KR20210026711A - Method and Apparatus for Developing Risk Prediction Model of Stroke of Diabetic Patients - Google Patents

Abstract

The present invention relates to a method and a device for designing a risk prediction model of a stroke of diabetic patients, which includes the steps of: receiving, by an electronic device, a design signal for a stroke risk prediction model; collecting, by the electronic device, cohort data according to the design signal; establishing a development cohort and a validation cohort for the diabetic patients based on the cohort data; calculating, by the electronic device, a risk level for a plurality of risk factors for the stroke; and designing the stroke risk prediction model by calculating the contribution of each risk factor based on the risk of each risk factor. Other embodiments are also applicable. According to the present invention, it is possible to more accurately predict the risk of stroke in Korean diabetic patients.

Description

Method and Apparatus for Developing Risk Prediction Model of Stroke of Diabetic Patients}

The present invention relates to a method and apparatus for designing a model for predicting stroke risk in diabetic patients.

Diabetes mellitus refers to a group of metabolic disorders in which high blood sugar levels persist for a long period of time, and the cause is that the pancreas does not make enough insulin or the cells of the body do not respond properly to the made insulin. Chronic hyperglycemia due to lack of insulin action, etc., accompanies a number of characteristic metabolic abnormalities. Since insulin is mainly involved in carbohydrate metabolism, diabetes is a basic problem in carbohydrate metabolism. However, since this affects the metabolism of all nutrients in the body, it can be said to be an overall metabolic disease.

Such diabetes is considered to be the most important chronic disease in the modern world, and in particular, it may cause complications such as cardio-cerebrovascular disease, stroke, chronic kidney failure, and diabetic ketoacidosis. Among complications caused by diabetes, cardio-cerebrovascular disease is the main cause of death. Compared to Westerners, Koreans have a higher incidence of cerebrovascular disease, that is, stroke. Various risk prediction models have been developed to predict the occurrence of these cerebrovascular diseases. However, the conventional risk prediction model does not target only diabetic patients, but is accurate in predicting the risk of cerebrovascular disease for Asians, especially Koreans, as there are many studies targeting the entire population or Westerners. There is a problem that is very deteriorated.

Embodiments of the present invention for solving such a conventional problem are to provide a method and apparatus for designing a stroke risk prediction model of a diabetic patient capable of predicting the risk of stroke occurrence in Korean diabetic patients.

In addition, embodiments of the present invention are a method for designing a stroke risk prediction model for diabetic patients that can provide a stroke risk prediction model that can be used as an important index for establishing a treatment plan in consideration of risk factors for causing stroke in Korean diabetic patients. And a device.

In the method of designing a model for predicting stroke risk of a diabetic patient according to an embodiment of the present invention, the electronic device receives a design signal for a stroke risk prediction model, and the electronic device collects cohort data according to the design signal. The step of, establishing a development cohort and a validation cohort for diabetic patients based on the cohort data, calculating, by the electronic device, a risk for each of a plurality of risk factors for stroke, and the And designing the stroke risk prediction model by calculating the contribution level for each risk factor based on the risk level for each risk factor.

In addition, the step of calculating the risk of multiple risk factors for stroke includes age, sex, smoking status, exercise level, BMI level, chronic kidney disease, coronary artery disease, duration of diabetes, use of insulin or hypoglycemic drugs. It is characterized in that it is a step of calculating the number, fasting blood sugar level, systolic blood pressure level, total cholesterol level, and risk of atrial fibrillation.

In addition, the step of designing the stroke risk prediction model by calculating the contribution of each risk factor is a step of designing the stroke risk prediction model by converting the contributions to the plurality of risk factors into a score between 0 and 100. To do.

In addition, it characterized in that it further comprises the step of verifying the stroke risk prediction model.

In addition, the verifying the stroke risk prediction model includes dividing the converted score for each of the plurality of risk factors into a plurality of sections, and a stroke occurrence prediction rate in each divided section based on the development cohort and the verification cohort. It characterized in that it comprises the step of performing the verification of the stroke risk prediction model by calculating each.

