-value Hospitalization period 0.406 28 days or less 437 (61.3%) 77 (61.1%) 29-56 days 136 (19.1%) 25 (19.8%) 57-70 days 110 (15.4%) 15 (11.9%) 70 days or more 30 (4.2%) 9 (7.1%) gender 0.008*** male 91 (72.2%) 590 (82.7%) female 35 (27.8%) 123 (17.3%) age 0.058* Under 29 9 (7.1%) 22 (3.1%) 30-39 22 (17.5%) 96 (13.5%) 40-49 29 (23.0%) 201 (28.2%) 50-59 30 (23.8%) 216 (30.3%) over 60 years old 36 (28.6%) 178 (25.0%) address 0.04** seoul 37 (29.4%) 144 (20.2%) game 75 (59.5%) 451 (63.3%) etc 14 (11.1%) 118 (16.5%) Department 0.015** psychiatry 111 (88.1%) 546 (76.6%) Gastroenterology 9 (7.1%) 104 (14.6%) etc 6 (4.8%) 63 (8.8%) Outpatient treatment for alcohol use disorder before hospitalization 0.000*** radish 35 (27.8%) 325 (45.6%) you 91 (72.2%) 388 (54.4%) diabetes 0.087* radish 109 (86.5%) 654 (91.7%) you 17 (13.5%) 59 (8.3%) liver disease 0.224 radish 107 (84.9%) 569 (79.8%) you 19 (15.1%) 144 (20.2%) depressive disorder 0.006*** radish 78 (61.9%) 529 (74.2%) you 48 (38.1%) 184 (25.8%) anxiety disorder 0.053* radish 104 (82.5%) 635 (89.1%) you 22 (17.5%) 78 (10.9%) Whether naltrexone is prescribed 0.000*** radish 93 (73.8%) 626 (87.8%) you 33 (26.2%) 87 (12.2%)

According to Table 1, at the significance level of 0.05 for gender, address, department, depressive disorder, outpatient treatment for alcohol use disorder before hospitalization, and whether or not naltrexone was prescribed, among patients with multiple alcohol use disorders, the outpatient treatment maintenance group and the outpatient treatment early discontinuation group A statistically significant difference can be confirmed between the two.

For example, patients who are male, do not have a depressive disorder, are not prescribed naltrexone, are outside of Seoul, are hospitalized in a department other than psychiatry, or are diagnosed with alcohol use disorder before hospitalization and do not receive outpatient treatment. The rate of early discontinuation of treatment may be high.

Through this verification process, differences in characteristics of alcohol use disorder patients between groups can be confirmed.

The preprocessor 110 may process data of a plurality of alcohol use disorder patients to generate processed data in order to apply one or more machine learning algorithms. Meanwhile, the preprocessor 110 may perform the above-described process on prediction target data even after the prediction model is created. This can improve the learning performance of the prediction model.

The preprocessor 110 can secure high-quality processed data and improve the accuracy of the prediction model by removing or modifying missing data, abnormal data, and duplicate data from the data of a plurality of alcohol use disorder patients.

In addition, the preprocessor 110 performs processes such as combining, dividing, filtering, sampling, generating derived variables, creating dummy variables, adjusting scale, changing data types, and normalizing data on the data of multiple alcohol use disorder patients to produce processed data. can be created.

As an example, the preprocessor 110 may modify and align digital information of numbers or letters derived empirically or experimentally and convert them into a simplified format.

The prediction model generator 120 receives processed data, sets whether or not early outpatient treatment of multiple alcohol use disorder patients discontinues outpatient treatment as a dependent variable, and uses one or more machine learning algorithms to apply all or part of the processed data based on the independent variables. You can create a prediction model by applying .

At this time, machine learning algorithms can be broadly classified into three types: supervised learning algorithms, unsupervised learning algorithms, and reinforcement learning algorithms.

A supervised learning algorithm is an algorithm used when there is an intended result. During learning, the machine learning algorithm model can adjust variables for input values and map them to the output.

