KR20200049290A - Method and system for screening diabetes using multiple biological signals - Google Patents

Abstract

A diabetes screening method and a diabetes screening system using multiple biosignals are disclosed. Diabetes screening method using multiple bio-signals according to an embodiment of the present invention comprises the steps of measuring multiple bio-signals about a patient, and using a plurality of bio-signals selected from the multiple bio-signals, a plurality of individual diagnostic models Step of establishing, and considering the performance related to the prediction of diabetes, determining the order of the plurality of individual diagnostic models, and integrating the plurality of individual diagnostic models according to the determined order to form an integrated diagnostic model, And determining whether the patient is diabetic based on the final predicted probability value calculated using the integrated diagnostic model.

Description

Diabetes screening method and diabetes screening system using multiple bio-signals {METHOD AND SYSTEM FOR SCREENING DIABETES USING MULTIPLE BIOLOGICAL SIGNALS}

The present invention relates to a non-invasive diabetes screening technique using multiple biosignals, and more particularly, to an improvement in the accuracy and reliability of diabetes screening through the integration of individual prediction models for diabetic discrimination.

Various prior literatures are known to present a technique for diagnosing diabetes using a non-invasive biosignal using a bioelectric method or a spectral method.

1) US Publication Number: 2012-0065514 (2012.03.15), "Cardiohealth Methods and Apparatus"

2) US Publication Number: 10-2010-0081941 (2010.04.01), "CARDIOVASCULAR HEALTH STATION METHODS AND APPARATUS"

3) Korean registration number: 10-1440735 (2014.09.04), "A blood sugar measurement system with minimal number of blood collection and its method"

Prior literatures measure the electrocardiogram by measuring the impedance and conductivity by applying a current to the human body, and measuring diabetes and body resistance based on the measured biometric information by shooting infrared rays, or measuring the pulse by applying infrared rays and diagnosing diabetes based on this. It discloses about detecting.

As described above, prior literatures suggest a non-invasive diabetes screening technique that directly predicts blood sugar based on the measured bio-information, but cannot provide a predictive probability value for diabetes using information obtained from multiple bio-signals.

In addition, in the existing prediction model, since the discrimination result may be different depending on the type of the biosignal, an integrated prediction model capable of predicting diabetes by analyzing a plurality of biosignals is required, but by simply merging individual prediction models. , May have limitations in improving the accuracy of diabetes prediction.

Accordingly, in the present specification, an individual prediction model constructed using multiple biosignals can be sequentially combined according to the high and low performances to construct an integrated prediction model that can more accurately predict the risk of suspected diabetes. I would like to propose a technique.

An embodiment of the present invention aims to improve the accuracy of diabetic discrimination through an integrated prediction model utilizing multiple bio signals.

According to an embodiment of the present invention, when forming an integrated predictive model for diagnosing diabetes by combining individual predictive models constructed by using multiple biosignals, the level of accuracy of diabetic discrimination is improved by It aims at improving reliability.

In the embodiment of the present invention, in the case of a suspected diabetic group whose predicted probability value is in the range of '0.45 to 0.55%,' a certain number of individual diagnostic models with high diagnostic performance for each biosignal are sequentially combined, and then any biosignal is displayed. It is an object to propose an integrated model capable of diagnosing diabetes more accurately by forming an integrated diagnostic model by additionally applying individual diagnostic models according to.

In an embodiment of the present invention, when an arbitrary biosignal is added to improve the accuracy of diabetes screening, a method of effectively integrating individual diagnostic models according to the arbitrary biosignals into an existing diagnostic model is proposed. It aims to improve accuracy through expansion of diagnostic models.

Diabetes screening method using multiple bio-signals according to an embodiment of the present invention comprises the steps of measuring multiple bio-signals about a patient, and using a plurality of bio-signals selected from the multiple bio-signals, a plurality of individual diagnostic models Step of establishing, and considering the performance related to the prediction of diabetes, determining the order for the plurality of individual diagnostic models, and integrating the plurality of individual diagnostic models according to the determined order to form an integrated diagnostic model, and And determining whether the patient is diabetic based on the final predicted probability value calculated using the integrated diagnostic model.

In addition, the diabetes screening system using multiple bio-signals according to an embodiment of the present invention includes a plurality of individual biomedical signals selected from the multiple bio-signals and a measurement unit for measuring multiple bio-signals related to a patient. A model building unit for constructing a diagnostic model, a performance evaluation unit for determining the order of the plurality of individual diagnostic models in consideration of performance related to diabetes prediction, and an integrated diagnostic model by integrating the plurality of individual diagnostic models according to the determined order It may include, and based on the final predicted probability value calculated using the integrated diagnostic model, may include a processing unit for determining whether the patient is diabetic.

