Prediction system and program

Abstract

To provide a prediction system and a program which can visualize which type of lifestyle-related illness a user is going to be prone to.SOLUTION: A prediction system 1 has a calculation unit 12 for calculating the probability P that a subject develops a lifestyle-related disease based on a prediction formula PR obtained through logistic regression analysis. The explanatory variables used in the analysis are a first variable EV1, for which a value N1 is obtained from a blood test, and a second variable EV2, for which a value N2 is obtained from a non-blood test. The objective variable is a disease onset probability of lifestyle-related illness after five years. The probability P of the subject is calculated on the basis of the values N1 and N2 of the first variable EV1 and the second variable EV2.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a prediction system and a program. [Background technology]

[0002] Patent Document 1 discloses an automatic diagnosis system, an automatic diagnosis program, and an automatic diagnosis method that are capable of automatically diagnosing a disease name based on the results of a health check. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2023-103760 A Summary of the Invention [Problem to be solved by the invention]

[0004] The attendance rate for specific health checkups, which began in 2008, is 53.4% ​​(2020), and the implementation rate of specific health guidance is 22.7%, which is by no means sufficient. Improving the attendance rate for specific health checkups, the implementation rate of specific health guidance, and the implementation rate of medical institutions is extremely important in preventing lifestyle-related diseases, and is important in terms of significantly reducing medical expenses and achieving healthy longevity. Due to the aging population, the proportion of medical expenses for cerebrovascular and cardiovascular diseases caused by lifestyle-related diseases is increasing, and the Basic Act on Measures against Cardiovascular Disease was enacted in 2018, further emphasizing the importance of prevention.

[0005] One of the reasons why the rates of specific health checkups, specific health guidance, and medical institution visits are not increasing is that the test results of specific health checkups are technical and difficult to understand, making it difficult to understand what lifestyle-related diseases people will be susceptible to in the future. Therefore, there is a need for a system that can "visualize" what lifestyle-related diseases people will be susceptible to in the future.

[0006] The automatic diagnosis system, automatic diagnosis program, and automatic diagnosis method disclosed in the above Patent Document 1 can perform automatic diagnosis of disease names based on the results of health checkups, but do not predict which lifestyle diseases an individual is likely to develop in the future.

[0007] The present invention has been made under the above circumstances, and an object thereof is to provide a prediction system and program that can "visualize" which lifestyle diseases an individual is likely to develop in the future.

Means for Solving the Problems

[0008] To achieve the above object, a prediction system according to a first aspect of the present invention is a variable that shows a significant difference in the presence or absence of the onset of lifestyle diseases after a predetermined number of years. Using the first variable whose numerical value is obtained from a blood test and the second variable whose numerical value is obtained from other than the blood test as explanatory variables, and performing logistic regression analysis with the onset probability of lifestyle diseases after a predetermined number of years as the objective variable, a calculation unit that calculates the onset probability of the subject based on the numerical values of the first variable and the second variable of the subject is provided.

[0009] When the lifestyle disease is hypertension, the first variable includes the presence or absence of diabetes and the presence or absence of hyperuricemia, and the second variable includes age, BMI, systolic blood pressure, diastolic blood pressure, the presence or absence of smoking, and the presence or absence of a family member with hypertension. This may be the case.

[0010] When the lifestyle disease is chronic kidney disease, the first variable includes the presence or absence of diabetes, the presence or absence of hyperuricemia, the presence or absence of dyslipidemia, and eGFR, and the second variable includes age, BMI, and the presence or absence of hypertension. This may be the case.

[0011] When the lifestyle disease is metabolic syndrome, The first variable includes fasting blood glucose level, HDL cholesterol, LDL cholesterol, and triglycerides. The second variable includes age, gender, BMI, systolic blood pressure, diastolic blood pressure, presence or absence of smoking, and presence or absence of drinking. This may be the case.

[0012] When the lifestyle disease is arteriosclerosis, in the first variable, when the subject is male, blood glucose level and triglycerides are included, while when the subject is female, blood glucose level is included. In the second variable, when the subject is male, age, BMI, systolic blood pressure, and diastolic blood pressure are included, while when the subject is female, age, systolic blood pressure, diastolic blood pressure, and presence or absence of smoking are included. This may be the case.

[0013] When the lifestyle disease is diabetes, the first variable includes fasting blood glucose level and HbA1c. The second variable includes age, gender, and BMI. This may be the case.

[0014] When the lifestyle disease is dyslipidemia, the first variable includes HDL cholesterol, LDL cholesterol, and triglycerides. The second variable includes age and presence or absence of smoking. This may be the case.

[0015] The prediction system according to the second aspect of the present invention comprises a calculation unit that calculates the onset probability of metabolic syndrome after a predetermined number of years for a subject based on the numerical values of the explanatory variables of the subject, using a prediction formula obtained by performing a logistic regression analysis with age, gender, BMI, presence or absence of hypertension, presence or absence of smoking, and presence or absence of drinking as explanatory variables and the onset probability of metabolic syndrome after a predetermined number of years as the objective variable.

[0016] The prediction system according to the third aspect of the present invention uses, as explanatory variables, age, gender, BMI, presence or absence of hypertension, presence or absence of family members with hypertension, and presence or absence of dyslipidemia, and uses a prediction formula obtained from a logistic regression analysis performed with the probability of developing diabetes after a predetermined number of years as the objective variable, and calculates the probability of developing diabetes in the subject after a predetermined number of years based on the numerical values of the explanatory variables of the subject. It is provided with a calculation unit.

[0017] It is provided with an image acquisition unit that acquires an image of an inspection result report, and an extraction unit that extracts the numerical values of the explanatory variables from the image. The calculation unit calculates the onset probability based on the numerical values of the explanatory variables extracted by the extraction unit. This may be the case.

