JP4728616B2 - Blood sugar level data analysis system, blood sugar level data analysis program - Google Patents

Description

The present invention, by processing the blood glucose level data relates blood glucose data analysis system for obtaining information useful for the prevention and diagnosis of diabetes, the blood glucose level data analysis program.

  The number of diabetic patients in Japan is large: “type 1 diabetes (insulin-dependent diabetes)”, which accounts for less than 5% of the total, and “type 2 diabetes (non-insulin-dependent diabetes)”, which accounts for more than 90% of the total. Separated. Type 1 diabetes is a type in which insulin-secreting cells in pancreatic islets are destroyed and insulin is abnormally deficient, and mainly affects children under 15 years of age. On the other hand, type 2 diabetes is a type that requires adults over 40 years of age and is related to the environment and aging.

  As a typical test for determining the diagnosis and treatment policy of such diabetes, “sugar tolerance test (75 g O-GTT)” can be mentioned. The glucose tolerance test is a test in which 75 g of sugar is taken, blood glucose, urine sugar, and blood insulin levels are measured and the time variation is observed. Furthermore, the Diabetes Society of Japan has proposed “Diabetes Diagnosis Procedure (New Guidelines)” (see, for example, Non-Patent Document 1). According to this “diagnosis procedure for diabetes”, blood glucose level in the fasting state, 2-hour value of blood glucose level data in the glucose tolerance test, whether or not the blood glucose level at any time exceeds the reference value, and the presence of other diabetes-related symptoms Shows a method for diagnosing whether or not a patient has diabetes.

  FIG. 10 is a diagram showing the criteria for diabetes diagnosis in a conventional glucose tolerance test.

  In FIG. 10, the vertical axis indicates the blood glucose level during fasting, and the horizontal axis indicates the 2-hour value of blood glucose level data in the glucose tolerance test. As shown in FIG. 10, if both the fasting blood glucose level and the 2-hour value of the blood glucose level data in the glucose tolerance test are equal to or less than the reference value, it is diagnosed as normal. Further, if one or both of the fasting blood glucose level and the 2-hour value of the blood glucose level data in the glucose tolerance test are within different reference ranges, the border type is diagnosed. Further, diabetes is diagnosed when one or both of the fasting blood glucose level and the two-hour blood glucose level data in the glucose tolerance test exceed a certain value.

  On the other hand, as a typical non-invasive or minimally invasive blood glucose level measuring device, there is a blood glucose meter that transmits a specific wavelength of light through a fingertip and measures the blood glucose level from the light attenuation rate (for example, Patent Document 1). reference). In this blood glucose meter, a time change pattern is calculated from the measured blood glucose level data of the patient. Then, the calculated time change pattern of the blood glucose level is compared with a known time change pattern of the blood glucose level stored in advance to determine the disease type and the dose of the medicine.

  In addition, as blood glucose level measurement data management software, diabetes data that performs processing such as average, maximum and minimum values, time series display, frequency display, hourly display, and ratio display within and outside the specified range within a certain period There is a management system (see, for example, Non-Patent Document 2).

Further, a chronic disease patient self-management support system is proposed that supports the patient himself / herself to manage life status information such as meal content, medication content, physical condition and the like of the chronic disease patient (see, for example, Patent Document 2). This system records patient life state information on an IC card, and can be edited or displayed on both the patient terminal and the medical institution terminal.
JP 2001-212114 A Japanese Patent Laid-Open No. 08-140944 Satoshi Kasama “Dementia and medical information disclosure website”, [online] Diabetes diagnosis procedure (new guidelines), [Search June 1, 2004], Internet <URL: http: //www.inetmie.or .jp / ~ kasamie / index.html> Soshinsha "Diabetes Network", [online], SKK, MEQNET DM Manager, [October 6, 2003 search], Internet <URL: http: //www.dm-net. co.jp/skk/mane.htm>

  In the diagnosis of diabetes, information useful for diagnosis is obtained from the blood glucose level data of the patient, for example, an index indicating the glucose metabolism function is obtained, the glucose metabolism function of the patient is specifically grasped, and the glucose metabolism function index It is important to obtain the necessary amount of meal and exercise as a specific value according to the value.