According to an embodiment of the present invention, when predicting diabetes based on individual diabetes prediction models constructed by analyzing multiple biosignals of two or more of bioimpedance, electrochemical skin conductivity, and near infrared rays, cases with unpredictable prediction (0.45 ~ 0.55%), the predictive model can be synthesized sequentially according to the performance of each predictive model to accurately predict diabetes.

According to an embodiment of the present invention, reliability and accuracy of screening can be improved through model integration when developing a non-invasive diabetes screening device using multiple bio signals.

According to an embodiment of the present invention, more accurate diagnosis can be performed because the probability of diabetes is predicted by deriving repetitive probability values using information obtained through multiple bio signals.

According to an embodiment of the present invention, a diabetes diagnosis model is formed by using information obtained through multiple biosignals, but the performance of each model is grasped, and a case designated as a suspected diabetes group is diagnosed with excellent performance. By re-diagnosing by sequentially applying the model, the accuracy of diabetes prediction can be improved.

According to an embodiment of the present invention, when using multiple biosignal measurement equipment such as 'BIA' and 'NIRS' as a diabetes screening tool, the accuracy can be improved.

According to one embodiment of the present invention, predicting diabetes by generating a predictive model for diabetes using variable information obtained by measuring bioimpedance and predicting diabetes using variable information obtained by measuring electrochemical skin conductivity. Predict diabetes by generating a model, predict the diabetes by generating a predictive model for diabetes using variable information obtained by shooting near infrared rays to the human body, and use each predictive model sequentially according to the height of diabetes prediction performance to suspect diabetes It is possible to accurately predict whether a group is at risk for diabetes.

1 is a block diagram showing the configuration of a diabetes screening system using multiple biosignals according to an embodiment of the present invention.
2A to 2C are graphs showing results of performance evaluation of diabetic discrimination according to individual diagnostic models in the diabetes screening system according to an embodiment of the present invention.
3 is a graph showing a result of performance evaluation according to an integrated diagnostic model in a diabetes screening system according to an embodiment of the present invention.
4 is a flowchart illustrating a procedure of a diabetes screening method using multiple bio signals according to an embodiment of the present invention.

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. However, various changes may be made to the embodiments, and the scope of the patent application right is not limited or limited by these embodiments. It should be understood that all modifications, equivalents, or substitutes for the embodiments are included in the scope of rights.

The terms used in the examples are used for illustrative purposes only and should not be construed as limiting. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this specification, the terms "include" or "have" are intended to indicate the presence of features, numbers, steps, actions, components, parts or combinations thereof described herein, one or more other features. It should be understood that the existence or addition possibilities of fields or numbers, steps, operations, components, parts or combinations thereof are not excluded in advance.

Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by a person skilled in the art to which the embodiment belongs. Terms, such as those defined in a commonly used dictionary, should be interpreted as having meanings consistent with meanings in the context of related technologies, and should not be interpreted as ideal or excessively formal meanings unless explicitly defined in the present application. Does not.

In addition, in the description with reference to the accompanying drawings, the same reference numerals are assigned to the same components regardless of reference numerals, and redundant descriptions thereof will be omitted. In describing the embodiments, when it is determined that detailed descriptions of related known technologies may unnecessarily obscure the subject matter of the embodiments, detailed descriptions thereof will be omitted.

1 is a block diagram showing the configuration of a diabetes screening system using multiple biosignals according to an embodiment of the present invention.

Referring to FIG. 1, a diabetes screening system 100 using multiple biosignals according to an embodiment of the present invention includes a measurement unit 110, a model building unit 120, a processing unit 130, and a performance evaluation unit 140 ) And the memory unit 150.

The measurement unit 110 measures multiple bio signals from a patient.

That is, the measurement unit 110 may simultaneously measure multiple bio-signals related to the patient's diabetes level, such as a bio-impedance signal, an electrochemical skin conductivity signal, and a near-infrared signal.

For example, the measurement unit 110 may measure a bioimpedance signal in response to an alternating current applied to a patient based on multiple frequencies through an impedance meter (BIA).

In addition, the measurement unit 110 may measure a conductivity signal in response to a DC current applied to the patient through a conductivity meter (ESC).