An unsupervised learning algorithm is an algorithm used when there is no intended result, and can classify input data sets into similar types of sets. Unsupervised learning algorithms can be used in data mining.

The reinforcement learning algorithm is an algorithm used when making decisions about input values. It is an algorithm in which the decision about a given input value gradually changes depending on success/failure when making a decision. As the reinforcement learning algorithm learns, it may be possible to predict the results of the input.

Meanwhile, the prediction model generator 120 may be implemented as a workstation server or a cloud server, for example.

The prediction unit 130 inputs all or part of the processed data generated by the pre-processing unit 110 into the prediction model generated by the prediction model generation unit 120 to determine whether to stop outpatient treatment early for patients with multiple alcohol use disorders. Prediction results can be generated.

The early discontinuation of outpatient treatment for patients with alcohol use disorder prediction system 100 can predict whether early discontinuation of outpatient treatment for patients with alcohol use disorder using the prediction results generated by the prediction unit 130, and can also determine whether early discontinuation of outpatient treatment for patients with alcohol use disorder. You can identify influencing variables.

In addition, the early cessation of outpatient treatment for patients with alcohol use disorder prediction system 100 uses the prediction results generated by the prediction unit 130 to provide special management according to the characteristics of the patient, helping the patient to receive treatment consistently. can do.

The output unit 140 may output the prediction result generated by the prediction unit 130. At this time, the output unit 140 may output the prediction result using a method such as screen output through a display or printing the prediction result using a printer.

Figure 2 is a flowchart showing a variable determination operation performed by the preprocessor 110 according to embodiments of the present invention.

Referring to FIG. 2, when determining independent variables, the preprocessor 110 calculates variance inflation factors (VIF) between the independent variables to solve the multicollinearity problem between the independent variables. And, the independent variables can be determined so that the variance inflation factor (VIF) remains below a preset threshold.

A multicollinearity problem refers to a problem in which part of an independent variable can be expressed as a combination of other independent variables. Multicollinearity problems can occur when independent variables are not independent from each other and have strong interrelationships. As a way to solve the multicollinearity problem, a method of eliminating variables that depend on other independent variables can be used, and in this case, the variance inflation factor (VIF) can be used.

Variance inflation factor (VIF) represents the performance of linear regression of one independent variable on another independent variable. The variance inflation factor (VIF) of the ith variable can be obtained through Equation 1 below.

is the coefficient of determination obtained by linear regression of the ith variable.

Since the value is less than 1, if it depends on the independent variable, The value becomes larger.

The preprocessor 110 may calculate variance inflation factor (VIF) values for all determined independent variables and determine independent variables so that all variance inflation factor (VIF) values are maintained below a preset threshold.

For example, the threshold value of the preset variance inflation factor (VIF) may be determined to be 5.

Referring to FIG. 2, the preprocessor 110 may determine an independent variable (S210).

The preprocessor 110 may calculate the variance inflation factor (VIF) for all determined independent variables using the method described above (S220).

If the variance inflation factor (VIF) for one independent variable exceeds the threshold (S230-Y), the preprocessor 110 may determine another independent variable (S240).

At this time, after determining another independent variable, the preprocessor 110 may enter step S220 and calculate the variance inflation factor (VIF) again for the determined independent variable.

If the variance inflation factor (VIF) for all independent variables is 5 or less (S230-N), the preprocessor 110 may end the variable determination process.

Figure 3 is a diagram showing the operation of the preprocessor 110 classifying alcohol use disorder patient data into a training data set and a test data set according to embodiments of the present invention.

Referring to FIG. 3, the preprocessor 110 may classify data of a plurality of alcohol use disorder patients into a learning data group and a test data group.

In order to create a prediction model using a machine learning algorithm, the preprocessor 110 classifies some of the data of a plurality of alcohol use disorder patients into a training data set and can be used to learn the prediction model. .

Meanwhile, the test data set can be used to test a prediction model created based on data from multiple alcohol use disorder patients.