In addition, the apparatus for designing a model for predicting stroke risk of a diabetic patient according to an embodiment of the present invention includes an input unit for inputting a design signal for a stroke risk prediction model, a communication unit for receiving cohort data according to the design signal, and the cohort. Based on the data, a development cohort and a validation cohort for diabetic patients are established, and the risk of multiple risk factors and the contribution of each risk factor are calculated to design the stroke risk prediction model. It characterized in that it comprises a control unit.

In addition, multiple risk factors include age, sex, smoking status, exercise level, BMI level, chronic kidney disease, coronary artery disease, duration of diabetes, the number of insulin or hypoglycemic drugs used, fasting blood sugar level, systolic blood pressure level. , Characterized in that the total cholesterol level and whether or not atrial fibrillation.

In addition, the control unit is characterized in that the stroke risk prediction model is designed by converting the contributions to the plurality of risk factors into a score between 0 and 100.

In addition, the control unit divides the converted score for each of the plurality of risk factors into a plurality of sections, and predicts the stroke risk by calculating a stroke occurrence prediction rate in each divided section based on the development cohort and the verification cohort, respectively. It is characterized in that the model is verified.

As described above, the method and apparatus for designing a stroke risk prediction model for diabetic patients according to the present invention can more accurately predict the risk of stroke occurrence in Korean diabetic patients by designing a stroke risk prediction model based on cohort data for Korean diabetic patients. There is an effect that can be.

In addition, the method and apparatus for designing a model for predicting stroke risk of a diabetic patient according to the present invention according to the present invention are important in establishing a treatment plan in consideration of risk factors for causing stroke in a diabetic patient in a diabetic patient's stroke risk prediction model. There is an effect that can be used as an indicator.

1 is a diagram showing an apparatus for designing a stroke risk prediction model according to an embodiment of the present invention.
2 is a flowchart illustrating a method of designing a stroke risk prediction model according to an embodiment of the present invention.
3 is a view obtained by converting the contribution of each risk factor for causing a stroke into a score according to an exemplary embodiment of the present invention.
4 is a graph showing a stroke incidence rate within 5 years corresponding to a total score calculated as a contribution score for each risk factor according to an embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. DETAILED DESCRIPTION OF THE INVENTION The detailed description to be disclosed below together with the accompanying drawings is intended to describe exemplary embodiments of the present invention, and is not intended to represent the only embodiments in which the present invention may be practiced. In the drawings, parts irrelevant to the description may be omitted in order to clearly describe the present invention, and the same reference numerals may be used for the same or similar components throughout the specification.

In an embodiment of the present invention, expressions such as “or” and “at least one” may represent one of words listed together, or a combination of two or more. For example, “A or B” and “at least one of A and B” may include only one of A or B, and may include both A and B.

1 is a diagram showing an apparatus for designing a stroke risk prediction model according to an embodiment of the present invention.

Referring to FIG. 1, the electronic device 100, which is an apparatus for designing a stroke risk prediction model according to the present invention, includes a communication unit 110, an input unit 120, a display unit 130, a memory 140, and a control unit 150. .

The communication unit 110 communicates with an external device (not shown) that provides cohort data, for example, a server managed by the National Health Insurance Service. To this end, the communication unit 110 may perform communication such as 5G (5 ^th generation communication), LTE-A (long term evolution-advanced), LTE (long term evolution), WiFi (wireless fidelity).

The input unit 120 generates input data in response to a user input of the electronic device 100. The input unit 120 includes at least one input means. Input unit 120 is a keyboard (key board), keypad (key pad), dome switch (dome switch), touch panel (touch panel), touch key (touch key), mouse (mouse), menu button (menu button), etc. It may include.