In addition, the measurement unit 110 may measure a near-infrared signal irradiated to the skin of a patient through a near-infrared measuring instrument (NIRS).

The measurement unit 110 may record multiple bio-signals measured for a period of time in the memory unit 150 in order to collect multiple bio-signals required for learning modeling.

In one example, the measurement unit 110 responds to a bioimpedance signal that reacts to the alternating current applied to the patient measured through an impedance meter (BIA), and a direct current applied to the patient measured through a conductivity meter (ESC). A biosignal of at least one of a conductivity signal and a near-infrared signal irradiated to the skin of the patient measured through a near-infrared instrument (NIRS) may be recorded in the memory unit 150 as the multiple bio-signals.

The model building unit 120 builds a plurality of individual diagnostic models using a plurality of bio signals selected from the multiple bio signals.

The model building unit 120 analyzes two or more types of biological signals selected from the memory unit 150 according to an ensemble technique among logistic regression models or machine learning algorithms to which a demerit function is applied, and analyzes each of the plurality of individual diagnostic models. Can build.

For example, the model building unit 120 analyzes the bioimpedance signal measured through an impedance measuring instrument (BIA) according to a logistic regression model to which a penalty function is applied, and calculates a predicted probability value for diabetic prediction. Build A.

In addition, the model construction unit 120 may construct a diagnostic model B that analyzes the conductivity signal measured through a conductivity meter (ESC) according to an ensemble technique and calculates a predicted probability value for diabetic prediction.

In addition, the model construction unit 120 analyzes the near-infrared signal measured through a near-infrared measuring instrument (NIRS) according to a combination of a logistic regression model and the ensemble technique to which a penalty function is applied, and calculates a predicted probability value for diabetic prediction Diagnostic model C can be constructed.

The performance evaluation unit 140 determines an order for the plurality of individual diagnostic models in consideration of performance related to diabetes prediction.

That is, the performance evaluation unit 140 analyzes the bioimpedance signal, the conductivity signal, and the near-infrared signal measured from the patient, respectively, and evaluates performance of the plurality of individual diagnostic models, based on the evaluation results of the performance, The order for sequentially integrating each of the plurality of individual diagnostic models may be determined.

The above sequence is a sequence number for sequentially and stepwisely integrating each of the plurality of individual diagnostic models, and the performance evaluation unit 140 evaluates the diabetes prediction performance of the plurality of individual diagnostic models, and the individual diagnostic model having excellent diabetes prediction performance. You can give a relatively high order of precedence.

Specifically, the performance evaluation unit 140 evaluates whether the diabetes prediction result according to the predicted probability value calculated using the individual diagnostic model is close to the actual diabetes diagnosis result of the patient, and based on the evaluation result, the The above sequence for sequentially integrating each of a plurality of individual diagnostic models may be determined.

For example, the performance evaluation unit 140 includes diagnostic model A constructed using the patient's bioimpedance signal, diagnostic model B constructed using the patient's conductivity signal, and diagnostic model C constructed using the near-infrared signal. Diabetes prediction is performed by calculating the predicted probability value, and according to the degree to which the results of the diabetes prediction coincide (proximally) with the diagnosis results of the actual patient, the prediction performance of the individual diagnosis model can be evaluated.

The performance evaluation unit 140 determines the order of diagnostic model A having the best performance as '1', and then determines the order of diagnostic model C having the highest performance as '2', and the order of diagnostic model B having the lowest performance Can be set to '3'.

In another example, the performance evaluation unit 140 evaluates an optimal coefficient and tuning parameter through cross-validation of each of a plurality of individual diagnostic models, and based on the evaluation results of the optimal coefficient and tuning parameter, the plurality of The above sequence for sequentially integrating each individual diagnostic model can be determined.

That is, the performance evaluation unit 140 may evaluate the superiority (high and low) of performance of each individual diagnostic model by comparing the optimal coefficients and tuning parameters of the individual diagnostic models selected through cross-validation verification.

For example, referring to the graphs of FIGS. 2A to 2C, the performance evaluation unit 140 performs '10 -fold 'cross-validation to select two tuning parameters that represent optimal performance from a plurality of individual diagnostic models. Through this, it is possible to select the tuning parameter having the maximum AUC value and compare the selected tuning parameter, thereby giving a relatively advanced order to individual diagnostic models having high tuning parameters.

In addition, the performance evaluation unit 140 evaluates a plurality of individual diagnostic models to estimate the optimal coefficients, so that the processing unit 130, which will be described later, can integrate the individual diagnostic models step by step in the order of high performance according to the optimal coefficients. In addition, it is possible to determine whether to repeatedly perform integration of individual diagnostic models by providing a graph (see 321 of FIG. 3) comparing the performance of the integrated diagnostic models in stages.