The test data set can be set large enough to produce statistically significant results, and can be the entire data of multiple alcohol use disorder patients. The test data group can be classified to have the same characteristics as the learning data group.

The prediction model generator 120 may use a test data set to derive a performance evaluation index to be used to measure the performance of the prediction model. The prediction model generator 120 can use performance evaluation indicators to check the objective performance of the prediction model and compare the performance between different prediction models.

Figure 4 is a diagram showing an operation of the preprocessor 110 performing oversampling on a specific class according to embodiments of the present invention.

Referring to FIG. 4, when processing data of multiple alcohol use disorder patients, the preprocessor 110 may apply a sampling method to a specific class to resolve class imbalance in the learning data group.

For example, for the outpatient treatment maintenance class and the outpatient treatment early discontinuation class, which are dependent variables for the learning data group, the number of data included in the outpatient treatment maintenance class and the outpatient treatment early discontinuation class is imbalanced (e.g. 85:15). Problems may arise.

If a prediction model is learned using data with class imbalance problems, biased results may be derived, and it may be difficult to accurately predict whether patients with alcohol use disorder will discontinue treatment. Therefore, sampling methods can be applied to solve the class imbalance problem of such data.

As an example, sampling methods can be divided into oversampling methods and undersampling methods.

The undersampling method is a method of reducing the majority class data group to the level of the minority class data group. The undersampling method removes a large number of class data, which can reduce computation time and reduce class overlap. However, the undersampling method can drastically reduce the total number of data used for learning, thereby deteriorating learning performance.

On the other hand, the oversampling method secures sufficient data for learning by increasing the data group of minority classes to the level of majority classes.

For example, oversampling methods include random oversampling to adjust the ratio by simply replicating existing minority classes, and SMOTE (Sythetic Minority), which is a method of generating new data between neighboring minority classes from data of a random minority class. Methods such as Over-Sampling Technique may be used.

When processing data, the preprocessor 110 can correct class imbalance using oversampling or undersampling methods and derive more precise predictions. Meanwhile, in embodiments of the present invention, the preprocessor 110 is configured to solve the imbalance problem between the outpatient treatment maintenance class and the outpatient treatment early discontinuation class, which are dependent variables for the learning data group, and the minority class among the two classes ( Oversampling can be applied to (e.g. outpatient treatment maintenance classes).

Referring to FIG. 4, as an example, the preprocessor 110 may generate duplicated data for a and b among data included in the minority class for a minority class.

Figure 5 is a diagram showing an operation of the prediction model generator 120 according to embodiments of the present invention to generate a prediction model by applying one or more machine learning algorithms to a group of learning data.

Referring to FIG. 5, the prediction model generator 120 may generate a prediction model by applying one or more machine learning algorithms to a portion of the processed data corresponding to the learning data group.

At this time, one or more machine learning algorithms may be one or more of Logistic Regression, Support Vector Machine (SVM), Random Forest, Gradient Boosting, and Adaboost. .

Logistic Regression is a statistical technique that uses a logistic function to estimate the causal relationship between a dependent variable that has only two values and an independent variable. The dependent variable is dichotomous (0 or 1), and the independent variable can be categorical or continuous.

The logistic regression model is a special form of the generalized linear model and is a functional model that draws an S-shaped curve. As a result of the logistic regression analysis, if the value of the dependent variable is greater than 0.5, the event is predicted to occur, and if it is less than 0.5, the event is predicted to not occur.

Support Vector Machine (SVM) is one of the machine learning fields. It is a supervised learning model for pattern recognition and data analysis, and is mainly used for classification and regression analysis. The support vector machine algorithm can create a non-probabilistic binary linear classification model that determines which category new data belongs to, given a set of data belonging to one of two categories.

In this case, the categories can be divided into a group of patients with alcohol use disorder maintaining outpatient treatment and a group of early discontinuation of outpatient treatment, and a support vector machine can be used to determine which of the two groups new data falls into.