The display unit 130 displays display data according to an operation of the electronic device 100. The display unit 130 includes a liquid crystal display (LCD), a light emitting diode (LED) display, an organic light emitting diode (OLED) display, and a micro electro mechanical systems (MEMS) display. Including ray and electronic paper displays. The display unit 130 may be combined with the input unit 120 to be implemented as a touch screen.

The memory 140 stores operation programs of the electronic device 100. The memory 140 may store cohort data received from the National Health Insurance Service, and may store a development cohort and a validation cohort built by the control unit 150. In addition, the memory 140 stores the risk of stroke calculated by the control unit 150 and stores the contribution of each risk factor converted into a score based on the risk. The memory 140 may store a stroke risk prediction model designed by the controller 150.

The control unit 150 establishes a development cohort and a validation cohort for diabetic patients based on the cohort data, and calculates the risk for each risk factor and the contribution level for each risk factor for a stroke, Design a stroke risk prediction model for

More specifically, when a design signal for designing a stroke risk prediction model is received from the input unit 120, the control unit 150 controls the communication unit 110 to collect cohort data from the National Health Insurance Service. In this case, the cohort data collected by the control unit 150 may be cohort data for Korean diabetic patients between 40 and 65 years of age, for example. The control unit 150 may classify cohort data based on 13 risk factors to establish a development cohort, and may establish a validation cohort for verification of the development cohort.

Subsequently, the control unit 150 calculates the risk for stroke of the 13 risk factors, converts the calculated risk to the contribution of each risk factor, and completes the design of the stroke risk prediction model. More specifically, the control unit 150 includes age, sex, smoking status, exercise level, BMI level, chronic kidney disease, coronary artery disease, diabetes duration, the number of insulin or hypoglycemic drugs used, fasting blood sugar level, systolic blood pressure. The contribution to each risk factor is converted into a score between 0 and 100 by assigning weights to risk factors such as numerical value, total cholesterol level, and atrial fibrillation. When it is confirmed that the design of the stroke risk prediction model is completed, the control unit 150 verifies the designed stroke risk prediction model. The control unit 150 divides the scores converted for each of a plurality of risk factors into a plurality of, for example, 10 sections, and calculates a stroke incidence prediction rate in each divided section based on the development cohort and the verification cohort, respectively. The control unit 150 performs verification of a stroke risk prediction model by comparing a stroke incidence prediction rate calculated based on the development cohort and the verification cohort in the same section.

2 is a flowchart illustrating a method of designing a stroke risk prediction model according to an embodiment of the present invention.

Referring to FIG. 2, in step 201, the control unit 150 checks whether or not a design signal for designing a stroke risk prediction model is received from the input unit 120. As a result of checking in step 201, when the design signal is received, the control unit 150 performs step 203, and if the design signal is not received, the control unit 150 waits for reception of the design signal.

In step 203, the controller 150 controls the communication unit 110 to collect cohort data from the National Health Insurance Service. In this case, the cohort data collected by the control unit 150 may be cohort data for Korean diabetic patients between 40 and 65 years of age, for example. In step 205, the controller 150 constructs a development cohort from the cohort data and performs step 207. In step 207, the control unit 150 establishes a validation cohort.

More specifically, the controller 150 extracts cohort data for a diabetic subject, that is, a patient, from the cohort data, and extracts cohort data of a diabetic and stroke patient. In this case, as the extracted cohort data, cohort data for 1,297,131 people may be extracted. The controller 150 may check clinical information of a diabetic patient and a diabetic and stroke patient from the extracted cohort data.

The control unit 150 may classify the extracted cohort data based on 13 risk factors to establish a development cohort as shown in Table 1 below, and may establish a verification cohort to confirm the verification of the development cohort. At this time, the verification cohort may be established as cohort data for 389,139 people.