The processing unit 130 integrates each of a plurality of individual diagnostic models according to the above-described order in consideration of diabetes prediction performance, thereby forming an integrated diagnostic model.

For example, the processing unit 130 selects and integrates two individual diagnostic models in order of high performance among a plurality of individual diagnostic models, and selects and integrates the individual diagnostic models having the next highest performance in the integrated two individual diagnostic models. In this way, a plurality of individual diagnostic models can be integrated sequentially and stepwise.

Particularly, when the prediction of diabetes by each individual diagnostic model is ambiguous or a risk of diabetes, the processor 130 uses the integrated diagnostic model in which individual diagnostic models according to different biosignals are synthesized step by step to predict diabetes (diabetes) Discrimination).

For example, the processing unit 130 has a predicted probability value calculated by an individual diagnostic model having the best performance among a plurality of individual diagnostic models corresponding to '0.45 to 0.55%', such that the diabetes prediction is ambiguous or belongs to a diabetes risk group For, diagnostic models can be sequentially synthesized according to the performance of each diagnostic model, and diabetes prediction can be performed through the synthesized integrated diagnostic model.

Hereinafter, a process of synthesizing a plurality of individual diagnostic models constructed with multiple biosignals in a stepwise and sequential manner in consideration of diabetes prediction performance will be described.

Specifically, the processor 130 calculates a first predicted probability value from a first linear predicted value calculated by the first individual diagnostic model having the highest order, and a range in which the first predicted probability value is selected (eg, '0.45 ~ 0.55% '), by calculating the average linear predicted value by adding the second linear predicted value calculated by the high-order second individual diagnostic model excluding the first individual diagnostic model to the first linear predicted value. , The first and second individual diagnostic models can be integrated.

Here, the selected range is a range for the predicted probability value when it is ambiguous to determine whether or not the patient is diabetic based on the predicted probability value calculated using the individual diagnostic model, or when it is estimated to belong to the diabetes risk group, as '0.45 to 0.55%' Can be illustrated.

As such, the processing unit 130 may improve the accuracy of diabetic discrimination through sequential and stepwise integration considering the performance of a plurality of individual diagnostic models.

On the other hand, for the case where the diabetes prediction is not ambiguous, the processing unit 130 does not need to integrate the second individual diagnostic model into the first individual diagnostic model, and the first individual diagnostic model having the best diabetes prediction performance. Diabetes can be used as it is (diabetes prediction).

That is, the processor 130, if the first predicted probability value calculated by the first individual diagnostic model having the best performance, does not fall within the selected range (eg, '0.45 to 0.55%'), the first prediction A probability value may be determined as the final predicted probability value, and the diabetes may be determined according to the final predicted probability value.

According to an embodiment, when a third individual diagnostic model different from the first and second individual diagnostic models exists, the processor 130 further sums the third linear predicted value calculated by the third individual diagnostic model, By calculating the average linear predicted value, the third individual diagnostic model may be further integrated into the integrated first and second individual diagnostic models.

For example, the processing unit 130 may select a third individual diagnostic model having a higher order than the second individual diagnostic model among a plurality of individual diagnostic models, and further integrate the first and second individual diagnostic models. have.

In addition, the processing unit 130 may arbitrarily select a third individual diagnostic model from among the plurality of individual diagnostic models except for the first and second individual diagnostic models regardless of the order, and to the first and second individual diagnostic models. It can also be integrated further.

In addition, the processing unit 130 selects a third individual diagnostic model constructed by using a biosignal different from the first and second individual diagnostic models among a plurality of individual diagnostic models, and the first and second individual diagnostic models. In addition to the integration.

Further, the processing unit 130 may further integrate any individual diagnostic model that is built before or after the plurality of individual diagnostic models, as the third individual diagnostic model, into the first and second individual diagnostic models. have.

As such, the processing unit 130 may further improve the accuracy of diabetic discrimination based on the integrated diagnostic model by additionally performing integration of individual diagnostic models constructed with different biosignals in a model in which the individual diagnostic models are integrated.

According to an embodiment, the processing unit 130 may repeat the process of additionally integrating as many as the number of the third individual diagnostic models, if a third individual diagnostic model that is not further integrated remains.

