KR20090003406A - Stroke risk prediction apparatus and method for the same - Google Patents

Abstract

An apparatus and a method for predicting risk of attack of stroke and a recording medium storing a program for predicting risk of attack of stroke are provided to perform prediction of risk of attack of stroke on Asian based on data obtained from Korean cancer prevention study. An apparatus for predicting risk of attack of stroke includes an input unit(10) for receiving data of a person subjected to examination, such as age, sex, systolic blood pressure, existence of diabetes, total cholesterol, height and weight, a computing unit(20) for computing risk of attack of stroke for ten years from now by using the data inputted to the input unit, and an output unit(30) for outputting the computing result of the computing unit.

Description

Stroke Risk Prediction Apparatus and Method for the same}

1 is a graph showing the predicted values of the risk of stroke and actual occurrence of stroke in Korean men using the Framingham model.

FIG. 2 is a graph showing the predicted values of the risk of stroke and actual occurrence of stroke in Korean women who applied the Framingham model over the next 10 years.

3 is a graph showing the predicted values of the risk of stroke and the actual risk of stroke in the next 10 years of a Korean male applying a corrected Framingham model.

4 is a graph showing the predicted values of the risk of stroke and actual occurrence of stroke in Korean women over the next 10 years using the corrected Framingham model.

5 is a block diagram showing the configuration of the apparatus for predicting stroke risk of stroke according to the present invention;

6 is a flowchart illustrating a method of predicting a stroke attack according to the present invention;

7 is a diagram illustrating an example in which an input screen of the apparatus for predicting stroke risk of stroke according to the present invention is output;

8 is a diagram illustrating an example of outputting the result after receiving male data in the apparatus for predicting stroke risk according to the present invention;

Figure 9 is a comparison of the results of reducing the risk accordingly after adjusting several factors in men,

10 is a view showing an example of outputting the result after receiving the female data in the apparatus for predicting stroke risk according to the present invention,

Figure 11 is a comparison of the result of reducing the risk accordingly after adjusting several factors in women,

12 is a graph showing the predicted results of the actual risk of stroke and stroke in accordance with the present invention for men;

13 is a graph showing the predicted results of the risk of stroke in accordance with the present invention for women and the actual stroke incidence rate,

14 is a graph comparing and comparing the prediction results of the stroke risk according to the present invention for men and women.

The present invention relates to an apparatus and method for diagnosing and predicting the risk of stroke, and more particularly, to an apparatus and method for predicting the risk of stroke that is suitable for Asians, including Koreans.

The stroke is the leading cause of death and disability in Koreans aged 65 and older. In 2004, the total mortality rate among Koreans was 506.97 deaths per 100,000 people. About 13.9% of all deaths were due to complications of stroke and stroke, and medical expenses for stroke treatment accounted for more than 10% of the total medical expenses. to be.

The first predictive model for stroke was the Framingham model, developed by Wolf et al. In 1991, which was developed as a sample from 55 to 84 years old. After adjusting for age, he reported more than 9.6% and 6.5% incidence of male and female for 10 years, respectively. The age range of this study ranged from 30 to 84 years. Comparing the two studies, the incidence of men and women aged 55 to 84 years, adjusted for 10 years, was 9.51% and 8.98%, respectively. Men's risk of stroke was similar to that of the US study for 10 years, but the risk of Korean women was higher than that of the US (6.5%).

The Framingham Heart Study contributed to identifying risk factors for stroke and developed a multivariate function for predicting the risk of stroke.

The Framingham Heart Study is the oldest and probably the most informative study for all prospective studies of heart disease risk.

However, there are limitations in the application of the risk efficacy obtained from this Framingham Study.

First, other ethnic groups differ from those in urban suburbs, such as Caucasian or Framingham, in terms of diet, lifestyle, social environment, and genetic distribution. Recent studies have reported significant discrepancies between the predicted heart disease incidence in Framingham and the actual incidence observed in southern Europe and China.

This study in China did not provide the risk of stroke in Asians with high risk of stroke, and no other study of Asians provided predictions of risk.

Risk factors considered in the development of the paralysis model in the Framingham study were age, systolic blood pressure, antihypertensive therapy, diabetes, smoking, Crdiovascular disease, and ventricular fibrillation. ), Left ventricular hypertrophy.

Among them, especially cardiovascular disease, ventricular fibrillation, left ventricular hypertrophy, etc. are difficult to be easily measured by the general public because they can be measured by ultrasound.

In addition, the Framingham study is based on Westerners, and differs from Asians in many factors such as constitution, physical condition, lifestyle, and genetic differences. Falls.

1 is a graph showing the predicted value of the risk of stroke and actual occurrence of stroke during the next 10 years of Korean men to which the Framingham model is applied. FIG. 2 is a graph showing the case of women.