  However, according to the conventionally proposed diagnostic procedure for diabetes (new guidelines), it is possible to determine whether or not a patient is diabetic based on blood glucose level data. In particular, it is impossible to obtain information such as the amount of exercise and the amount of meal necessary for specifically indicating or controlling the blood glucose level.

  In addition, since the conventional blood glucose meter simply determines the rate of change in blood glucose level over time and compares it with known data, it is necessary to input the patient's meal amount and exercise amount separately. If there is an error in the momentum or the momentum, there is a drawback that the determination result is distorted. Further, when data is used in determination of blood glucose level data, there is a possibility that the user's arbitraryness may be affected, and there is a possibility that the determination result varies for each user.

  On the other hand, a conventional diabetes data management system calculates statistical data such as an average value of blood glucose level within a certain period, and supports parameters for directly reflecting a patient's glucose metabolism function although it assists in the diagnosis of diabetes. Since it is not calculated, it is difficult to obtain information such as the necessary amount of exercise and the amount of meal and to grasp the patient's glucose metabolism function specifically.

  In addition, the chronic disease patient self-management support system only records or displays history information about various information related to patients such as a meal information list and a test information list and life information such as a meal information graph. It does not require information such as parameters that directly reflect. For this reason, there is a possibility that blood glucose level data cannot be sufficiently used for diagnosis of diabetes.

The present invention has been made in order to cope with such a conventional situation. The sugar metabolism function is extracted from the patient's blood glucose level data as patient-specific information and digitized to serve as an index. function and an object diabetes prevention and diagnosis can obtain useful information to the blood glucose level data analysis system, to provide a blood glucose level data analysis program based on.

Blood glucose level data analysis system according to the present invention, a blood glucose level measurement data database for storing measurement data of the blood glucose level at each time of a subject, at least one of meal size and momentum of the subject to an impulse input, its impulse An impulse response y (t) to an input is expressed by a linear system in which the subject's glucose metabolism function is modeled by a linear system represented by a second-order lag element representing a temporal change in blood glucose level. Diabetes index value which is a means for obtaining an index, and outputs a parameter relating to a time constant of the mathematical formula obtained based on the measurement data read from the blood glucose level measurement data database as a numerical value of an index representing the glucose metabolism function And calculating means .

Blood glucose level data analysis system according to the present invention, Oite the blood glucose level data analysis program is to quantify the patient's glucose metabolism function, information useful for the prevention and diagnosis of diabetes based on the digitized sugar metabolism Obtainable.

Blood glucose level data analysis system according to the present invention will be described with reference to the accompanying drawings showing a preferred embodiment of the blood glucose level data analysis program.

  FIG. 1 is a functional block diagram showing a first embodiment of a blood sugar level data analysis system according to the present invention.

  The blood glucose level data analysis system 1 reads a blood glucose level data analysis program into a computer 4 equipped with a monitor 2 and an input device 3, thereby allowing a blood glucose level measurement data database 5, a diabetes index value calculation means 6, a diabetes index value determination means. 7 is made to function.

  In the blood glucose level measurement data database 5, blood glucose level measurement data at each time of a subject such as a patient or a patient is measured and stored in advance by any method.

  The diabetes index value calculation means 6 reads blood sugar level measurement data from the blood sugar level measurement data database 5 and uses the mathematical formula obtained by modeling the subject's sugar metabolism function as a linear system, It has a function to obtain an index of metabolic function as a numerical value. That is, the subject's glucose metabolism function can be approximated by a linear system.

  Here, diabetes is roughly classified into type 1 diabetes (insulin-dependent diabetes) and type 2 diabetes (non-insulin-dependent diabetes). The main cause of type 1 diabetes is insulin secretion failure, which does not produce enough insulin to lower blood sugar levels. On the other hand, the main cause of type 2 diabetes is so-called insulin resistance, in which insulin levels are relatively high, and insulin secretion does not work well despite the increase in glucose secretion and glucose levels do not decrease. Is to have.