[0018] The program according to the fourth aspect of the present invention causes a computer to function as a calculation unit that uses, as explanatory variables, a first variable for which numerical values are obtained from a blood test and a second variable for which numerical values are obtained from other than the blood test, which are variables having a significant difference in the presence or absence of onset of lifestyle-related diseases after a predetermined number of years, and uses a prediction formula obtained by performing a logistic regression analysis with the onset probability of lifestyle-related diseases after a predetermined number of years as the objective variable, and calculates the onset probability of the subject based on the numerical values of the first variable and the second variable of the subject. to function as.

Effects of the Invention

[0019] According to the present invention, it is possible to "visualize" what lifestyle-related diseases a person is likely to develop in the future.

Brief Description of the Drawings

[0020]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

[0021] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In each drawing, the same or equivalent parts are denoted by the same reference numerals. In the following embodiments, expressions such as "having", "including", or "containing" also include the meaning of "consisting of" or "composed of".

[0022] Embodiment 1 First, Embodiment 1 of the present invention will be described. As shown in FIG. 1, the prediction system 1 according to the present embodiment includes a terminal device 10. The terminal device 10 predicts the onset probability P of lifestyle-related diseases after a predetermined number of years, for example, after 5 years. In the present embodiment, as lifestyle-related diseases, for example, as shown in FIG. 2, hypertension, chronic kidney disease, metabolic syndrome, arteriosclerosis, diabetes, and dyslipidemia are targeted. However, the lifestyle-related diseases to be predicted are not limited to these.

[0023] [Criteria for Judging Lifestyle-Related Diseases] The criterion for determining whether hypertension has developed is that the systolic blood pressure (maximum blood pressure) is 140 mmHg or higher, or the diastolic blood pressure (minimum blood pressure) is 90 mmHg or higher.

[0024] In addition, the criteria for determining whether chronic kidney disease has developed are that either or both of the following (1) and (2) persist for three months or more. (1) The presence of kidney damage is revealed by urine tests, imaging diagnoses, blood tests, pathology, etc., especially proteinuria of 0.15 g / gCr or more (albuminuria of 30 mg / gCr or more). (2) The glomerular filtration rate (GFR) is less than 60 mL / min / 1.73m 2 ².

[0025] In addition, the criteria for determining whether metabolic syndrome has developed are that in Japan, the waist circumference (abdomen circumference at the height of the navel) is 85 cm or more for men and 90 cm or more for women, and two or more of blood pressure, blood sugar, and lipids deviate from the reference values.

[0026] In addition, the criteria for determining whether arteriosclerosis has developed are that the reference value of the arteriosclerosis index is less than 3.0.

[0027] In addition, the criteria for determining whether diabetes has developed are that the test result of blood glucose level is 126 mg / dL or more on an empty stomach (before breakfast), 200 mg / dL or more two hours after a meal, or the test result of hemoglobin A1c (HbA1c) is 6.5% or more, or diabetes drug treatment is being carried out. Note that the numerical value of HbA1c is strictly HbA1c (NGSP).

[0028] In addition, the criteria for determining whether dyslipidemia has developed are LDL (Low Density Lipoprotein) cholesterol: 140 mg / dL or more, or triglyceride: 150 mg / dL or more on an empty stomach, 175 mg / dL or more in the non-fasting state, or Non-HDL (High Density Lipoprotein) cholesterol (total cholesterol - HDL cholesterol): 170 mg / dL or more, HDL cholesterol less than 40 mg / dL, or lipid drug treatment is being carried out.

[0029] Even if the current specific medical examination does not meet the above-mentioned criteria, this prediction system 1 predicts the probability of meeting the above-mentioned criteria, that is, the onset probability P of lifestyle-related diseases, five years after the specific medical examination based on the examination results of the specific medical examination and the like.

[0030] [Database] The prediction system 1 includes a database 2. As shown in FIG. 1, the database 2 stores learning data BD for generating a prediction formula PR for the onset probability P of lifestyle-related diseases five years later. This learning data BD includes, among the variables representing the risk factors related to lifestyle-related diseases, the numerical value of the first variable EV1 obtained from the blood test of the examinee in the specific medical examination, the numerical value of the second variable EV2 related to the examinee obtained from sources other than the blood test, and whether or not the examinee has developed a lifestyle-related disease five years after the specific medical examination was conducted, that is, the presence or absence (result) R of lifestyle-related diseases five years later. The more the amount of this learning data BD, the better.

[0031] As the first variable EV1, any item can be included as long as it can be directly obtained from the blood test of the specific medical examination or its numerical value can be calculated based on the calculation or determination result. For example, as shown in FIG. 2, the first variable EV1 includes the presence or absence of diabetes, HbA1c (mmol / mol), the presence or absence of hyperuricemia, the presence or absence of dyslipidemia, eGFR (mL / min / 1.73m 2 )), fasting blood glucose level (mg / dl), HDL cholesterol (mg / dl), LDL cholesterol (mg / dl), and triglyceride (mg / dl). Here, the presence or absence of diabetes is a variable calculated from the numerical value of the blood test, meets the criteria for diabetes determination, has a value of 1 if it is diabetes, and a value of 0 if it is not diabetes. The same applies to the presence or absence of hyperuricemia and the presence or absence of dyslipidemia. The first variable EV1 is not limited to those shown in FIG. 2.

[0032] As the second variable EV2, any information regarding the examinee obtained from sources other than the blood test of the specific medical examination can be included. For example, as shown in FIG. 2, the second variable includes age, gender, BMI (Body Mass Index) (kg / m 2 ), presence or absence of hypertension, systolic blood pressure (mmHg), diastolic blood pressure (mmHg), presence or absence of a family with hypertension (hypertensive family history), presence or absence of smoking, and presence or absence of drinking. The age can be used as the value of the variable as it is. For gender, male is set to 1 and female is set to 0. For the presence or absence of hypertension, presence is 1 and absence is 0. Similarly, for smoking, presence is 1 and absence is 0, and for drinking, in daily life, it is 1 when the alcohol intake per unit number of days exceeds a predetermined amount, and 0 otherwise. The second variable EV2 is not limited to those shown in FIG. 2.