In the present invention, 13 years of the Korean Cancer Prevention Study (KCPS) was used to predict the risk of stroke in Koreans registered with the National Health Insurance Corporation. 1 and 2, there is a difference between the predicted value and the actual incidence rate.

For a more accurate comparison, the estimates used in the Framingham model were recalibrated by changing the mean and survival rate of each risk factor.

FIG. 3 is a graph showing the predicted values of the risk of stroke and actual occurrence of stroke during the next 10 years of Korean men to which a corrected framingham model is applied, and FIG. 4 is a graph showing the case of women.

Referring to FIG. 3 and FIG. 4, it can be seen that there is a significant difference between the predicted risk and the actual incidence rate when the Framingham model is applied to Koreans even after correction. Prediction / diagnosis is problematic in terms of its accuracy.

The present invention is based on 13 years of Korean Cancer Prevention Study (KCPS) data for Koreans registered with the National Health Insurance Corporation and predicts stroke risk. It is an object of the present invention to provide a risk prediction apparatus and method.

An apparatus for predicting stroke risk according to the present invention for solving the above problems is input unit for receiving test subject data including age, sex, systolic blood pressure, diabetes, total cholesterol, drinking, smoking, exercise, height and weight And, using the respective data input to the input unit includes a calculation unit for calculating the risk of stroke during the next 10 years, and an output unit for outputting the calculation result of the calculation unit.

In addition, the method for predicting the risk of stroke caused by the present invention is a first step of receiving test subject data including age, sex, systolic blood pressure, diabetes, total cholesterol, drinking, smoking, exercise, height and weight, each of the above A second step of calculating a first operation value by multiplying the risk coefficient by adding the risk coefficients after adding the correction to a predetermined correction value, a third step of calculating a second operation value by applying an exponential function to the first operation value, A fourth step of calculating a third operation value, which is a risk of stroke, for the next 10 years of the test subject using the second operation value and the average survival rate, and a risk of stroke attack for the next 10 years of the test subject calculated in the fourth step; And a fifth step of outputting.

In addition, the computer-readable medium in which the risk prediction program of the stroke according to the present invention is recorded may include the age, sex, systolic blood pressure, diabetes, total cholesterol, drinking, smoking, exercise, A first step of receiving data including height and weight, a second step of causing a computer computing device to correct each data to a predetermined correction value, multiplying the risk coefficients, and adding all of them to calculate the first calculation value, A third step of causing the computing device of the computer to apply an exponential function to the first calculated value to calculate a second calculated value; the computing device of the computer uses the second calculated value and the average survival rate for the next 10 years A fourth step of calculating a third operation value, which is a risk of stroke, and a computer output device for stroke for the next 10 years of the test subject calculated in the fourth step The stroke onset risk prediction program is recorded for executing a fifth step which outputs the disease risk.

Hereinafter, exemplary embodiments of the present invention will be described with reference to the accompanying drawings.

In the present invention, a 13-year cohort study (KCPS) was conducted to predict the risk of stroke in Koreans registered with the National Health Insurance Service. Data on blood pressure, smoking, total cholesterol, fasting blood glucose, and other risk factors were collected from 1.3 million Koreans. This study utilizes trace data compiled from records collected and accumulated at national level for Koreans.

The research group used in the present invention is specifically as follows. The study involved 1,329,525 Koreans aged 30 to 95 who were diagnosed at least once every two years by the National Health Insurance (NHIC) between 1992 and 1995. Among the participants, 784,879 (59%) were registered in 1992, 367,903 (27.7%) in 1993, 98,417 (7.4%) in 1994, and 78,335 (5.9%) in 1995.

904 participants who died before 1 January 1993 were excluded to avoid a false link between stroke and pre-existing disease and were treated with cancer, liver disease, cardiovascular disease and respiratory disease. Patients were also excluded. In addition, information such as body mass index (BMI), blood pressure, total cholesterol, aspartate aminotransferase, fasting blood sugar and alcohol consumption were lost, or extremely low body mass index (BMI less than 16kg / m ² ), height (1.30m) And 29,074 older participants (over 85 years) were excluded.

A 10-year predictive study of stroke was determined to be between 30 and 84 years old, with a final population of 1,223,740.

Since data was customarily collected from medical records, no specific written consent was obtained, and the research procedure was approved by the Yonsei Institutional Review Boards.

The National Health Insurance Corporation (NHIC) registrants used for data collection in the present invention received a standard test at a local hospital every two years. Participants' smoking, weight, height, and blood pressure were measured during visits between 1992 and 1995. Fasting blood samples were taken for fasting clinical biochemistry. Data quality control procedures followed the guidelines of the Korean Association of Laboratory Quality Control.