At this time, each time the processing unit 130 performs the integration of the third individual diagnostic model, the performance evaluation unit 140 evaluates the performance related to diabetes prediction, and determines whether to further integrate another third individual diagnostic model. You can decide.

If the diabetic prediction performance is evaluated to be excellent by the performance evaluation unit 140, the processing unit 130 may stop further integration of the third individual diagnostic model.

As described above, the processing unit 130 may continuously improve the accuracy of diabetic discrimination of the integrated diagnostic model by repeatedly performing integration with the individual diagnostic model constructed with an arbitrary biosignal for the integrated diagnostic model.

The processor 130 determines whether the patient is diabetic based on the final predicted probability value calculated using the integrated diagnostic model.

Specifically, the processor 130 calculates a predicted probability value from the average linear predicted value calculated by the integrated diagnostic model, determines the predicted probability value as the final predicted probability value, and determines whether or not the diabetes according to the final predicted probability value. Can be determined.

Here, the average linear predicted value may be calculated as an average value of linear predicted values calculated by each of a plurality of individual diagnostic models combined upon formation of the integrated diagnostic model.

For example, the performance evaluation unit 140 may be configured with individual diagnostic models ('diagnostic model A') constructed from the bioimpedance signals measured by an impedance meter (BIA), and the conductivity signals measured by a conductivity meter (ESC). Among the individual diagnostic models ('diagnostic model B') constructed with and the individual diagnostic models ('diagnostic model C') constructed with the near-infrared signal measured by a near-infrared instrument (NIRS), according to diabetes prediction performance, performance The order of the best diagnostic model A can be set to '1', and then the order of the best diagnostic model C is set to '2', and the order of the lowest diagnostic model B can be set to '3'.

The processing unit 130 is a predicted probability value (p ₁ ) calculated using the diagnostic model A having the best diabetes prediction performance, and predicts diabetes, but the predicted probability value (p ₁ ) is a selected range ('0.45≤p ₁ ≤ 0.55 '), it can be determined that the result of diabetes prediction is ambiguous or belongs to the diabetes risk group.

)을 연산하고, 상기 평균 선형예측값으로부터 예측확률값(p₁₊₃)을 다시 산출할 수 있다.In this case, the processing unit 130 sums the linear predicted value (η ₁ ) calculated by the diagnostic model A having the best performance, and the linear predicted value (η ₃ ) calculated by the diagnostic model C having the next best diabetes prediction performance. After averaging, the average linear predicted value (

) Is calculated, and the predicted probability value (p _{1 + 3} ) can be calculated again from the average linear predicted value.

)으로부터, 최종 예측확률값(p₁₊₂₊₃)을 산출할 수 있다.Subsequently, the processing unit 130 averages the linear predicted value calculated by further summing the linear predicted values (η ₃ ) calculated by the diagnostic model B constructed using the bio signals different from the diagnostic models A and C (the 'skin conductivity signal').

), The final predicted probability value (p _{1 + 2 + 3} ) can be calculated.

As described above, according to the present invention, in the case of a suspected diabetic group having a predicted probability value of, for example, '0.45 to 0.55%,' a certain number of individual diagnostic models with high diagnostic performance for each biosignal are sequentially combined, and then By forming an integrated diagnostic model by additionally applying an individual diagnostic model according to a biosignal, the accuracy of diabetic discrimination can be further improved.

2A to 2C are graphs showing results of performance evaluation of diabetic discrimination according to individual diagnostic models in the diabetes screening system according to an embodiment of the present invention.

FIG. 2A shows a graph showing the performance of diagnostic model A constructed from bioimpedance signals measured by an impedance meter (BIA).

2B is a graph showing the performance of Diagnostic Model B constructed with a conductivity signal measured by a conductivity meter (ESC).

2C, a graph showing the performance of the diagnostic model C constructed with a near infrared signal measured by a near infrared measuring instrument (NIRS) is shown.

Diabetes screening system, through the graph of FIGS. 2A to 2C, compares the performance of individual diagnostic models constructed through multiple biosignals, and integrates a plurality of individual diagnostic models in order of superior performance (diagnostic model A> Diagnostic model C> Diagnostic model B) can be determined.

For example, the diabetes screening system selects the tuning parameter having the maximum AUC value through '10 -fold 'cross-validation, and selects the two tuning parameters showing optimal performance from a plurality of individual diagnostic models. By comparing the tuning parameters, a relatively advanced order can be given to individual diagnostic models with high tuning parameters.

3 is a graph showing a result of performance evaluation according to an integrated diagnostic model in a diabetes screening system according to an embodiment of the present invention.