[0033] As shown in FIG. 1, the prediction system 1 includes an analyzer 3. The analyzer 3 uses the first variable EV1 obtained from the blood test of the subject and the second variable EV2 obtained from sources other than the blood test as explanatory variables, and performs logistic regression analysis with the onset probability of the lifestyle disease of the subject after 5 years as the target variable to generate a prediction formula PR. The prediction formula PR is defined as follows. P = 100 / (1 + e -(a0+a1(X1)+a2(X2)+a3(X3)···) ) ··· (1) Here, P [%] is the onset probability of the lifestyle disease. a0 is the intercept, X1, X2, X3,... are explanatory variables, and a1, a2, a3,... are regression variables. Generating the prediction formula PR corresponds to determining the intercept a0, regression variables a1, a2, a3,... by logistic regression analysis. Note that the intercept a0, regression variables a1, a2, a3,... are generated for each lifestyle disease and for each of the various conditions described below. The various conditions include the presence or absence of blood sampling or gender, and the analyzer 3 generates a prediction formula PR for each lifestyle disease, and also generates a prediction formula PR for each of the presence and absence of blood sampling or for each gender, and generates the prediction formula PR with the highest AUC (Area Under Curve) as the prediction formula PR for the lifestyle disease under that condition.

[0034] Note that among the first variable EV1 and the second variable EV2, there are some that do not have a significant difference depending on the lifestyle disease. The analyzer 3 performs statistical processing such as a t-test on the first variable EV1 and the second variable EV2 for each type of lifestyle disease to determine the presence or absence of a correlation between the lifestyle disease and the variable, and selects the first variable EV1 and the second variable EV2 that have a significant difference, that is, are correlated with the onset of the lifestyle disease after 5 years have passed.

[0035] The selection results are shown, for example, in FIG. 2. As shown in FIG. 2, for example, the variables (risk factors) that have a significant difference between those who developed hypertension after 5 years and those who did not are the age of the examinee at the specific health check, BMI, systolic blood pressure, diastolic blood pressure, family history of hypertension, presence or absence of smoking, presence or absence of diabetes, and presence or absence of hyperuricemia. Therefore, when the lifestyle disease is hypertension, the first variable EV1 includes the presence or absence of diabetes and the presence or absence of hyperuricemia, and the second variable EV2 includes age, BMI, systolic blood pressure, diastolic blood pressure, family history of hypertension, presence or absence of smoking, and presence or absence of alcohol consumption.

[0036] Also, the risk factors that have a significant difference between those who developed chronic kidney disease after 5 years and those who did not are the age, gender, presence or absence of hypertension, presence or absence of diabetes, presence or absence of hyperuricemia, presence or absence of dyslipidemia, and eGFR at the specific health check. Therefore, when the lifestyle disease is chronic kidney disease, the first variable EV1 includes the presence or absence of diabetes, the presence or absence of hyperuricemia, the presence or absence of dyslipidemia, and eGFR, and the second variable EV2 includes age, gender, and the presence or absence of hypertension.

[0037] Also, when blood sampling test results are available, the variables that have a significant difference between those who developed metabolic syndrome after 5 years and those who did not are age, gender, BMI, systolic blood pressure, diastolic blood pressure, fasting blood glucose level, HDL cholesterol, LDL cholesterol, and triglycerides. Therefore, when the lifestyle disease is metabolic syndrome and blood sampling test results are available, the first variable EV1 includes the fasting blood glucose level, HDL cholesterol, LDL cholesterol, and triglycerides, and the second variable EV2 includes age, BMI, systolic blood pressure, and diastolic blood pressure.

[0038] In addition, when the results of blood tests cannot be obtained, the first variable EV1 and the second variable EV2 that show significant differences between those who developed metabolic syndrome after 5 years and those who did not are age, gender, BMI, systolic blood pressure, diastolic blood pressure, presence or absence of smoking, and presence or absence of drinking at the time of health check. Therefore, when the lifestyle disease is metabolic syndrome and the results of blood tests are obtained, the first variable EV1 includes age, gender, BMI, systolic blood pressure, diastolic blood pressure, presence or absence of smoking, and presence or absence of drinking.

[0039] In addition, when the results of blood tests cannot be obtained, the first variable EV1 and the second variable EV2 that show significant differences between those who developed arteriosclerosis after 5 years and those who did not are age, BMI, systolic blood pressure, diastolic blood pressure, fasting blood glucose level, and triglycerides for men, and age, systolic blood pressure, diastolic blood pressure, smoking, and blood glucose level for women at the time of health check. Therefore, when the lifestyle disease is arteriosclerosis, for the first variable EV1, fasting blood glucose level and triglycerides are included if the examinee is male, while blood glucose level is included if the examinee is female. For the second variable EV2, age, BMI, systolic blood pressure, and diastolic blood pressure are included if the examinee is male, while age, systolic blood pressure, diastolic blood pressure, and presence or absence of smoking are included if the examinee is female.

[0040] In addition, when the results of blood tests are obtained, the first variable EV1 and the second variable EV2 that show significant differences between those who developed diabetes after 5 years and those who did not are age, gender, BMI, HbA1c, and fasting blood glucose level. When the results of blood tests cannot be obtained, they are age, gender, BMI, presence or absence of hypertension, and presence or absence of dyslipidemia. Therefore, when the lifestyle disease is diabetes and the results of blood tests are obtained, the first variable EV1 includes fasting blood glucose level and HbA1c, and the second variable EV2 includes age, gender, and BMI. On the other hand, when the results of blood tests cannot be obtained, the second variable EV2 includes age, gender, BMI, and presence or absence of hypertension.