Referring to Table 1 below, the control unit 150 divides the age range into three stages (40-49, 50-59, 60-64) to establish a development cohort, and checks whether smoking or not and exercise level. . In addition, the control unit 150 checks the BMI level by dividing it into three stages (<18.5, 18.5-23, >=23), and whether chronic kidney disease (CKD), coronary artery disease (CHD), and diabetes duration Check if it is more than 5 years. The control unit 150 divides and checks the number of insulin or hypoglycemic drugs used and the fasting blood glucose level (FBG) in five steps (<100, 100-140, 140-160, 160-180, >-180), and systolic blood pressure. Check the number (SBP) by dividing it into 5 steps (<130, 130-140, 140-150, 150-160, >=160). In addition, the control unit 150 divides and checks the total cholesterol level into four steps (<160, 160-200, 200-240, >=240), and checks whether atrial fibrillation has occurred.

Development cohort (n=1,297,131) Verification cohort (n=389,139) Stroke(-) Stroke(+) P-value Stroke(-) Stroke(+) P-value N 1,262,444 34,687 378,683 10,456 Age <.0001 <.0001 40-49 358,773(28.4) 4,649(13.4) 107,162(28.3) 1,397(13.4) 50-59 590,308(46.8) 15,689 (45.2) 177,342(46.8) 4,706(45.0) 60-64 313,363(24.8) 14,349(41.4) 94,179 (24.9) 4,353(41.6) Sex <.0001 <.0001 Male 807,750 (64.0) 23,765(68.5) 242,190 (64.0) 7,129 (68.2) Female 454,697(36.0) 10,922(31.5) 136,493(36.0) 3,327(31.8) Smoking status <.0001 <.0001 current 362,007 (28.7) 12,792(36.9) 108,805(28.7) 3,852(36.8) Regular exercise <.0001 <.0001 YES 669,098 (53.0) 15,731(45.4) 200,702(53.0) 4,783(45.7) BMI <.0001 <.0001 <18.5 13,161 (1.08) 595(1.7) 4,154(1.1) 174(1.7) 18.5-23 292,281(23.2) 8,959(25.8) 87,819 (23.2) 2,705(25.9) >=23 956,547(75.8) 25,133(72.5) 286,710(75.7) 7,577(72.5) CKD <.0001 <.0001 YES 83,548(6.6) 4,423(12.8) 24,935(6.6) 1,320(12.6) CHD <.0001 <.0001 YES 44,668(3.5) 2,25996.5) 13,446(3.6) 703(6.7) Diabetic duration(years) <.0001 <.0001 >=5 378,745(30.0) 15,516(44.7) 113,477 (30.0) 4,675(44.7) Number of OHAs or use or insulin <.0001 <.0001 >=2, insulin 581,115 (46.0) 20,834(60.1) 174,191(46.0) 6,336(60.6) FBG <.0001 <.0001 <100 102,100(8.1) 3,415(9.9) 30,421(8.0) 1,055 (10.1) 100-140 602,665 (47.7) 13,351(38.5) 181,398(47.9) 4,105(39.3) 140-160 237,681 (18.8) 5,774(16.7) 70,843 (18.7) 1,703(16.3) 160-180 114,636 (9.1) 3,513(10.1) 34,557(9.1) 1,034(9.9) >=180 205,362(16.3) 8,634(24.9) 61,464(16.2) 2,559(24.9) SBP <.0001 <.0001 <130 637,804(50.5) 14,558 (42.0) 191,027 (50.5) 4,423(42.3) 130-140 373,140(29.6) 10,279(29.6) 111,861(29.5) 3,095(29.6) 140-150 129,162(10.2) 4,373(12.6) 38,924(10.3) 1,341(12.8) 150-160 73,932(5.9) 2,865(8.3) 22,391(5.9) 848(8.1) >=160 48,406(3.8) 2,612(7.5) 14,480(3.8) 749(7.2) TC <.0001 <.0001 <160 234,298 (18.6) 6,394(18.4) 70,435 (18.6) 1,909(18.3) 160-200 466,221(36.9) 12,249(35.3) 139,928(34.0) 3,686(35.3) 200-240 376,532 (29.8) 10,358 (29.9) 112,613(29.7) 3,169(30.3) >=240 185,393 (14.7) 5,686(16.4) 55,707(1407) 1,692(16.2) Atrial fibrillation <.0001 <.0001 YES 10,4719(0.8) 1,021(2.9) 3,117(0.8) 298(2.9)