Referring to Figure 3, the diabetes screening system according to an embodiment of the present invention, among the plurality of individual diagnostic models A, B, C (311, 312, 313), the most excellent diagnostic model A (311), The diagnostic model C 313 having the next best performance is first synthesized, and then the diagnostic model B 312 having the best performance is further synthesized to form an integrated diagnostic model 322.

That is, the diabetes screening system is a diagnostic model A (311) constructed with a bioimpedance signal measured by an impedance measuring instrument (BIA), and a diagnostic model C (313) constructed with a near infrared signal measured by a near infrared measuring instrument (NIRS). Is synthesized, and the synthesized diagnostic model (A + C) 321 is further synthesized by synthesizing the diagnostic model B 312 constructed from the conductivity signal measured by the conductivity meter (ESC), and the integrated diagnostic model 322. Can form.

As described above, the diabetes screening system can improve the accuracy and reliability of diabetic discrimination through step-by-step integration in consideration of the performance of individual prediction models constructed using multiple bio signals.

Hereinafter, FIG. 4 describes in detail the workflow of the diabetes screening system 100 according to embodiments of the present invention.

4 is a flowchart illustrating a procedure of a diabetes screening method using multiple bio signals according to an embodiment of the present invention.

The diabetes screening method according to this embodiment may be performed by the above-described diabetes screening system 100.

Referring to FIG. 4, in step 410, the diabetes screening system 100 measures multiple biosignals of a patient for a period of time.

As an example, the diabetes screening system 100 measures a bioimpedance signal responsive to an alternating current applied to a patient based on multiple frequencies through an impedance meter (BIA), and through a conductivity meter (ESC) to a patient. The conductivity signal in response to the applied DC current can be measured, and a near-infrared signal irradiated to the skin of a patient can be measured through a near infrared meter (NIRS).

In step 420, the diabetes screening system 100 constructs a plurality of individual diagnostic models using two or more types of bio signals selected from the multiple bio signals.

The diabetes screening system 100 may analyze two or more types of biosignals according to an ensemble technique of a logistic regression model or a machine learning algorithm to which a penalty function is applied, and construct each of the plurality of individual diagnostic models.

In step 430, the diabetes screening system 100 integrates a plurality of individual diagnostic models step by step in a determined order in consideration of performance related to diabetes prediction.

As an example, the diabetes screening system 100 evaluates an optimal coefficient and tuning parameter through cross-validation of each of a plurality of individual diagnostic models, and based on the evaluation results regarding the optimal coefficient and tuning parameter, the plurality of The above sequence for sequentially integrating each individual diagnostic model can be determined.

For example, referring to the graphs of FIGS. 2A to 2C, the diabetes screening system 100 cross-validates '10 -fold 'in order to select two tuning parameters indicating optimal performance from a plurality of individual diagnostic models. Through this, it is possible to select the tuning parameter having the maximum AUC value, compare the selected tuning parameters, and give a relatively advanced order to individual diagnostic models having high tuning parameters.

In step 440, the diabetes screening system 100 determines whether the patient is diabetic based on the final predicted probability value calculated according to the integrated diagnostic model.

In this step 440, the diabetes screening system 100, the predicted probability value calculated by the individual diagnostic model having the best performance among the plurality of individual diagnostic models corresponds to '0.45 ~ 0.55%', the diabetes prediction is ambiguous or , For the cases belonging to the diabetes risk group, the diagnostic model can be sequentially synthesized according to the high and low performance of each diagnostic model, and the diabetes prediction can be performed through the synthesized integrated diagnostic model.

For example, referring to FIGS. 2A to 2C, the diabetes screening system 100 includes a performance graph (FIG. 2A) of a diagnostic model A constructed with a bioimpedance signal measured by an impedance meter (BIA), and a near infrared meter. Performance graph by comparing the performance graph of the diagnostic model C (FIG. 2B) constructed with the near-infrared signal measured by (NIRS), and the diagnostic model B (FIG. 2C) constructed with the conductivity signal measured by the conductivity meter (ESC). In this excellent order, a sequence to be integrated into a plurality of individual diagnostic models (diagnostic model A> diagnostic model C> diagnostic model B) is given, and an integrated diagnostic model can be formed according to the above sequence.