[0041] In addition, the first variable EV1, which shows a significant difference between those who developed dyslipidemia after 5 years and those who did not, is age and smoking status, and the second variable EV2 is HDL cholesterol, LDL cholesterol, and triglycerides. Therefore, when the lifestyle disease is dyslipidemia, the first variable EV1 includes HDL cholesterol, LDL cholesterol, and triglycerides, and the second variable EV2 includes the age of the subject and smoking status.

[0042] The analyzer 3 performs logistic regression analysis using, as explanatory variables, the first variable EV1 and the second variable EV2 to which ○ is assigned for each lifestyle disease in the table of FIG. 2, and obtains the intercept a0, regression variables a1, a2, a3, ··· in Equation (1) to generate a prediction formula PR. In the analyzer 3, the prediction formulas PR generated for each lifestyle disease and each type of condition are as follows, for example. Note that the explanatory variables X1, X2, X3, ··· differ in content depending on the prediction formula PR. For clarity, the variable "presence or absence of ~" is expressed as "~".

[0043] (1) Prediction formula for the possibility of developing hypertension after 5 years (%) (with blood sampling) P = 100 / (1 + exp(-(-16.3122 + 0.0364 × age + 0.0592 × BMI + 0.0667 × systolic blood pressure + 0.0435 × diastolic blood pressure + 0.2186 × diabetes + 0.1592 × hyperuricemia + 0.3350 × smoking + 0.3675 × family history of hypertension)))) …(1-1)

[0044] (2) Prediction formula for the possibility of developing chronic kidney disease after 5 years (%) (with blood sampling) P = 100 / (1 + exp(-(9.4876 + 0.0311 × age + 0.2400 × gender + 0.3470 × hypertension + 0.0893 × dyslipidemia + 0.3444 × diabetes + 0.0832 × hyperuricemia + (-0.1980) × eGFR))) …(2-1)

[0045] (3) Prediction formula for the possibility of developing metabolic syndrome after 5 years (%) With blood sampling P = 100 / (1 + exp(-(14.7123 + 0.0111 × age + 0.014 × fasting blood glucose + 0.3135 × BMI + 0.0167 × diastolic blood pressure + (-0.0152) × HDL cholesterol + 0.0040 × LDL cholesterol + 0.0083 × systolic blood pressure + 0.0015 × triglyceride + 0.2479 × alcohol consumption + 1.1735 × gender + 0.2049 × smoking))) …(3-1) No blood sampling P = 1 / (1 + exp(-(-13.2623 + 0.0119 × age + 0.2847 × BMI + 0.0137 × diastolic blood pressure + 0.0088 × systolic blood pressure + 0.0707 × alcohol consumption + 1.2056 × gender + 0.4464 × smoking))) …(3-2)

[0046] (4) Prediction formula for the possibility of arteriosclerosis onset after 5 years (%) With blood sampling (for men) P = 100 / (1 + exp(-(-15.91172 + 0.06697 × age - 0.03948 × BMI + 0.02527 × systolic blood pressure + 0.02231 × diastolic blood pressure + 1.59574 × log(blood glucose value) + 0.21281 × log(triglyceride)))) …(4-1) With blood sampling (for women) P = 100 / (1 + exp(-(-16.34973 + 0.09834 × age + 0.02888 × systolic blood pressure + 0.01876 × diastolic blood pressure + 1.31674 × log(blood glucose value) + 0.49505 × smoking))) …(4-2)

[0047] (5) Prediction formula for the possibility of diabetes onset after 5 years (%) With blood sampling P = 100 / (1 + exp(-(-30.3726 + 2.5471 × HbA1c + 0.1101 × fasting blood glucose value + 0.0905 × BMI - 0.0184 × age + 0.2876 × gender))) …(5-1) No blood sampling P = 100 / (1 + exp(-(-8.4968 + 0.1522 × BMI + 0.0200 × age + 0.5096 × male + 0.4644 × hypertension))) …(5-2)

[0048] (6) Prediction formula for the possibility of developing dyslipidemia (%) after 5 years P = 1 / (1 + exp(-(-5.7225 - 0.0117 × age + 0.3204 × smoking + 0.0369 × LDL cholesterol - 0.0047 × HDL cholesterol + 0.0161 × triglyceride))) ···(6-1) However, in this formula, the value of "smoking" is set to 1 only when currently smoking.

[0049] In addition, in the above formulas (1-1) to (6-1), the AUC of the ROC (Receiver Operating Characteristic) curve is as follows. (1-1) Hypertension: AUC = 0.79 (2-1) Chronic kidney disease (with blood sampling): AUC = 0.89 (3-1) Metabolic syndrome (with blood sampling): AUC = 0.85 (3-2) Metabolic syndrome (without blood sampling): AUC = 0.83 (4-1) Arteriosclerosis (male): AUC = 0.71 (4-2) Arteriosclerosis (female): AUC = 0.77 (5-1) Diabetes (with blood sampling): AUC = 0.89 (5-2) Diabetes (without blood sampling): AUC = 0.71 (6-1) Dyslipidemia (with blood sampling): AUC = 0.73 Thus, for the same lifestyle-related diseases, the prediction formula PR including the first variable EV1, i.e., the one with blood sampling, has a larger AUC. Also, for arteriosclerosis, the AUC is larger when the prediction formula PR is divided by gender, so prediction formulas PR for men and women are generated.

[0050] Returning to Figure 1, the analyzer 3 generates the prediction formulas PR of the above (1-1) to (6-1). The data of the generated prediction formula PR is downloaded to the terminal device 10 via a communication network (not shown).

[0051] [Terminal device] The terminal device 10 is, for example, a smartphone, a mobile terminal, a tablet computer, or a personal computer. Basically, the terminal device 10 is a device that enables user operation input and has a man-machine interface with a display screen. The terminal device 10 includes an input unit 11, a calculation unit 12, and an output unit 13.