Random Forest is a type of ensemble learning method used in regression analysis, etc., and operates by outputting classification or average prediction values from multiple decision trees constructed during the training process.

The random forest testing process using an ensemble model can derive the final result by averaging, multiplying, or majority voting the results obtained from the decision tree. These tests can be performed in parallel, resulting in high computational efficiency.

Gradient Boosting is a machine learning algorithm that can perform regression or classification analysis, and is an algorithm that belongs to the boosting family of ensemble methodologies of machine learning algorithms.

Boosting is the process of creating a strong classifier by combining weak classifiers, and gradient boosting takes the error of the data predicted by the model in the previous step and creates a new step model with the goal of reducing this error to 0. This model It is an algorithm that combines them to create a model.

Adaboost is a machine learning algorithm that expresses the final result by adding weight to the results of other learning algorithms.

Meanwhile, the above-described machine learning algorithm is an example, and embodiments of the present invention are not limited thereto.

The prediction model generator 120 may apply a machine learning algorithm to a training data set to generate a prediction model corresponding to the machine learning algorithm.

Figure 6 is a diagram showing an operation of the prediction model generator 120 determining a prediction model according to a performance evaluation index according to embodiments of the present invention.

Referring to FIG. 6, when there are multiple machine learning algorithms, the prediction model generator 120 generates a plurality of machine learning algorithms by applying them to the portion corresponding to the learning data group among the processed data received from the preprocessor 110. For each of the plurality of candidate prediction models, test results can be derived by inputting a portion of the processed data corresponding to the test data group.

And the prediction model generator 120 may calculate a performance evaluation index for each of the plurality of candidate prediction models using the derived test results.

In addition, the prediction model generator 120 may determine the candidate prediction model with the highest performance evaluation index among the plurality of candidate prediction models as the prediction model.

At this time, the performance evaluation index may be, for example, one of Accuracy, Sensitivity, Specificity, and AUC (Area under the ROC curve).

Accuracy is the number of data with the same prediction result (TP + TN) divided by the total number of prediction data (TP + FP + FN + TN), and is an indicator that determines how much the predicted data is the same as the actual data. Accuracy refers to the proportion of all patients who correctly decided whether to discontinue outpatient treatment or maintain outpatient treatment. At this time, TP is the number of data that was actually positive when the prediction model was predicted to be positive, FP is the number of data that was actually negative when the prediction model was predicted to be positive, and FN is the number of data that was actually negative when the prediction model was predicted to be positive. TN refers to the number of data that was predicted to be positive but was actually positive, and TN refers to the number of data that the prediction model predicted to be negative but was actually negative.

Sensitivity is also called recall rate or hit rate, and is the ratio of what is actually positive (TP) among what the prediction model predicts to be positive (TP + FP), and is the ratio of what is actually positive (TP), which results in stopping actual outpatient treatment. This refers to the percentage of patients in whom the prediction model was correct.

Specificity is the ratio of actual negatives (TN) among those predicted by the prediction model to be negative (TN + FP), and refers to the proportion of patients maintained in outpatient treatment who were actually hit by the prediction model.

AUC can be obtained from the ROC (Receiver Operating Characteristics) curve and means the True Positive Rate according to the False Positive Rate, which means (1 - specificity) according to sensitivity.

AUC is the area under the ROC curve, the maximum is 1, and the better the prediction model, the closer the AUC value is to 1.

The prediction model generator 120 may select a prediction model that can predict whether a patient with alcohol use disorder will discontinue outpatient treatment early with the highest probability using the above-mentioned performance evaluation index among the plurality of candidate prediction models.

As an example, the prediction model generator 120 may calculate a performance evaluation index for each candidate prediction model and generate Table 2.

Model AUC Accuracy Sensitivity Specificity Logistic Regression 0.6914 0.6130 0.7058 0.6026 SVM 0.6797 0.7023 0.6470 0.7086 Random Forest 0.6365 0.7380 0.4705 0.7682 Gradient Boosting 0.6022 0.7083 0.4117 0.7417 Adaboost 0.7241 0.6428 0.7647 0.6291

According to Table 2, if the prediction model is determined based on AUC or sensitivity, a candidate prediction model using the Adaboost algorithm may be determined as the prediction model.