Referring to FIG. 3, the diabetes screening system 100 is constructed with a near infrared signal measured by a near infrared ray meter (NIRS) in a diagnostic model A 311 built with a bioimpedance signal measured by an impedance meter (BIA). One diagnostic model C (313) is synthesized first, and the synthesized diagnostic model (A + C) 321 is further synthesized by constructing a diagnostic model B (312) constructed from conductivity signals measured by a conductivity meter (ESC). Thus, an integrated diagnostic model 322 can be formed.

As described above, according to an embodiment of the present invention, when diabetic prediction is performed based on an individual diabetes prediction model constructed by analyzing multiple biosignals of two or more of bioimpedance, electrochemical skin conductivity, and near-infrared, a case with unpredictable prediction For (0.45 ~ 0.55%), prediction models can be synthesized sequentially according to the performance of each prediction model to accurately predict diabetes.

According to an embodiment of the present invention, reliability and accuracy of screening can be improved through model integration when developing a non-invasive diabetes screening device using multiple bio signals.

According to an embodiment of the present invention, more accurate diagnosis can be performed because the probability of diabetes is predicted by deriving repetitive probability values using information obtained through multiple bio signals.

According to an embodiment of the present invention, a diabetes diagnosis model is formed by using information obtained through multiple biosignals, but the performance of each model is grasped, and a case designated as a suspected diabetes group is diagnosed with excellent performance. By re-diagnosing by sequentially applying the model, the accuracy of diabetes prediction can be improved.

According to an embodiment of the present invention, when using multiple biosignal measurement equipment such as 'BIA' and 'NIRS' as a diabetes screening tool, the accuracy can be improved.

According to an embodiment of the present invention, a predictive model for diabetes is predicted by generating a predictive model for diabetes using variable information obtained by measuring bioimpedance and prediction for diabetes using variable information obtained by measuring electrochemical skin conductivity. Predict diabetes by generating a model, predict the diabetes by generating a predictive model for diabetes using variable information obtained by shooting near-infrared rays to the human body, and use each predictive model sequentially according to the height of diabetes prediction performance to suspect diabetes It is possible to accurately predict whether a group is at risk for diabetes.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, or the like alone or in combination. The program instructions recorded in the medium may be specially designed and configured for the embodiments or may be known and usable by those skilled in computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs, DVDs, and magnetic media such as floptical disks. -Hardware devices specifically configured to store and execute program instructions such as magneto-optical media, and ROM, RAM, flash memory, and the like. Examples of program instructions include high-level language codes that can be executed by a computer using an interpreter, etc., as well as machine language codes produced by a compiler. The hardware device described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

The software may include a computer program, code, instruction, or a combination of one or more of these, and configure the processing device to operate as desired, or process independently or collectively You can command the device. Software and / or data may be interpreted by a processing device, or to provide instructions or data to a processing device, of any type of machine, component, physical device, virtual equipment, computer storage medium or device. , Or may be permanently or temporarily embodied in the transmitted signal wave. The software may be distributed over networked computer systems, and stored or executed in a distributed manner. Software and data may be stored in one or more computer-readable recording media.

As described above, although the embodiments have been described with limited drawings, those skilled in the art can apply various technical modifications and variations based on the above. For example, the described techniques are performed in a different order than the described method, and / or the components of the described system, structure, device, circuit, etc. are combined or combined in a different form from the described method, or other components Alternatively, even if replaced or substituted by equivalents, appropriate results can be achieved.

Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

100: diabetes screening system
110: measuring unit 120: model building unit
130: processing unit 140: performance evaluation unit
150: memory unit

Claims (15)