Paralysis risk factors considered in the present invention include age, sex, systolic blood pressure, diabetes, total cholesterol, drinking, smoking, exercise, body mass index. Using data collected from 1992 to 2005, classify participants as smokers (participants who have been smoking for at least one year), non-smokers (non-smokers), and non-smokers (participants who have not smoked), based on their level of smoking. It was. Hypertension was defined as systolic blood pressure above 140 mmHg or diastolic blood pressure above 90 mmHg. Total cholesterol was determined in the appropriate group (blood cholesterol level, <5.17 mmol / L [200 mg / dL]), borderline group (blood cholesterol level, 5.17-6.20 mmol /) according to the National Cholesterol Education Program guidelines. L [200-239 mg / dL] and risk group (serum cholesterol level,> = 6.21 mmol / L [240 mg / dL]) Diabetes was diagnosed using the National Diabetes Data Group diagnostic criteria. Fasting serum glucose levels were defined as those above 6.99 mmol / L (126 mg / dL).

The main incidence variables of stroke attack considered in the present invention are the incidence of subtype and whole stroke caused by cerebral infarction and cerebral hemorrhage. The statistical analysis included only the first event for those who experienced more than one stroke during the follow-up period from January 1993 to December 2005.

Fatal and non-fatal outbreaks were identified from the National Health Insurance Corporation's accident certificate and death certificate, respectively. Automated retrieval of death certificates was performed by National Statistical offices using identification numbers assigned by birth date.

The follow-up of the deceased is therefore approximately 100% and the follow-up is close to 100% for those cases clearly identified by hospital discharge diagnoses. This is because hospitals will not be covered if they do not notify NHIC of their planned discharge date.

Professional recorders separated the causes of death according to criteria defined by the International Classification of Diseases (ICD-10) and used them as causes of death.

In the present invention, the cohort data database is divided into two groups. 50% of the cohort data (random sample 1) was used for model development, and the remaining 50% (random sample 2) was used to test the data.

Cox's Cox proportional hazards regression model was used to model the stroke model using a method similar to that of Framingham researchers.

In the present invention, the 10-year stroke risk (P) is separately calculated for each gender. The basic equation of the risk of stroke for 10 years of the present invention is as shown in Equation 1 below.

P = 1-S (t) ^{exp (f [x, M])} where f [x, M] = β ₁ (x ₁ -M ₁ ) +. β _p (x _p -M _p ).

Where β ₁ ... β _p is the regression coefficient and x ₁ ... x _p Represents an individual risk factor. M ₁ . M _p Is the mean value of the risk factors in the cohort group, and S (t) represents the survival rate when the risk factor is the mean value in t years. Discriminatory power of the model was evaluated by the region of the receiver operating curve (ROC). The test data developed by “male” was used to compare the ROCs of the two models. In order to measure how closely the predicted risk is related to the actual risk, the KCPS participants were divided into the 10-year quartile risk decile predicted by KCPS factors, and the Hosmer-Lemeshow crossover was used to predict and actual risk within each decile. Comparison was made using analysis (X ² test). All statistical analysis was performed using SAS software (ver.8.0) and STATA software (ver.8.2).

Table 1 is a table showing the statistical data on risk factors according to the present invention.

Referring to Table 1, the mean age is 45 years for males and 49.4 years for females. The mean values of systolic blood pressure, fasting blood glucose, and total blood glucose were similar in men and women. But men smoke the majority, while women do not.

Table 2 is a table showing the incidence of stroke among Korean men and women according to the 13-year follow-up survey used in the present invention.

Referring to Table 2, during the 13-year follow-up, 47,233 paralyzed patients (including 35,461 nonfatal and 11,772 fatal) occurred in a cohort of 777,502 males and 446,238 females. Cerebral infarction (ischemic stroke) was 60.6% among men and 58.9% among women, respectively.

Intracerebral hemorrhage was more common than subarachnoid hemorrhage during hemorrhagic stroke. Subarachnoid hemorrhage was excluded from subsequent analysis because the pathogenesis of subarachnoid hemorrhage is different from that of other paralytic subclasses.

Table 3 shows the incidence of 10-year strokes according to the ages of men and women.

Referring to Table 3, the mean incidence of strokes after 10 years was 3.81% for men and 3,53% for women. Outbreaks increased gradually with age, ranging from 0.63% for men aged 30-34 to 11.12% for 80-84 years and 0.25% to 11.45% for women in the same age range. In addition, the average incidence of stroke in 10 years from 55 to 84 years was 9.96% for men and 9.00% for women.

Table 4 is a diagram showing the risk coefficient (Coefficient) and the hazard rate (Hazard Rate) according to each risk factor in the 10-year stroke attack prediction model according to the present invention.

In the present invention, Cox proportional hazards regression model was used.