On the other hand, if the prediction model is determined based on Accuracy or Specificity, a candidate prediction model using the Random Forest algorithm may be determined as the prediction model.

Figure 7 is a diagram showing AUC, one of the performance evaluation indicators according to embodiments of the present invention.

Referring to FIG. 7 , as an example, the prediction model generator 120 may determine a prediction model using AUC as a performance evaluation index.

As an example, the ROC curve may be determined as shown in FIG. 7. AUC refers to the area under the ROC curve, and the prediction model generator 120 can check the AUC value by calculating the area under the ROC curve for each candidate prediction model.

Referring to FIG. 7, since the AUC value of Adaboost is the largest, the prediction model generator 120 may select a prediction model using Adaboost.

Figure 8 is a diagram showing a method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder according to embodiments of the present invention.

Referring to FIG. 8, the method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder may include a data collection step (S810) of collecting data on a plurality of patients with alcohol use disorder.

And the method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder is an independent variable that determines a plurality of independent variables to which one or more machine learning algorithms will be applied to create a prediction model for early discontinuation of outpatient treatment for patients with alcohol use disorder. A variable determination step (S820) may be included.

In addition, the method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder may include a preprocessing step (S830) of generating processed data by processing data of a plurality of patients with alcohol use disorder.

Meanwhile, the above-described data collection step (S810), independent variable determination step (S820), and preprocessing step (S830) may be executed by the above-described preprocessor 110.

And the method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder is to receive processed data, set whether or not early discontinuation of outpatient treatment for multiple alcohol use disorder patients is a dependent variable, and based on the independent variables, determine the total of the processed data. Alternatively, it may include a prediction model generation step (S840) in which a prediction model is generated by applying one or more machine learning algorithms to some parts. Meanwhile, the prediction model creation step (S840) may be executed by the prediction model creation unit 120 described above.

In addition, the method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder includes a prediction step (S850) in which all or part of the processed data is input into a prediction model to generate a prediction result regarding whether multiple alcohol use disorder patients will initially discontinue outpatient treatment. may include. Meanwhile, the prediction step (S850) may be performed by the prediction unit 130 described above.

Additionally, the method for predicting early cessation of outpatient treatment for patients with alcohol use disorder may include an output step (S860) that outputs the prediction result. Meanwhile, the output step (S860) may be executed by the output unit 140 described above.

For example, in the independent variable determination step (S820), when determining the independent variables, calculate Variance Inflation Factors (VIF) for the independent variables to solve the multicollinearity problem between the independent variables. , independent variables can be determined so that the variance expansion coefficient remains below a preset threshold.

As an example, the preprocessing step (S830) may include classifying a plurality of alcohol use disorder patient data into a learning data group and a test data group.

And the preprocessing step (S830) may include applying oversampling to a minority class among the outpatient treatment maintenance class and the outpatient treatment early discontinuation class, which are dependent variables for the learning data group.

For example, in the prediction model creation step (S840), a prediction model may be generated by applying one or more machine learning algorithms to a portion of the processed data corresponding to the learning data group. At this time, one or more machine learning algorithms include: 1) Logistic Regression, 2) Support Vector Machine (SVM), 3) Random Forest, 4) Gradient Boosting, and 5 ) It may be one or more of Adaboost.

Meanwhile, the prediction model creation step (S840) is, for example, 1) when there are multiple machine learning algorithms, a plurality of candidate prediction models generated by applying a plurality of machine learning algorithms to the portion corresponding to the learning data group among the processed data. For each of these, a step of deriving test results by inputting a portion of the processed data corresponding to the test data group, 2) calculating a performance evaluation index for each of a plurality of candidate prediction models using the test results, 3) ) It may include the step of determining the candidate prediction model with the highest performance evaluation index among the plurality of candidate prediction models as the prediction model.