Measuring multiple biosignals related to the patient;
Constructing a plurality of individual diagnostic models using a plurality of biological signals selected from the multiple biological signals;
Determining an order for the plurality of individual diagnostic models in consideration of performance related to diabetes prediction;
Forming an integrated diagnostic model by integrating the plurality of individual diagnostic models according to the predetermined order; And
Determining whether the patient is diabetic based on the final predicted probability value calculated using the integrated diagnostic model
Diabetes screening method using multiple bio-signals comprising a. According to claim 1,
The step of determining the order,
Evaluating whether the diabetes prediction result according to the predicted probability value calculated using the individual diagnostic model is close to the actual diabetes diagnosis result of the patient; And
Determining the order for sequentially integrating each of the plurality of individual diagnostic models based on the evaluation result
Diabetes screening method comprising a. According to claim 1,
The step of determining the order,
Evaluating an optimal coefficient and tuning parameter through cross-validation of each of the plurality of individual diagnostic models; And
Determining the order for sequentially integrating each of the plurality of individual diagnostic models based on the evaluation results regarding the optimal coefficients and tuning parameters
Diabetes screening method comprising a. According to claim 1,
The step of determining the order,
Evaluating performance of the plurality of individual diagnostic models by analyzing bioimpedance signals, conductivity signals, and near infrared signals measured from the patient, respectively; And
Determining the order for sequentially integrating each of the plurality of individual diagnostic models based on the evaluation result regarding the performance
Diabetes screening method comprising a. According to claim 1,
The step of forming the integrated diagnostic model,
Calculating a first predicted probability value from a first linear predicted value calculated by the first individual diagnostic model having the highest order; And
When the first predicted probability value falls within a selected range,
The first and second individual by calculating the average linear predicted value by adding the second linear predicted value calculated by the second individual diagnostic model having a high order, excluding the first individual diagnostic model, to the first linear predicted value Steps to Integrate Diagnostic Models
Diabetes screening method comprising a. The method of claim 5,
When there is a third individual diagnostic model different from the first and second individual diagnostic models,
The step of forming the integrated diagnostic model,
The third individual diagnostic model is further integrated into the integrated first and second individual diagnostic models by further summing the third linear predicted value calculated by the third individual diagnostic model and calculating the average linear predicted value. To do; And
If the third individual diagnostic model that has not been further integrated remains, repeating the further integration step
Diabetes screening method further comprising. The method of claim 5,
The step of determining whether or not the diabetes,
Determining a predicted probability value calculated from the average linear predicted value as the final predicted probability value, and determining whether or not the diabetes according to the final predicted probability value
Diabetes screening method comprising a. The method of claim 5,
If the first predicted probability value does not fall within the selected range,
The step of determining whether or not the diabetes,
Determining the first predicted probability value as the final predicted probability value, and determining whether or not the diabetes according to the final predicted probability value
Diabetes screening method comprising a. According to claim 1,
A bioimpedance signal responsive to an alternating current applied to the patient measured through an impedance meter (BIA), a conductivity signal responsive to a direct current applied to the patient measured through a conductivity meter (ESC), and a near infrared meter (NIRS ) Recording at least one bio-signal of the near-infrared signal irradiated to the skin of the patient measured through) as the multiple bio-signals, in a memory unit
Diabetes screening method further comprising. The method of claim 9,
The step of constructing the plurality of individual diagnostic models,
Analyzing two or more types of biological signals selected from the memory unit according to an ensemble technique among a logistic regression model or a machine learning algorithm to which a penalty function is applied, and constructing each of the plurality of individual diagnostic models
Diabetes screening method comprising a. A measurement unit for measuring multiple bio-signals about the patient;
A model building unit for constructing a plurality of individual diagnostic models using a plurality of bio signals selected from the multiple bio signals;
A performance evaluation unit that determines an order for the plurality of individual diagnostic models in consideration of performance related to diabetes prediction; And
A processing unit that forms an integrated diagnostic model by integrating the plurality of individual diagnostic models according to the predetermined order, and determines whether or not diabetes is present in the patient based on a final predicted probability value calculated using the integrated diagnostic model
Diabetes screening system using multiple bio-signals comprising a. The method of claim 11,
The performance evaluation unit,
Optimal coefficients and tuning parameters are evaluated through cross-validation of each of the plurality of individual diagnostic models,
Based on the evaluation result regarding the optimal coefficient and tuning parameter, the order for sequentially integrating each of the plurality of individual diagnostic models is determined.
Diabetes screening system. The method of claim 11,
The processing unit,
The first predicted probability value is calculated from the first linear predicted value calculated by the first individual diagnostic model having the highest order,
When the first predicted probability value falls within a selected range,
The first and second individual by calculating the average linear predicted value by adding the second linear predicted value calculated by the second individual diagnostic model having a high order, excluding the first individual diagnostic model, to the first linear predicted value Incorporating diagnostic models
Diabetes screening system. The method of claim 13,
When there is a third individual diagnostic model different from the first and second individual diagnostic models,
The processing unit,
The third individual diagnostic model is further integrated into the integrated first and second individual diagnostic models by further summing the third linear predicted value calculated by the third individual diagnostic model and calculating the average linear predicted value. And if the third individual diagnostic model that is not further integrated remains, repeating the further integration process
Diabetes screening system. The method of claim 11,
The model building unit,
Two or more types of biosignals are selected from the memory unit recording the multiple biosignals, and the selected two or more types of biosignals are analyzed according to an ensemble method among logistic regression models or machine learning algorithms to which a penalty function is applied. To build each diagnostic model
Diabetes screening system.

Images