Referring to Table 4, the β factor of the KCPS Cox model and the mean value of risk factors in the KCPS cohort (Random Sample I) were used in the Korean stroke risk prediction (KSRP) model. As expected, risk factors such as age, systolic blood pressure, diabetes, total cholesterol, smoking, and BMI increased the risk of stroke in Korean men and women. Table 4 shows the risk ratio of stroke during 1 mmHg increase in systolic blood pressure (2.2% for men and 1.6% for women) and 1kg / m ² increase in BMI (3.2% for men and 2.2% for women).

On the basis of this, in the present invention, males predict the risk of stroke after 10 years by the following Equations 2 to 4.

Here, KMT is a median value for calculating the risk of stroke during 10 years, and the median calculated value is calculated by multiplying the risk factors of Table 4 by adding each of the risk factors after correcting each risk factor with a correction value for each factor. to be.

In addition, AGE refers to age and SBP refers to systolic blood pressure.

In addition, the TDM value refers to the presence or absence of diabetes, and if there is diabetes, it is 1, and if there is no diabetes, it is 0.

EXSMOK tells you if you quit smoking. Here, the value of 0 is applied when smoking is stopped.

CUSMOK refers to whether or not you are currently smoking. If you are currently smoking, a value of 1 or 0 is applied. For non-smokers, both EXSMOK and CUSMOK have a value of zero.

In addition, EXER1 indicates whether the user is currently exercising. If one exercise is not performed, a value of 0 is applied.

BMI refers to the body mass index, which is calculated by inputting height and weight, not directly from the subject. Body mass index (BMI) = body weight (kg) / height (m) ² to be.

Next, consider the risk of stroke in accordance with drinking or not. In Equation 2, DRINK1 is applied with a value of 1 if no alcohol is drinked at all, DRINK2 is applied with a value of 1 if it is 1-2 drinks per day, and DRINK4 has a value of 1 when drinking 5-8 glasses. DRINK 5 is authorized for a value of 1 for more than 8 drinks per day.

That is, the values of 1 for DRINK1 to DRINK5 are applied only to the variable corresponding to the main quantity of the value of each variable, and 0 is applied to the remaining variables. If you drink three or four glasses, all DRINK values will be zero.

Next, consider total cholesterol. NTTC2 is 1 when total cholesterol is 200 or more and 239 or less, and NTTC3 is applied when the total cholesterol is 240 or more.

In the present invention, since the effects of drinking are greater in the case of men than in women, the risk of drinking is further divided.

In particular, when calculating the risk of stroke, the effects of drinking and total cholesterol are considered, and all other risk factors can be measured by a simple blood test. Thus, the risk of stroke can be predicted without ultrasound or other complicated tests. The risk coefficients used in Equation 2 are regression coefficients analyzed on the basis of a database of data tracked in the present invention.

Next, Equation 3 below calculates the comparative risk for the average population of Korea using the calculated median calculated value (KMT).

The exponential function is applied to the calculated median value KMT to calculate a comparative risk for the average population of Korea.

Prediction risk of stroke according to the present invention is calculated using Equation 4 using the comparative risk and average survival rate for the average population of Korea.

Here, 0.98485 represents the 10-year average survival rate of men when the risk factor is the mean value.

In the present invention, the risk of stroke occurrence is predicted after 10 years by the following Equations 2 to 4 in women.

Here, the remaining variables except for TDRINK are the same as those of the male. TDRINK is assigned a value of 1 if a woman drinks alcohol, and a value of 0 otherwise.

Since each variable value in Equations 6 and 7 has substantially the same meaning as that of the male, description thereof is omitted.

5 is a block diagram showing the configuration of the apparatus for predicting stroke risk according to the present invention.

An apparatus for predicting stroke risk according to the present invention includes an input unit 10, a calculation unit 20, and an output unit 30.

The input unit 10 receives test subject data including age, sex, systolic blood pressure, diabetes, total cholesterol, drinking, smoking, exercise, height and weight of the test subject.

The input unit 10 may be an input device such as a keyboard or a mouse.

In addition, the input unit 10 may include a blood analysis device that collects blood and the like from the test subject and analyzes the blood, for example, through which the data regarding diabetes and total cholesterol may be input.

In addition, the input unit 10 may include a device for measuring height, weight, blood pressure, etc. directly from the test subject, and data such as height, weight, and blood pressure may be input therefrom.

The calculation unit 20 calculates the stroke risk for the next 10 years by using the respective data input from the input unit 10. The calculating unit 20 may be implemented by a microcomputer or a computer.

The calculation unit 20 calculates the risk of stroke in the next 10 years by processing the input data according to Equations 2 to 4 when the test subject is a male, and when the test subject is a female, the equations 5 to 7 Process the inputted data to calculate the risk of stroke in the next 10 years.

The calculation unit 20 corrects the data to a predetermined correction value, multiplies the risk coefficients, adds all of them, and calculates the first operation value by applying the exponential function to the first operation value. A second calculation unit 22 for calculating a second calculation value and a third calculation unit 23 for calculating a third calculation value, which is a risk of stroke, over the next ten years, using the second calculation value and the average survival rate.