At this time, the performance evaluation index may be AUC (Area under the ROC curve).

The system 100 for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder described above may be implemented by a computing device that includes at least some of a processor, memory, user input device, and presentation device. Memory is a medium that stores computer-readable software, applications, program modules, routines, instructions, and/or data that are coded to perform specific tasks when executed by a processor. The processor may read and execute computer-readable software, applications, program modules, routines, instructions, and/or data stored in memory. A user input device may be a means for allowing a user to input a command that causes the processor to execute a specific task or to input data required to execute a specific task. User input devices may include a physical or virtual keyboard, keypad, key buttons, mouse, joystick, trackball, touch-sensitive input means, or microphone. Presentation devices may include displays, printers, speakers, or vibrating devices.

Computing devices may include a variety of devices such as smartphones, tablets, laptops, desktops, servers, and clients. A computing device may be a single stand-alone device or may include multiple computing devices operating in a distributed environment comprised of multiple computing devices cooperating with each other through a communication network.

In addition, the method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder described above includes a processor, and is a computer-readable software and application coded to perform an imaging diagnosis method using a deep learning model when executed by the processor. , can be executed by a computing device having a memory storing program modules, routines, instructions, and/or data structures, etc.

The above-described embodiments can be implemented through various means. For example, the present embodiments may be implemented by hardware, firmware, software, or a combination thereof.

In the case of hardware implementation, the present embodiments include one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), and Field Programmable Gates (FPGAs). Arrays), processors, controllers, microcontrollers, or microprocessors.

For example, the method for predicting early cessation of outpatient treatment for patients with alcohol use disorder according to embodiments can be implemented using an artificial intelligence semiconductor device in which neurons and synapses of a deep neural network are implemented with semiconductor devices. there is. At this time, the semiconductor device may be currently used semiconductor devices, such as SRAM, DRAM, or NAND, or may be next-generation semiconductor devices such as RRAM, STT MRAM, or PRAM, or a combination thereof.

When implementing the method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder according to embodiments using an artificial intelligence semiconductor device, the results (weights) of learning the deep learning model with software are transferred to the synaptic mimetic elements arranged in an array. Learning can also be carried out in artificial intelligence semiconductor devices.

In the case of implementation by firmware or software, the method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder according to the present embodiments may be implemented in the form of a device, procedure, or function that performs the functions or operations described above. . Software code can be stored in a memory unit and run by a processor. The memory unit is located inside or outside the processor and can exchange data with the processor through various known means.

Additionally, terms such as "system", "processor", "controller", "component", "module", "interface", "model", or "unit" described above generally refer to computer-related entities hardware, hardware and software. It may refer to a combination of, software, or running software. By way of example, but not limited to, the foregoing components may be a process, processor, controller, control processor, object, thread of execution, program, and/or computer run by a processor. For example, both an application running on a controller or processor and the controller or processor can be a component. One or more components may reside within a process and/or thread of execution, and the components may be located on a single device (e.g., system, computing device, etc.) or distributed across two or more devices.

Meanwhile, another embodiment provides a computer program stored in a computer recording medium that performs the method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder described above. Another embodiment also provides a computer-readable recording medium recording a program for realizing the method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder described above.

The program recorded on the recording medium can be read, installed, and executed on the computer to execute the above-described steps.

In this way, in order for the computer to read the program recorded on the recording medium and execute the functions implemented by the program, the above-mentioned program is a C, C++ program that the computer's processor (CPU) can read through the computer's device interface (Interface). , may include code coded in computer languages such as JAVA and machine language.

These codes may include functional codes related to functions defining the above-mentioned functions, etc., and may also include control codes related to execution procedures necessary for the computer processor to execute the above-described functions according to predetermined procedures.

In addition, these codes may further include memory reference-related codes that determine which location (address address) in the computer's internal or external memory the additional information or media required for the computer's processor to execute the above-mentioned functions should be referenced. .