Subsequently, in step 209, the control unit 150 calculates the risk (95% confidence intervals) of the 13 risk factors for stroke. The risk is as in Table 2 below, and when the number of each risk factor is collected in Table 2 below, it is the same as the number of development cohorts (1,297,131) in Table 1. In Table 2 below, the values for Model 1 are values considering only age and sex, and Model 2 is age, sex, smoking status, exercise level, BMI level, chronic kidney disease, coronary artery disease, diabetes duration, insulin Or, the number of hypoglycemic drugs used, fasting blood sugar levels, systolic blood pressure levels, total cholesterol levels, and risk factors such as atrial fibrillation are all considered. For example, in the 60 to 64 year old group, the stroke risk according to age and gender is 3.53 (95% CI 3.41-3.65), and the stroke risk according to multiple risk factors in the same group is 3.18 (95% CI 3.07-). 3.29).

Number Events Incidence rate(per 1000 person-years) Model 1 Model 2 Age 40-49 363,422 4,649 1.82 1(ref.) 1(ref.) 50-59 605,997 15,689 3.64 2.05 (1.99-2.12) 1.96 (1.90-2.03) 60-64 327,712 14,349 6.10 3.53 (3.41-3.65) 3.18 (3.07-3.29) Sex Male 831,515 23,765 4.04 1(ref.) 1(ref.) Female 465,616 10,922 3.27 0.70 (0.69-0.72) 0.81 (0.79-0.83) Smoking status No 922,332 21,895 3.32 1(ref.) 1(ref.) current 374,799 12,792 4.87 1.69 (1.65-1.73) 1.65 (1.61-1.69) Regular exercise NO 612,302 18,956 4.37 1(ref.) 1(ref.) YES 684,829 15,731 3.22 0.75 (0.73-0.76) 0.79 (0.78-0.81) BMI <18.5 14,211 595 6.01 1(ref.) 1(ref.) 18.5-23 301,240 8,959 4.18 0.69 (0.63-0.74) 0.73 (0.67-0.79) >=23 981,680 25,133 3.60 0.60 (0.56-0.65) 0.66 (0.60-0.71) CKD NO 1,209,160 30,264 3.52495 1(ref.) 1(ref.) YES 87,971 4,423 6.97195 1.64 (1.59-1.70) 1.52 (1.47-1.57) CHD NO 1,250,204 32,428 3.65 1(ref.) 1(ref.) YES 46,927 2,259 6.77 1.48 (1.42-1.54) 1.37 (1.31-1.43) Diabetic duration(years) <5 902,870 19,171 3.01 1(ref.) 1(ref.) >=5 394,261 15,516 5.44 1.55 (1.52-1.59) 1.31 (1.27-1.34) Number of OHAs or use or insulin 0, 1 695,182 13,853 2.85 1(ref.) 1(ref.) >=2, insulin 601,949 20,834 4.77 1.52 (1.49-1.56) 1.31(1.28-1.34) FBG <100 105,515 3,415 4.47 1.33(1.28-1.38) 1.21(1.16-1.25) 100-140 616,016 13,351 3.06 1(ref.) 1(ref.) 140-160 243,455 5,774 3.34 1.13 (1.10-1.17) 1.07 (1.04-1.11) 160-180 118,149 3,513 4.16 1.42 (1.37-1.47) 1.25 (1.20-1.30) >=180 213,996 8,634 5.66 2.04 (1.99-2.10) 1.65 (1.60-1.70) SBP <130 652,362 14,558 3.14 1(ref.) 1(ref.) 130-140 383,419 10,279 3.77 1.15 (1.12-1.17) 1.186 (1.156-1.217) 140-150 133,535 4,373 4.62 1.34(1.29-1.39) 1.39 (1.34-1.44) 150-160 76,797 2,865 5.26 1.51 (1.45-1.58) 1.55 (1.49-1.61) >=160 51,018 2,612 7.25 2.09 (2.00-2.17) 2.07 (1.98-2.15) TC <160 240,692 6,394 3.73 1(ref.) 1(ref.) 160-200 478,470 12,249 3.59 1.03 (1.00-1.06) 1.09 (1.05-1.12) 200-240 386,890 10,358 3.77 1.15 (1.11-1.19) 1.24 (1.20-1.28) >=240 191,079 5,686 4.22 1.36 (1.31-1.41) 1.42 (1.37-1.48) Atrial fibrillation NO 1,285,691 33,666 3.68 1(ref.) 1(ref.) YES 11,440 1,021 12.89 2.82 (2.66-3.00) 2.67 (2.50-2.85)