[0052] [Input Unit] The input unit 11 inputs the numerical value N1 of the first variable EV1 of the subject and the numerical value N2 of the second variable EV2 obtained from other than blood sampling tests. The inspection device 4 is an inspection device used for specific medical checkups such as a weighing scale and a blood pressure monitor. The input unit 11 inputs the numerical value N1 of the first variable EV1 and the numerical value N2 of the second variable EV2 from this inspection device 4.

[0053] The input unit 11 can also input the numerical value N1 of the first variable EV1 and the numerical value N2 of the second variable EV2 by operating the terminal device 10. For example, as shown in FIG. 3, on the display screen of the terminal device 10, an input image for inputting the numerical value N1 of the first variable EV1 and the numerical value N2 of the second variable EV2 is displayed. By operating while looking at this input image, the numerical value N1 of the first variable EV1 such as age, gender, height, weight, BMI, systolic blood pressure, diastolic blood pressure, presence or absence of drinking, presence or absence of smoking, and family history of hypertension can be input. Note that the input of the numerical values N1 and N2 to the input unit 11 may be performed only by operation input.

[0054] [Calculation Unit] The calculation unit 12 calculates the onset probability P of lifestyle-related diseases after 5 years to be predicted using at least one of the prediction formulas PR of the above formulas (1-1) to (6-1) based on the numerical value N1 of the first variable EV1 and the numerical value N2 of the second variable EV2 input to the input unit 11.

[0055] Regarding hypertension, the calculation unit 12 calculates the onset probability P of hypertension after 5 years using the above formula (1-1) based on the presence or absence of diabetes and the presence or absence of hyperuricemia as the first variable EV1 and the numerical values of age, BMI, systolic blood pressure, diastolic blood pressure, smoking, presence or absence of drinking, and family history of hypertension as the second variable EV2.

[0056] Regarding chronic kidney disease, the calculation unit 12 calculates the onset probability P of chronic kidney disease after 5 years using the above formula (2-1) based on the presence or absence of diabetes, hyperuricemia, and dyslipidemia determined from blood tests as the first variable EV1, the eGFR, and the age, BMI, and presence or absence of hypertension of the subject as the second variable EV2.

[0057] When the lifestyle disease is metabolic syndrome and blood sampling is performed, the calculation unit 12 calculates the onset probability P of metabolic syndrome after 5 years using the above formula (3-1) based on the fasting blood glucose, HDL cholesterol, LDL cholesterol, and triglycerides of the subject as the first variable EV1, and the age, gender, systolic blood pressure, and diastolic blood pressure of the subject as the second variable EV2. Further, when the lifestyle disease is metabolic syndrome and no blood sampling is performed, the calculation unit 12 calculates the onset probability P of metabolic syndrome after 5 years using the above formula (3-2) based on the fasting blood glucose, HDL cholesterol, LDL cholesterol, and triglycerides of the subject as the first variable EV1, and the age, BMI, systolic blood pressure, and diastolic blood pressure of the subject as the second variable EV2.

[0058] Regarding arteriosclerosis, when the subject is male, the calculation unit 12 calculates the onset probability P of arteriosclerosis after 5 years using the above formula (4-1) based on the blood glucose level and triglycerides as the first variable EV1, and the age, BMI, systolic blood pressure, and diastolic blood pressure as the second variable EV2. When the subject is female, the calculation unit 12 calculates the onset probability P of arteriosclerosis after 5 years using the above formula (4-2) based on the age, systolic blood pressure, diastolic blood pressure, and presence or absence of smoking as the first variable EV1, and the blood glucose level as the second variable EV2.

[0059] Furthermore, regarding diabetes, when blood sampling is available, the calculation unit 12 calculates the onset probability P of diabetes after 5 years using the above formula (5-1) based on the fasting blood glucose level and HbA1c as the first variable EV1, and the age, gender, and BMI of the subject as the second variable EV2. When blood sampling is not available, the calculation unit 12 calculates the onset probability P of diabetes after 5 years using the above formula (5-2) based on the age, gender, BMI, and presence or absence of hypertension of the subject as the second variable EV2.

[0060] Furthermore, regarding dyslipidemia, the calculation unit 12 calculates the onset probability of dyslipidemia after 5 years using the above formula (6-1) based on the age, gender, BMI, blood pressure, family history of hypertension, and presence or absence of dyslipidemia of the subject.

[0061] The lifestyle-related diseases to be calculated by the calculation unit 12 may be any one of hypertension, chronic kidney disease, metabolic syndrome, arteriosclerosis, diabetes, and dyslipidemia, or may be a plurality of lifestyle-related diseases among hypertension, chronic kidney disease, metabolic syndrome, arteriosclerosis, diabetes, and dyslipidemia, or the onset probability P may be calculated for all of these. Which lifestyle-related disease is to be predicted can be selected by the user operating the terminal device 10. The user can select a plurality of lifestyle-related diseases as prediction targets.

[0062] [Output unit] The output unit 13 outputs the onset probability P of the lifestyle-related disease after 5 years calculated by the calculation unit 12. The output unit 13 may output the onset probability P of the lifestyle-related disease after 5 years by display or by voice. For example, when the input / output image shown in FIG. 3 is displayed on the display screen of the terminal device 10, the onset probabilities of hypertension and metabolic syndrome can be displayed.

[0063] [Hardware configuration of terminal device 10] The terminal device 10 shown in FIG. 1 is realized, for example, by a computer having the hardware configuration shown in FIG. 4 executing a software program. Specifically, the terminal device 10 includes a CPU (Central Processing Unit) 31 that controls the entire device, a main memory 32 that operates as a work area of the CPU 31, an external memory 33 that stores the operation program of the CPU 31, a man-machine interface 34, a communication interface 35, and an internal bus 38 that connects these components.