In addition, if the computer's processor needs to communicate with any other remote computer or server in order to execute the above-mentioned functions, the code is It may further include communication-related codes for how to communicate with other computers, servers, etc., and what information or media should be transmitted and received during communication.

Recording media that can be read by a computer recording the above-described program include, for example, ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical media storage, etc., and also include carrier wave (e.g. , transmission via the Internet) may also be implemented.

Additionally, computer-readable recording media can be distributed across computer systems connected to a network, so that computer-readable code can be stored and executed in a distributed manner.

In addition, the functional program for implementing the present invention and the code and code segments related thereto are designed by programmers in the technical field to which the present invention belongs, taking into account the system environment of the computer that reads the recording medium and executes the program. It can also be easily inferred or changed by .

The method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder can also be implemented in the form of a recording medium containing instructions executable by a computer, such as an application or program module executed by a computer. Computer-readable media can be any available media that can be accessed by a computer and includes both volatile and non-volatile media, removable and non-removable media. Additionally, computer-readable media may include all computer storage media. Computer storage media includes both volatile and non-volatile, removable and non-removable media implemented in any method or technology for storage of information such as computer-readable instructions, data structures, program modules or other data.

The method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder can be executed by an application installed by default on the terminal (this may include programs included in the platform or operating system, etc. installed by default on the terminal), and the user can access the application store. It may also be executed by an application (i.e. program) installed directly on the master terminal through an application providing server such as a server, application, or web server related to the service. In this sense, the method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder described above is implemented as an application (i.e., program) installed by default on the terminal or directly installed by the user and stored on a computer-readable recording medium such as the terminal. can be recorded

The description of the present invention described above is for illustrative purposes, and those skilled in the art will understand that the present invention can be easily modified into other specific forms without changing the technical idea or essential features of the present invention. will be. Therefore, the embodiments described above should be understood in all respects as illustrative and not restrictive. For example, components described as single may be implemented in a distributed manner, and similarly, components described as distributed may also be implemented in a combined form.

The scope of the present invention is indicated by the claims described later than the detailed description above, and all changes or modified forms derived from the meaning and scope of the claims and their equivalent concepts should be construed as being included in the scope of the present invention. do.

100: Prediction system for early discontinuation of outpatient treatment for patients with alcohol use disorder
110: Preprocessing unit
120: Prediction model generation unit
130: prediction unit
140: output unit

Claims (14)