In step 211, when the risk of each risk factor is calculated as shown in Table 2, the control unit 150 converts the contribution of each risk factor to complete the design of the stroke risk prediction model. More specifically, the control unit 150 includes age, sex, smoking status, exercise level, BMI level, chronic kidney disease, coronary artery disease, diabetes duration, the number of insulin or hypoglycemic drugs used, fasting blood sugar level, systolic blood pressure. The contribution to each risk factor is converted into a score between 0 and 100 by assigning weights to risk factors such as numerical value, total cholesterol level, and atrial fibrillation. At this time, the contribution of each risk factor is shown in Table 3 and Fig. 3 below, and the stroke incidence rate corresponding to the total score calculated based on the contribution of the risk factor is shown in Fig. 4. 3 is a view obtained by converting the contribution of each risk factor for causing a stroke into a score according to an exemplary embodiment of the present invention. 4 is a graph showing a stroke incidence rate within 5 years corresponding to a total score calculated as a contribution score for each risk factor according to an embodiment of the present invention.

No. Variable Categories and score One Age (years) 40-49 50-59 60-64 0 59 100 2 Sex Male Female 19 0 3 Current smoker No Current 0 41 4 Physical activity No Yes 21 0 5 BMI <18.5 18.5-23 >23.0 24 12 0 6 CKD No Yes 0 38 7 CHD No Yes 0 28 8 Diabetic duration <5 ≥5 0 25 9 Number of OHAs or Use of Insulin 0-1 ≥2 or insulin 0 22 10 FBG levels <100 100-140 140-160 160-180 ≥180 16 0 7 20 43 11 SBP levels <130 130-140 140-150 150-160 ≥160 0 16 29 40 66 12 TC levels <160 160-200 200-240 ≥240 0 6 16 29 13 Atrial fibrillation No YES 0 88

Taking Table 3 and Figure 3 as an example, a 62-year-old male (19 points) with diabetes is a smoker (41 points), does not exercise at all (21 points), and BMI is 22 (12 points). , CKD and CHD are absent, and the duration of diabetes is more than 5 years (25 points). In addition, if the patient is on insulin (22 points), FBG is 170 (20 points), SBP is 155 (40 points), TC is 170 points (6 points) and there is no cardiac fibrillation, the risk of this patient is determined. The total score for this is 306 points. In this case, it can be confirmed that the risk of the patient having a stroke within 5 years is about 8% as shown in FIG. 4. In step 213, if the design completion of the stroke risk prediction model is confirmed, the control unit 150 performs step 215, and if the design completion is not confirmed, the control unit 150 returns to step 205 and performs the above operations again. In step 215, the control unit 150 verifies the designed stroke risk prediction model.

To this end, the control unit 150 divides 500 points, which is the total risk score shown in Fig. 3, as shown in Table 4 below, into a plurality of, for example, 10 sections (D1-D10), and develops cohorts and verifications for each section. Based on the cohort, the predicted stroke incidence rate can be calculated. For example, in the D1 section, the ratio calculated by the development cohort is 0.907, the ratio calculated by the validation cohort is 0.836, the ratio calculated by the development cohort in the D2 section is 1.361, and the ratio calculated by the validation cohort is 1.382, so stroke It can be seen that the risk prediction model is very accurate.