[0064] As will be described later, the CPU 31 realizes various functions of the terminal device 10 by executing the program 39 stored in the main memory 32.

[0065] The main memory 32 is composed of a RAM (Random Access Memory) or the like. The program 39 to be executed by the CPU 31 is loaded from the external memory 33 into the main memory 32. Also, the main memory 32 is used as a work area (temporary data storage area) of the CPU 31. The functions of the input unit 11, calculation unit 12, and output unit 13 shown in FIG. 1 are realized by the execution of the program of the CPU 31.

[0066] The external memory 33 is composed of a non-volatile memory such as a flash memory or a hard disk. The program 39 for the CPU 31 to execute is stored in the external memory 33 in advance.

[0067] The man-machine interface 34 includes devices such as a keyboard and a mouse, and an interface device that connects these devices to the internal bus 38. Also, the man-machine interface 34 includes a display device such as a CRT (Cathode Ray Tube) or a liquid crystal monitor. A touch panel can be adopted as the man-machine interface 34. The determination result of the presence or absence of a disease or the like is displayed by the man-machine interface 34.

[0068] The communication interface 35 is an interface for data communication. Through the communication interface 35, data communication with external devices becomes possible. Through this communication interface 35, data communication with the analyzer 3 or the inspection device 4 is performed.

[0069] The functions of the terminal device 10 can be implemented in a computer system composed of one or more computers including one or more processors and one or more storage devices including a non-transitory storage medium. The plurality of computers communicate with each other via a connected communication network while realizing the functions of the terminal device 10. For example, a part of the plurality of functions of the terminal device 10 may be implemented in one computer, and another part may be implemented in another computer.

[0070] Note that the database 2 and the analyzer 3 are also realized by a computer having the hardware configuration shown in FIG. 4 executing a software program.

[0071] Next, the prediction process executed by the terminal device 10 according to the present embodiment will be described. In this prediction process, it is assumed that the lifestyle-related diseases to be calculated are preset.

[0072] As shown in FIG. 5, when the prediction process application is started, for example, an input / output image as shown in FIG. 3 is displayed (step S1). Subsequently, the input unit 11 waits until a calculation command using the prediction formula PR is input (step S2; No). At this stage, the onset probability P [%] of the lifestyle-related disease is not yet displayed. The user can input the numerical value N1 of the first variable EV1 and the numerical value N2 of the second variable EV2 by operating input via the man-machine interface 34.

[0073] Here, for example, referring to the input / output images shown in FIG. 3, operation inputs such as age, gender, height, weight, alcohol consumption, smoking, family history of hypertension, etc. are performed. At this stage, the input unit 11 inputs the data sent from the inspection device 4, and it is assumed that the inspection results of the inspection device 4, such as systolic blood pressure, diastolic blood pressure, etc., are displayed. If the inspection device 4 and the terminal device 10 are not communicably connected, the inspection results of the inspection device 4 are obtained by operation input.

[0074] When a calculation command is input (step S2; Yes), the input unit 11 determines whether all the numerical values N1 of the necessary first variable EV1 and the numerical values N2 of the second variable EV2 have been input (step S3). If there are still numerical values N1 and N2 that have not been input (step S3; No), the output unit 13 performs a display output that prompts the user to input, such as "~ has not been input" (step S4), and returns to step S2. Thereafter, the input unit 11 waits again for the input of the calculation command (step S2; No).

[0075] On the other hand, when a calculation command is input (step S2; Yes) and all the numerical values N1 and N2 have been input (step S3; Yes), the calculation unit 12 uses the above formulas (1-1) to (6-1) corresponding to the lifestyle disease to be predicted based on the numerical value N1 of the first variable EV1 and the numerical value N2 of the second variable EV2 input by the input unit 11, and calculates the onset probability P of the lifestyle disease after 5 years (step S5). Subsequently, the output unit 13 outputs the onset probability P calculated by the calculation unit 12 (step S6). As a result, the onset probability P% of the lifestyle disease to be predicted, such as hypertension and metabolic syndrome as shown in FIG. 3, is displayed.

[0076] Subsequently, the input unit 11 determines whether an end command operation input for the prediction process application has been made (step S7). If the end command has not been input (step S7; No), the input unit 11 returns to step S2.

[0077] Here, the terminal device 10 again waits for the input of a calculation command (step S2; No). At this point, the user can change the numerical value N1 of the first variable EV1 and the numerical value N2 of the second variable EV2 by operating the input. Then, when the numerical values N1 and N2 are changed and a calculation command is input (step S2; Yes), after passing through the determination in step S3, the calculation unit 12 calculates the onset probability P of lifestyle-related diseases using the prediction formula PR based on the changed numerical values N1 and N2 (step S5), and the output unit 13 outputs the updated onset probability P (step S6).

[0078] In this way, it is possible to change the numerical value N1 of the first variable EV1 and the numerical value N2 of the second variable EV2 at any time and update the onset probability P. Thereby, for example, when the weight is reduced by several kilograms, it is possible to confirm how much the onset probability P of lifestyle-related diseases changes. Furthermore, it is possible to confirm how much the onset probability P of lifestyle-related diseases changes when smoking is changed from having it to not having it, or when drinking alcohol is changed from having it to not having it.

[0079] When an end command is input (step S7; No), the terminal device 10 ends the prediction process. Note that the lifestyle-related diseases to be calculated can be changed as appropriate. Also, all lifestyle-related diseases for which the prediction formula PR has been generated may be used as the calculation targets. In this case, for lifestyle-related diseases whose onset probability P is equal to or higher than the threshold value, the onset probability P may be highlighted so as to be easily distinguishable from that of other lifestyle-related diseases.