The first operation unit 21 calculates the first operation values KMT and KFT by Equation 2 or Equation 5, and the second operation unit 22 calculates the second operation value by Equation 3 or Equation 6 KMT1 and KFT1 are calculated, and the third calculation unit 23 calculates the third calculation values KMTP and KFTP according to equation (4) or (7).

Finally, the third calculated value is the risk of developing a stroke during the next 10 years.

The output unit 30 outputs the 10-year stroke occurrence risk calculated by the calculating unit 20. The output unit 30 may be implemented as an output device such as a monitor, an LCD panel, and the like.

7 is a diagram illustrating an example in which an input screen of the stroke prevalence risk prediction apparatus according to the present invention is output, and FIG. 8 is a result obtained after receiving male data from the stroke prevalence risk prediction apparatus according to the present invention. 9 is a diagram illustrating an example, and in the case of men, a comparison of the results of reducing the risk accordingly after adjusting several factors.

10 is a diagram showing an example of outputting the result after receiving the female data in the apparatus for predicting stroke risk according to the present invention, Figure 11 is a risk according to the female after adjusting some factors The result of decreasing is a figure shown comparatively.

The screen of the output unit 30 may include a first display unit W1 for displaying a query for inputting the respective data, an average value of overall risk of developing the disease after inputting the data, a risk of developing the next 10 years of the test subject, and an age group to which the test subject belongs. And a second display portion W2 for outputting a graph of an average risk of risk, and a third display portion W3 for displaying the contents of the second display portion in letters.

The first display unit W1 displays a query for receiving data including age, sex, systolic blood pressure, diabetes, total cholesterol, drinking alcohol, smoking, exercise, height and weight. The test subject may input the data to an input device such as a keyboard or a mouse with respect to the matter output on the first display unit.

In addition, the first display unit W1 displays a command button A for inputting all of the data and then transmitting the data to the calculation unit 20 to calculate the risk of stroke in the next 10 years.

In addition, a plurality of command buttons B may be displayed on the first display unit W1 to re-enter some data to adjust a risk factor, perform a new operation, and output a result.

After inputting the information, the second display unit W2 outputs a graph of the average value of the overall risk rate, the risk rate of the test subject, and the average risk rate of the age group to which the test subject belongs.

In the third display unit W3, the risk rate and the risk level of the test subject are displayed in text, and the change in the risk rate, which is a result of the new calculation by the command button B, is displayed in a graph.

The third display portion W3 displays the contents of the second display portion W2 displayed in graphs in letters, and outputs how many times the risk of stroke occurs among the test subjects compared to the average values.

The output unit 30 further includes a fourth display unit W4 which sequentially displays a list of previously input data and a list of newly input data when a new operation is performed by modifying a part of data or inputting new data. Include.

The fourth display unit W4 compares the previously input data with the contents of the input data to calculate the risk of stroke, which is currently being output, so that the change in the risk of stroke can be observed according to the change of the risk factor. Can be.

By modifying data related to risk factors in advance, we can find ways to reduce the risk of stroke.

In this case, the third display unit W3 outputs a graph of the risk of stroke incurred by the previous data of the test subject and the risk of stroke caused by newly inputted data.

6 is a flowchart illustrating a method of predicting a stroke attack according to the present invention.

Referring to FIG. 6, the method for predicting stroke risk according to the present invention first receives test subject data including age, sex, systolic blood pressure, diabetes, total cholesterol, drinking, smoking, exercise, height and weight ( S10)

Next, after correcting the input data to a predetermined correction value, multiplying the risk coefficients and adding them all to calculate a first operation value (S20). At this time, the first operation value is calculated by Equation 2 or Equation 4 above.

Next, a second calculation value is calculated by applying an exponential function to the first calculation value (S30). In this case, the second calculation value is calculated by Equation 3 or Equation 5.

Next, using the second operation value and the average survival rate, a third operation value, which is a risk of stroke, is calculated for the next 10 years of the test subject (S40). At this time, the third operation value is calculated by Equation 4 or Equation 7.

Lastly, a third operation value, which is a risk of stroke, in the next 10 years of the test subject calculated in the fourth step, is output to the output device (S50).

In addition, the method for predicting stroke risk according to the present invention may further include a sixth step of receiving the data of the test subject again and repeating the second to fifth steps. That is, some data is input again to calculate the modified result.

The method may further include a seventh step of comparing the calculated risk of stroke incurred after the re-entry with the calculated risk of stroke incurred before the re-entry. By comparing the results of the previous data with the results of the newly input data, the test subjects can see the effect of each input data (risk factor) on the risk of stroke.

The present invention can be embodied as a recording medium which is produced as a computer program and stored therein, so that the method for predicting stroke risk according to the present invention can be performed on a computer.