Collecting data from a plurality of alcohol use disorder patients, determining a plurality of independent variables to which one or more machine learning algorithms will be applied to generate a prediction model for early discontinuation of outpatient treatment for the plurality of alcohol use disorder patients, A pre-processing unit that processes data of multiple alcohol use disorder patients and generates processed data;
Receiving the processed data, setting whether or not early outpatient treatment of the plurality of patients with alcohol use disorder is discontinued as a dependent variable, and applying the one or more machine learning algorithms to all or part of the processed data based on the independent variables a prediction model generator that generates the prediction model;
a prediction unit that inputs all or part of the processed data into the prediction model to generate a prediction result regarding whether to discontinue outpatient treatment early for the plurality of alcohol use disorder patients; and
A system for predicting early cessation of outpatient treatment for patients with alcohol use disorder, including an output unit that outputs the prediction result.
According to paragraph 1,
The preprocessor,
When determining the independent variables, calculate Variance Inflation Factors (VIF) for the independent variables to solve the multicollinearity problem between the independent variables,
A system for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder, wherein the independent variables are determined to maintain the variance inflation coefficient below a preset threshold.
According to paragraph 1,
The preprocessor,
A system for predicting early cessation of outpatient treatment for alcohol use disorder patients, which classifies the data of the plurality of alcohol use disorder patients into a learning data group and a test data group.
According to paragraph 3,
The preprocessor,
A system for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder that applies oversampling to the minority class among the outpatient treatment maintenance class and the outpatient treatment early discontinuation class, which are dependent variables for the learning data group.
According to paragraph 4,
The prediction model generator,
Generating the prediction model by applying the one or more machine learning algorithms to a portion of the processed data corresponding to the learning data group,
The one or more machine learning algorithms include:
Predicting early discontinuation of outpatient treatment in patients with alcohol use disorder using one or more of Logistic Regression, Support Vector Machine (SVM), Random Forest, Gradient Boosting, and Adaboost. system.
According to clause 5,
The prediction model generator,
When the machine learning algorithm is plural, each of a plurality of candidate prediction models generated by applying a plurality of machine learning algorithms to a portion corresponding to the learning data group among the processed data, and the test data group among the processed data Enter the corresponding part to derive the test result,
Calculate a performance evaluation index for each of the plurality of candidate prediction models using the test results,
A system for predicting early cessation of outpatient treatment for patients with alcohol use disorder, wherein among the plurality of candidate prediction models, the candidate prediction model with the highest performance evaluation index is determined as the prediction model.
According to clause 6,
The performance evaluation indicators are,
A system for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder using AUC (Area under the ROC curve).
In the method of predicting early discontinuation of outpatient treatment in patients with multiple alcohol use disorders,
A data collection step of collecting data of the plurality of alcohol use disorder patients through a preprocessor included in the system according to any one of claims 1 to 7, and whether to discontinue outpatient treatment early for the plurality of alcohol use disorder patients. an independent variable determination step of determining a plurality of independent variables to which one or more machine learning algorithms will be applied to create a prediction model for, and a pre-processing step of generating processed data by processing the data of the plurality of alcohol use disorder patients;
Receiving the processed data through a prediction model generator included in the system according to any one of claims 1 to 7, and setting as a dependent variable whether or not outpatient treatment for the plurality of alcohol use disorder patients is stopped early, A prediction model generation step of generating the prediction model by applying the one or more machine learning algorithms to all or part of the processed data based on the independent variables;
Through the prediction unit included in the system according to any one of claims 1 to 7, all or part of the processed data is input into the prediction model to determine whether to initially discontinue outpatient treatment for the plurality of alcohol use disorder patients. A prediction step of generating a prediction result regarding: and
A method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder, comprising: an output step of outputting the prediction result through an output unit included in the system according to any one of claims 1 to 7.
According to clause 8,
The independent variable determination step is,
When determining the independent variables, calculate Variance Inflation Factors (VIF) for the independent variables to solve the multicollinearity problem between the independent variables,
A method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder, wherein the independent variables are determined to maintain the variance inflation coefficient below a preset threshold.
According to clause 8,
The preprocessing step is,
A method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder, comprising: classifying the plurality of data on patients with alcohol use disorder into a learning data group and a test data group.
According to clause 10,
The preprocessing step is,
A method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder, comprising: applying oversampling to minority classes among the outpatient treatment maintenance class and the outpatient treatment early discontinuation class, which are dependent variables for the learning data group.
According to clause 11,
The prediction model creation step is,
Generating the prediction model by applying the one or more machine learning algorithms to a portion of the processed data corresponding to the learning data group,
The one or more machine learning algorithms include:
Predicting early discontinuation of outpatient treatment in patients with alcohol use disorder using one or more of Logistic Regression, Support Vector Machine (SVM), Random Forest, Gradient Boosting, and Adaboost. method.
According to clause 12,
The prediction model creation step is,
When the machine learning algorithm is plural, for each of a plurality of candidate prediction models generated by applying a plurality of machine learning algorithms to a portion corresponding to the learning data group among the processed data, the test data group among the processed data Deriving a test result by inputting the corresponding part;
calculating a performance evaluation index for each of the plurality of candidate prediction models using the test results; and
A method for predicting early discontinuation of outpatient treatment for patients with alcohol use disorder, comprising: determining a candidate prediction model with the highest value of the performance evaluation index among the plurality of candidate prediction models as the prediction model.
According to clause 13,
The performance evaluation indicators are,
A method for predicting early discontinuation of outpatient treatment in patients with alcohol use disorder using AUC (Area under the ROC curve).

Images