[0080] Embodiment 2 Embodiment 2 of the present invention will be described. The prediction system 1 according to the present embodiment is the same as the prediction system 1 according to Embodiment 1 above in that, for each lifestyle-related disease and, if necessary, for each condition of the subject, the onset probability P of lifestyle-related diseases after 5 years is calculated using the simplified prediction formula PR (the above formulas (1-1) to (6-1)).

[0081] As shown in FIG. 6, the prediction system 1 according to the present embodiment is different from the prediction system 1 according to Embodiment 1 above in that it includes an image acquisition unit 20 and an extraction unit 21 instead of the input unit 11.

[0082] The image acquisition unit 20 acquires an image of the inspection result report TR captured using a camera or the like built in the terminal device 10. The image acquisition unit 20 may input an image of the inspection result report TR captured by a camera other than the terminal device 10 via a communication network (not shown).

[0083] The extraction unit 21 extracts the numerical value N1 of the first variable EV1 and the numerical value N2 of the second variable EV2 from the image acquired by the image acquisition unit 20.

[0084] The calculation unit 12 calculates the onset probability P of lifestyle-related diseases after 5 years based on the numerical values N1 and N2 of the first variable EV1 and the second variable EV2 extracted by the extraction unit 21. The output unit 13 outputs the calculated onset probability P.

[0085] For example, as shown in FIG. 7, in the examination data area of the display screen of the terminal device 10, the numerical values N1 and N2 of the first variable EV1 and the second variable EV2 obtained from the image of the inspection result report TR are displayed, and the onset probabilities P after 5 years of hypertension, diabetes, chronic kidney disease, and metabolic syndrome are respectively displayed. It is also possible to perform a comprehensive determination based on the onset probabilities P of all lifestyle-related diseases and display the result. The criteria for the comprehensive determination can be determined based on, for example, the number of lifestyle-related diseases with an onset probability of 50% or more.

[0086] The numerical values N1 and N2 displayed in the examination data area can be changed in the same manner as in the first embodiment. When the numerical values N1 and N2 are changed, recalculation using the prediction formula PR is performed each time, and the onset probabilities P after 5 years of hypertension, diabetes, chronic kidney disease, and metabolic syndrome are updated.

[0087] As described in detail above, according to the prediction system 1 according to the present embodiment, using the prediction formula PR generated by logistic regression analysis, based on the numerical value N1 of the first variable EV1 and the numerical value N2 of the second variable EV2, the onset probability P of lifestyle-related diseases after 5 years can be calculated and presented. Therefore, it is possible to "visualize" what kind of lifestyle-related diseases one is likely to develop in the future.

[0088] Also, according to the prediction system 1, by changing the numerical value N1 of the first variable EV1 and the numerical value N2 of the second variable EV2, it is also possible to "visualize" how much the onset probability P of lifestyle-related diseases changes. As a result, it becomes easier for the person himself or the person providing lifestyle guidance to grasp which items need to be improved in order to reduce the onset probability P of lifestyle-related diseases.

[0089] In the above embodiment, by changing the numerical value N1 of the first variable EV1 and the numerical value N2 of the second variable EV2, the onset probability P of lifestyle-related diseases after 5 years is updated at any time. However, it may be possible to display both the onset probability P with the original numerical values N1 and N2 and the onset probability P with the changed numerical values N1 and N2 in a comparable manner. Furthermore, the relationship between the change in the numerical values N1 and N2 of the first variable EV1 and the second variable EV2 and the change in the onset probability P may be graphically displayed. By doing so, it becomes easier to grasp the correlation between the numerical values N1 and N2 of the first variable EV1 and the second variable EV2 and the onset probability P.

[0090] Also, according to the prediction system 1 according to the second embodiment described above, it is possible to obtain an image of the test result report TR of a specific medical check-up including the test results of a blood test, and read the numerical value N1 of the first variable EV1 and the numerical value N2 of the second variable EV2 from the image of the test result report TR of the specific medical check-up and use them as inputs to the prediction formula PR. Since the test result report TR of the specific medical check-up has a nationally unified format and the medical check-up items are determined, it is possible to easily read the numerical values from the image. As a result, it is possible to prevent the occurrence of data loss that makes it difficult to calculate the prediction formula PR, and thus it is possible to reliably predict the onset probability P of each lifestyle-related disease.

[0091] In addition, in the prediction system 1, it is possible to calculate the onset probability P of metabolic syndrome and diabetes using a prediction formula PR that does not use the numerical value N1 of the first variable EV1 obtained from the results of blood tests as an explanatory variable. By doing so, without performing a blood test, the subject himself / herself can use this application of the prediction process to check the onset probability P of future lifestyle-related diseases and perform health management.

[0092] Also, for the same lifestyle-related diseases, the onset probabilities P predicted with and without blood sampling may be displayed in a comparable manner.

[0093] In the above embodiment, the terminal device 10 includes the calculation unit 12. However, this is not limiting. The terminal device 10 may include an input unit 11 and an output unit 13, and the calculation unit 12 may be provided in a server computer communicably connected to the terminal device 10. It is possible to flexibly design which device the input unit 11, the calculation unit 12, the output unit 13, the image acquisition unit 20, and the extraction unit 21 are incorporated into.

[0094] In the above embodiment, the onset probability of lifestyle-related diseases after 5 years has passed is predicted, but this is not limiting. The number of years may be 1 to 4 years, 6 years or more, or even a few months later.

[0095] In addition, the hardware configuration and software configuration of the terminal device 10 are examples and can be arbitrarily changed and modified.

[0096] The central part that processes the terminal device 10, which is composed of a CPU 31, a main memory 32, an external memory 33, a man-machine interface 34, a communication interface 35, an internal bus 38, etc., may be constructed as a dedicated system as described above, or may be realized using a normal computer system. For example, a computer program for executing the above operations may be stored and distributed in a non-temporary computer-readable recording medium (flexible disk, CD-ROM, DVD-ROM, etc.), and the terminal device 10 that executes the above processing may be configured by installing the computer program on a computer. Also, the computer program may be stored in a storage device of a server device on a communication network such as the Internet, and the terminal device 10 may be configured by downloading it with a normal computer system, etc.