That is, the computer-readable medium that records the stroke predictive risk prediction program according to the present invention includes a computer's input device such as age, sex, systolic blood pressure, diabetes, total cholesterol, drinking, smoking, exercise, height, etc. And a first step of receiving data including body weight, a second step of causing the computing device of the computer to correct the respective data to a predetermined correction value, multiplying the risk coefficients, and adding them all to calculate the first calculation value. A third step of causing a computing device to apply an exponential function to the first calculated value to calculate a second calculated value, wherein the computing device of the computer uses the second calculated value and the average survival rate to stroke the test subject for the next 10 years. The fourth step of calculating the third operation value, which is the risk of onset, and the output device of the computer causes the stroke to occur during the next ten years of the inspection box calculated in the fourth step. The program including a fifth step of outputting the risk is recorded.

In the program, the first operation value is calculated by Equation 2 or Equation 5, the second operation value is calculated by Equation 3 or Equation 6, and the third operation value is Equation 4 Or it is calculated by the computer computing unit by the equation (7).

Further, in the computer-readable medium in which the stroke prevalence risk prediction program according to the present invention is recorded, the computer input device may further perform a sixth step of repeating the second to fifth steps by receiving the data of the subject again. Can be stored

In addition, the computer-readable medium that records the stroke prevalence risk prediction program according to the present invention includes a seventh program that causes the computer output device to compare and output the stroke impact risk before re-entry with the stroke impact risk after re-entry. A program may be stored to further execute the step.

12 is a graph showing the predicted results and actual stroke incidence of stroke in accordance with the present invention for men, Figure 13 is a graph showing the prediction results and actual stroke incidence of stroke in accordance with the present invention for women to be. In addition, Figure 14 is a graph showing a comparison of the prediction results of the risk of stroke in accordance with the present invention for men and women.

With reference to Figures 12 to 14 will review the prediction results of the risk of stroke in accordance with the present invention.

12 and 13 contain a performance plot for men and women in the KCPS cohort (Random Sample II). In general, the actual rates of stroke in the KCPS cohort were similar to those predicted by the KSRP model for men and women (p-values from the cross-sectional analysis were 0.5641 for men and 0.1623 for women).

In addition, the risk ratio for actual stroke was gradually increased in the KCPS cohort (Random Sample II) as the predicted risk of stroke was increased according to the present invention.

The relationship between the predicted stroke risk factor and the actual stroke occurrence is 63.17 for men with the highest and lowest deciles, and the 95% confidence interval is 52.30–76.31. Similarly, for women, the risk ratio is 120.34 and the 95% confidence interval is between 85.31 and 169.77.

In discriminant analysis, the Receiver Operating Curve (ROC), which was sufficient to distinguish between stroke and non-group, was 0.8165 (95% confidence interval, 0.7993-0.8337) in men and 0.8095 (95% confidence interval in women). , 0.7875-0.8315).

More than 45,000 strokes have occurred in this cohort of more than 1 million people over 13 years in the database data used in the present invention. Apparatus, method and program for predicting stroke risk according to the present invention have been developed based on this data. Compared to actual stroke cases, the present invention applied to Korean cohort showed good prediction.

In the present invention, eight major risk factors are included in the risk prediction.

Variables in the Framingham study were age, systolic blood pressure, hypertension treatment, diabetes, smoking, heart disease, fibrillation of the arteries, and left ventricular hypertrophy. Variables of age, systolic blood pressure, diabetes, and smoking were commonly used in this study and in the Framingham study.

Hypertension is a major risk factor for stroke and stroke incidence is proportional to blood pressure in the present invention. In the present invention, whenever the 10 mmHg systolic blood pressure increases, the risk of stroke increases by 22% in men and 15% in women.

The present invention is characterized by including other variables such as total cholesterol, physical activity, physical activity, BMI, drinking amount. Although total cholesterol and alcohol consumption had a low impact on stroke risk, this study found that these two variables significantly increased stroke risk, especially in men. Many studies have been conducted on the link between total cholesterol and stroke. These studies show that increased total cholesterol is associated with or without stroke

The present invention showed a prediction rate of 82% for males and 81% for females compared to actual stroke cases. Although the data of the database used in the present invention is for Koreans, it can be applied to Asians including Koreans.

As described above, the apparatus and method for predicting stroke risk according to the present invention have been described with reference to the illustrated drawings. However, the present invention is not limited by the embodiments and drawings disclosed herein, and it is applied within the scope of the technical idea. Can be.

The recording medium storing the apparatus and method and program for predicting stroke risk according to the present invention has developed a system for predicting stroke risk using stroke data of 1.3 million Koreans to accurately predict the risk of stroke attack among Asians including Koreans with high accuracy. It has a predictable effect.