Deciles Incidence rate of stroke(per 1000 person-years) Development cohort Verification cohort D1 0.907 0.836 D2 1.361 1.382 D3 1.699 1.754 D4 2.272 2.148 D5 2.687 2.832 D6 3.238 3.279 D7 3.863 4.086 D8 4.711 4.903 D9 6.223 6.327 D10 10.459 10.126

The embodiments of the present invention disclosed in the present specification and drawings are only provided for specific examples to easily explain the technical content of the present invention and to aid understanding of the present invention, and are not intended to limit the scope of the present invention. Therefore, the scope of the present invention should be construed that all changes or modified forms derived based on the technical idea of the present invention in addition to the embodiments disclosed herein are included in the scope of the present invention.

Claims (9)

Receiving, by the electronic device, a design signal for a stroke risk prediction model;
Collecting, by the electronic device, cohort data according to the design signal;
Establishing a development cohort and a validation cohort for diabetic patients based on the cohort data;
Calculating, by the electronic device, a risk level for each of a plurality of risk factors for stroke; And
Designing the stroke risk prediction model by calculating a contribution level for each risk factor based on the risk level for each risk factor;
A method of designing a model for predicting stroke risk of a diabetic patient, comprising: a. The method of claim 1,
The step of calculating the risk for each of a plurality of risk factors for the stroke,
Age, sex, smoking, exercise level, BMI level, chronic kidney disease, coronary artery disease, duration of diabetes, number of insulin or hypoglycemic drugs used, fasting blood glucose level, systolic blood pressure, total cholesterol and atrial fibrillation A method of designing a model for predicting stroke risk of a diabetic patient, characterized in that the step of calculating a risk for The method of claim 2,
Designing the stroke risk prediction model by calculating the contribution of each risk factor,
Designing the stroke risk prediction model of a diabetic patient, characterized in that the step of designing the stroke risk prediction model by converting contributions to the plurality of risk factors into a score between 0 and 100. The method of claim 3,
Verifying the stroke risk prediction model;
A method of designing a model for predicting stroke risk of a diabetic patient, characterized in that it further comprises. The method of claim 4,
The step of verifying the stroke risk prediction model,
Dividing the converted score for each of the plurality of risk factors into a plurality of sections; And
Verifying the stroke risk prediction model by calculating a stroke incidence prediction rate in each of the divided sections based on the development cohort and the verification cohort, respectively;
A method of designing a model for predicting stroke risk of a diabetic patient, comprising: a. An input unit for inputting a design signal for a stroke risk prediction model;
A communication unit for receiving cohort data according to the design signal; And
Based on the cohort data, a development cohort and a validation cohort for diabetics are established, and the risk of each risk factor and the contribution of each risk factor for stroke are calculated, and the stroke risk prediction model is established. A control unit to design;
A device for designing a model for predicting stroke risk of a diabetic patient, comprising: a. The method of claim 6,
The plurality of risk factors are,
Age, sex, smoking, exercise level, BMI level, chronic kidney disease, coronary artery disease, duration of diabetes, number of insulin or hypoglycemic drugs used, fasting blood glucose level, systolic blood pressure, total cholesterol and atrial fibrillation A device for designing a model for predicting stroke risk of a diabetic patient, characterized in that. The method of claim 7,
The control unit,
The stroke risk prediction model design apparatus of a diabetic patient, characterized in that the stroke risk prediction model is designed by converting contributions to the plurality of risk factors into a score between 0 and 100. The method of claim 8,
The control unit,
The converted score for each of the plurality of risk factors is divided into a plurality of sections, and the stroke occurrence prediction rate in each divided section is calculated based on the development cohort and the verification cohort, respectively, and verification of the stroke risk prediction model is performed. A device for designing a model for predicting stroke risk of a diabetic patient, characterized in that.

Images