[0097] When realizing the functions of a computer through the division of labor between an OS (operating system) and an application program, or through the cooperation between the OS and the application program, etc., only the application program part may be stored in a recording medium or a storage device.

[0098] It is also possible to superimpose a computer program on a carrier wave and distribute it via a communication network. For example, a computer program may be posted on a bulletin board (BBS, Bulletin Board System) on a communication network, and the computer program may be distributed via the network. Then, this computer program may be started and configured to be able to execute the above processing by being executed in the same manner as other application programs under the control of the OS.

[0099] The present invention can be implemented in various embodiments and variations without departing from the broad spirit and scope of the present invention. Further, the above-described embodiments are for explaining the present invention and do not limit the scope of the present invention. That is, the scope of the present invention is indicated by the claims rather than the embodiments. And various modifications made within the scope of the claims and within the scope of the meaning of the invention equivalent thereto are considered to be within the scope of the present invention.

Industrial Applicability

[0100] The present invention can be applied to predicting lifestyle diseases of a target person after a predetermined number of years have passed.

Explanation of Reference Numerals

[0101] 1 Prediction system, 2 Database, 3 Analysis device, 4 Inspection device, 10 Terminal device, 11 Input unit, 12 Calculation unit, 13 Output unit, 20 Image acquisition unit, 21 Extraction unit, 31 CPU, 32 Main memory, 33 External memory, 34 Man-machine interface, 35 Communication interface, 38 Internal bus, 39 Program, BD Learning data, EV1 First variable, EV2 Second variable, N1, N2 Numerical values, PR Prediction formula, P Onset probability, R Presence or absence of lifestyle disease after 5 years (result), TR Inspection result report

Claims

1. A variable that shows a significant difference in the presence or absence of lifestyle-related diseases after a specified number of years has passed. Using as explanatory variables a first variable whose numerical value is obtained from a blood test and a second variable whose numerical value is obtained from sources other than the blood test, and performing a logistic regression analysis with the probability of developing a lifestyle-related disease after a specified number of years as the objective variable, a calculation unit that calculates the probability of onset of the subject based on the numerical values of the first variable and the second variable of the subject is provided. A prediction system.

2. When the lifestyle-related disease is hypertension, the first variable includes the presence or absence of diabetes and the presence or absence of hyperuricemia, the second variable includes age, BMI, systolic blood pressure, diastolic blood pressure, the presence or absence of smoking, and the presence or absence of a family member with hypertension. The prediction system according to claim 1.

3. When the lifestyle-related disease is chronic kidney disease, the first variable includes the presence or absence of diabetes, the presence or absence of hyperuricemia, the presence or absence of dyslipidemia, and eGFR, the second variable includes age, BMI, and the presence or absence of hypertension. The prediction system according to claim 1.

4. When the lifestyle-related disease is metabolic syndrome, the first variable includes fasting blood glucose level, HDL cholesterol, LDL cholesterol, and triglycerides, the second variable includes age, gender, BMI, systolic blood pressure, diastolic blood pressure, the presence or absence of smoking, and the presence or absence of alcohol consumption. The prediction system according to claim 1.

5. When the lifestyle-related disease is arteriosclerosis, in the case where the subject is male, the first variable includes blood glucose level and triglycerides, while in the case where the subject is female, the first variable includes blood glucose level. In the case where the subject is male, the second variable includes age, BMI, systolic blood pressure, and diastolic blood pressure, while in the case where the subject is female, the second variable includes age, systolic blood pressure, diastolic blood pressure, and the presence or absence of smoking. The prediction system according to claim 1.

6. When the lifestyle-related disease is diabetes, the first variable includes fasting blood glucose level and HbA1c, the second variable includes age, gender, and BMI. The prediction system according to claim 1.

7. When the lifestyle-related disease is dyslipidemia, the first variable includes HDL cholesterol, LDL cholesterol, and triglycerides, the second variable includes age and the presence or absence of smoking. The prediction system according to claim 1.

8. A prediction formula obtained by performing a logistic regression analysis with age, gender, BMI, presence or absence of hypertension, presence or absence of smoking, and presence or absence of alcohol consumption as explanatory variables and the onset probability of metabolic syndrome after a predetermined number of years as the objective variable is used to calculate the onset probability of metabolic syndrome in a subject after a predetermined number of years based on the numerical values of the explanatory variables of the subject. A calculation unit is provided. Prediction system.

9. With age, gender, BMI, presence or absence of hypertension, presence or absence of family members with hypertension, and presence or absence of dyslipidemia as explanatory variables, and the onset probability of diabetes after a predetermined number of years as the objective variable, using a prediction formula obtained from a logistic regression analysis performed with the onset probability of diabetes after a predetermined number of years as the objective variable, based on the numerical values of the explanatory variables of the subject, a calculation unit for calculating the onset probability of diabetes in the subject after a predetermined number of years is provided. Prediction system.

10. An image acquisition unit for acquiring an image of an inspection result report, An extraction unit for extracting the numerical values of the explanatory variables from the image, and The calculation unit is Based on the numerical values of the explanatory variables extracted by the extraction unit, the onset probability is calculated. The prediction system according to any one of claims 1 to 9.

11. A computer, A variable that shows a significant difference in the presence or absence of lifestyle-related diseases after a predetermined number of years, with a first variable whose numerical value is obtained from a blood test and a second variable whose numerical value is obtained from other than the blood test as explanatory variables, and using a prediction formula obtained by performing a logistic regression analysis with the onset probability of lifestyle-related diseases after a predetermined number of years as the objective variable, based on the numerical values of the first variable and the second variable of the subject, a calculation unit for calculating the onset probability of the subject. A program for functioning as.

Images