In addition, by modifying and inputting the test subject's data and comprehensively comparing the results of the test results, it is possible to easily recognize the effect of each risk factor on the risk of stroke, and to change the input data of the test subject. There is an advantage that can be easily recognized at a glance the change in risk of stroke.

Claims (30)

An input unit for receiving test subject data including age, sex, systolic blood pressure, diabetes, total cholesterol, drinking, smoking, exercise, height and weight; A calculating unit calculating a risk of stroke during the next 10 years by using the data inputted to the input unit; And An apparatus for predicting stroke risk including an output unit for outputting a calculation result of the operation unit. The method according to claim 1, The calculation unit comprises a first calculation unit for calculating the first operation value by multiplying the risk coefficient after correcting each data to a predetermined correction value and adding them all together; A second operation unit calculating a second operation value by applying an exponential function to the first operation value; And And a third operation unit calculating a third operation value, which is a risk of stroke, over the next 10 years by using the second operation value and the average survival rate. The method according to claim 2, And calculating the first calculation value KMT according to the following equation when the first operator is a male.

(Where AGE is age, SBP is systolic blood pressure, TDM is 0 if diabetic 1, EXSMOK is 1 if you quit smoking, 0 is 0, CUSMOK is 1 or 0 if you don't smoke, 0 if EXER1 is not exercise, 0 is BMI Body mass index = body weight (kg) / height (m) ² , DRINK1 is not drinking at all 1, DRINK2 is 1 to 2 cups per day shochu, DRINK4 is 5-8 cups, DRINK 5 is 8 cups or more 1 NTTC2 is 200 or more than 239 for total cholesterol 1, NTTC3 is more than 240 for total cholesterol 1) The method according to claim 3, And calculating the second calculation value KMT1 according to the following formula when the second operator is a male.

The method according to claim 4, When the third operation unit test subject is a male calculates the third operation value (KMTP) by the following formula characterized in that the stroke risk risk prediction device.

The method according to claim 2, And calculating the first computational value (KFT) according to the following formula when the first operator inspects the female.

(Where AGE is age, SBP is systolic blood pressure, TDM is 0 if diabetic is 1, EXSMOK is 1 if you quit smoking, 0 is 0, CUSMOK is 1 or 0 if you don't smoke, 0 if EXER1 is not exercised, 0 is BMI Body mass index = body weight (kg) / height (m) ² , TDRINK is 1 or 0 when drinking alcohol, NTTC2 is 1 when total cholesterol is 200 or more and 239 or less, 1 is when NTTC3 is 240 or more. The method according to claim 6, And calculating the second calculation value KFT1 according to the following equation when the second operator is a female.

The method according to claim 7, And calculating the third calculation value (KFTP) according to the following formula when the third operator inspects the female.

The method according to claim 1, The output unit may include a first display unit for displaying a query for receiving the respective data; A second display unit for outputting a graph showing an average of the overall risk of risk, a risk of developing the next 10 years of the test subject, and an average of the risk level of the age group belonging to the test subject after inputting each data; And And an output screen including a third display unit in which the contents of the second display unit are displayed in letters. The method according to claim 9, And the third display unit further displays a ratio of the onset risk of the test subject to the onset risk average value of the age group to which the test subject belongs. The method according to claim 9, The third display unit, if the re-entry by modifying a part of each of the data of the subject to be re-input before the re-entry risk data, each occurrence risk according to the re-input data, characterized in that the apparatus for predicting stroke attack. The method according to claim 11, The output unit further includes a fourth display unit for outputting a list of data before and after the correction sequentially, when a part of each data of the test subject is corrected, characterized in that the output of the stroke attack. A first step of receiving test subject data including age, sex, systolic blood pressure, diabetes, total cholesterol, drinking, smoking, exercise, height and weight; A second step of correcting the respective data with a predetermined correction value and multiplying the risk coefficients to add all of them to calculate a first operation value; A third step of calculating a second operation value by applying an exponential function to the first operation value; A fourth step of calculating a third operation value, which is a risk of windbreak, among the next 10 years of the test subject, by using the second operation value and the average survival rate; And And a fifth step of outputting a risk of stroke in the next 10 years of the test subjects calculated in the fourth step. The method according to claim 13, In the second step, when the test subject is a male, the first operation value KMT is calculated by the following equation.

(Where AGE is age, SBP is systolic blood pressure, TDM is 0 if diabetic 1, EXSMOK is 1 if you quit smoking, 0 is 0, CUSMOK is 1 or 0 if you don't smoke, 0 if EXER1 is not exercise, 0 is BMI Body mass index = weight / height ² , DRINK1 is 1 if no alcohol is consumed at all, DRINK2 is 1 or 2 drinks per day based on soju, 1 to 5 to 8 cups of DRINK4, 8 or more to DRINK 5, 1, NTTC2 is total cholesterol 200 or more 239 or less 1, NTTC3 is total cholesterol 240 or more 1) The method according to claim 14, In the third step, when the test subject is a male, the second operation value KMT1 is calculated by the following formula, wherein the risk of stroke is predicted.

The method according to claim 15, In the fourth step, when the test subject is a male, the third operation value KMTP is calculated by the following equation, wherein the stroke prevalence risk predicting device is characterized in that the following.

The method according to claim 13, In the second step, when the test subject is a female, the first operation value KFT is calculated by the following formula, wherein the risk of stroke attack is estimated.

(Where AGE is age, SBP is systolic blood pressure, TDM is 0 if diabetic 1, EXSMOK is 1 if you quit smoking, 0 is 0, CUSMOK is 1 or 0 if you don't smoke, 0 if EXER1 is not exercise, 0 is BMI Body mass index = weight / height ² , TDRINK is 1 if drinking alcohol or 0, NTTC2 is total cholesterol 200 or more and 239 or less 1, NTTC3 is total cholesterol 240 or more 1) The method according to claim 17, In the third step, when the test subject is a female, the second operation value KFT1 is calculated by the following formula, wherein the risk of stroke attack is estimated.

The method according to claim 18, In the fourth step, when the test subject is a female, the third operation value (KFTP) is calculated by the following equation.

The method according to claim 13, And a sixth step of receiving the data of the test subject again and repeating the second to fifth steps. The method of claim 20, And a seventh step of comparing the risk of stroke occurrence calculated after the re-entry with the risk of stroke occurrence calculated before re-entry. A first step of receiving data including an age, a sex, systolic blood pressure, diabetes, total cholesterol, drinking, smoking, exercise, height, and weight of a test subject through a computer input device; A second step of causing the computing device of the computer to correct the respective data to a predetermined correction value, multiplying the risk coefficients, and adding all of them to calculate the first calculated value; A third step of causing a computing device of a computer to apply an exponential function to the first calculated value to calculate a second calculated value; A fourth step of causing a computing device of the computer to calculate a third calculated value of the risk of stroke during the next ten years of the test subject by using the second calculated value and the average survival rate; And A computer-readable medium having recorded a stroke prevalence risk prediction program for executing a fifth step of causing the output device of the computer to output the risk of stroke attack during the next 10 years of the test subject calculated in the fourth step. The method according to claim 22, If the test subject is a male, a computer-readable medium recording a stroke prevalence risk prediction program for performing the second step by the following formula.

(Where KMT is the first computed value, AGE is age, SBP is systolic blood pressure, TDM is 0 if diabetic is 1, EXSMOK is 1 if no smoking, 0 is 0, CUSMOK is currently 1 or 0 if smoking, and EXER1 is not exercised 1 If exercise, 0, BMI = body mass index = ² weight / height ² , DRINK1 is not drinking at all 1, DRINK2 is 1 ~ 2 cups per day shochu, DRINK4 5 ~ 8 cups, DRINK 5 8 cups or more 1, NTTC2 is more than 200 and less than 239 total cholesterol 1, NTTC3 is more than 240 total cholesterol 1) The method according to claim 23, And a computer-readable medium recording a stroke prevalence risk prediction program for causing the third step to be performed by the following equation when the test subject is a male.

(Where KMT1 is a second operation value) The method of claim 24, The computer-readable medium having recorded the stroke risk prediction program for causing the fourth step to be performed by the following formula when the test subject is a male.

Where KMTP is the third computed value The method according to claim 22, And a computer-readable medium recording a stroke prevalence risk prediction program for causing the second step to be performed by the following formula when the test subject is a female.

(Where KFT is the first computed value, AGE is age, SBP is systolic blood pressure, TDM is 0 if diabetic is 1, EXSMOK is 1 if no smoking, 0 is 0, CUSMOK is currently 1 or 0 if smoking, and EXER1 is not exercised. 1 for exercise, 0 for BMI, body mass index = weight / height ² , TDRINK for drinking 1 or 0, NTTC2 for total cholesterol above 200 and 239 1, NTTC3 for total cholesterol above 240 1) The method of claim 26, And a computer-readable medium recording a stroke prevalence risk prediction program for causing the third step to be performed according to the following formula when the test subject is a female.

(Where KFT1 is the second operation value) The method of claim 27, And a computer-readable medium having recorded a risk assessment program for stroke attack, which causes the fourth step to be performed by the following formula when the test subject is a female.

Where KFTP is the third computed value The method according to claim 22, And a computer-readable apparatus for recording a stroke prevalence risk prediction program, wherein the computer input device further inputs the test subject's data and executes the sixth step of repeating the second to fifth steps. The method of claim 29, And a seventh step of causing the computer output device to perform the seventh step of comparing the risk of stroke inflation before re-input with the risk of stroke inflation after re-entry.

Images