  To determine whether such diabetics are type 1 diabetics, type 2 diabetics, or healthy individuals by examining the temporal changes in blood glucose level and insulin level in the glucose tolerance test. Can do. The glucose tolerance test takes 75 g (300 calories) of sugar, measures blood glucose, urine sugar, and blood insulin levels after 30 minutes, 1 hour, and 2 hours, and observes the blood glucose level change pattern over time. Is.

  FIG. 2 is a diagram showing temporal changes in blood glucose level and insulin level in a glucose tolerance test of healthy subjects, type 1 diabetes patients, and type 2 diabetes patients.

  2A, 2B, and 2C, the right vertical axis represents the blood glucose level, the left vertical axis represents the insulin value, and the horizontal axis represents time. 2 (a), 2 (b), and 2 (c), the solid line and the square mark or circle mark are blood glucose level measurement data D1 indicating the time change of the blood glucose level, and the dotted line and triangle mark are secreted insulin. It is insulin value data D2 which shows the time change of a value.

  FIG. 2A is a diagram showing blood glucose level measurement data D1 and insulin level data D2 of a healthy person. As shown in FIG. 2 (a), in the case of a healthy person, when the blood glucose level measurement data D1 increases due to the sugar load, the insulin level data D2 also increases as the blood glucose level measurement data D1 increases. It can be seen that the blood glucose level measurement data D1 decreases due to the action of insulin. Further, when the blood glucose level measurement data D1 decreases, the insulin level data D2 also decreases following the blood glucose level measurement data D1, and the maximum value of the blood glucose level measurement data D1 appears within 2 hours.

  FIG. 2B is a diagram showing blood glucose level measurement data D1 and insulin level data D2 of a type 1 diabetes patient. As shown in FIG. 2 (b), in the case of a type 1 diabetic patient, the insulin value data D2 before sugar loading is low. Further, even if the blood glucose level measurement data D1 increases due to the sugar load, the insulin level data D2 does not increase because there is no insulin secretion, and the blood glucose level measurement data D1 does not decrease because there is no action of insulin. It shows a high price.

  FIG.2 (c) is a figure which shows the measurement data D1 and insulin level data D2 of the blood glucose level of a type 2 diabetes patient. As shown in FIG. 2C, in the case of a type 2 diabetic patient, the insulin value data D2 before sugar loading is relatively high. In addition, when the blood glucose level measurement data D1 increases due to the sugar load, insulin secretion increases and the insulin level data D2 increases, but insulin does not work sufficiently, so the blood glucose level measurement data D1 decreases. It shows a high price.

  That is, as shown in FIGS. 2 (a), 2 (b), and 2 (c), the time-varying pattern of blood glucose level measurement data D1 and insulin level data D2 for a glucose tolerance test is different between a healthy person and a diabetic patient. You can see that they are different. For example, in FIG. 2 (a), focusing only on the time change of the blood glucose level measurement data D1, the blood glucose level measurement data D1 once rises to the maximum value in a healthy person, and thereafter the blood glucose level measurement data D1 decreases. is doing. However, as shown in FIGS. 2 (b) and 2 (c), in the type 1 diabetic patient with insulin secretion deficiency and the type 2 diabetic patient with insulin resistance, the time required for the blood glucose level measurement data D1 to rise is increased. The change pattern is shown.

  Further, as shown in FIGS. 2B and 2C, the difference between the type 1 diabetic patient and the type 2 diabetic patient can be identified by the time change pattern of the insulin value data D2.

  Here, in FIG. 2B and FIG. 2C, if it is assumed that the blood glucose level measurement data D1 of type 1 diabetic patients and type 2 diabetic patients continues to rise, the type 1 diabetic patients and type 2 diabetic patients Therefore, if the follow-up observation of the time change pattern of the blood glucose level measurement data D1 over 2 hours or more is allowed for type 1 diabetic patients and type 2 diabetic patients, measurement of the blood glucose level is surely required. The data D1 should have a time change pattern that decreases after reaching the maximum value.

  That is, as a characteristic common to type 1 diabetic patients and type 2 diabetic patients, the sugar metabolic function of diabetic patients is lowered and the rate of sugar metabolism is slow. Therefore, focusing on only the viewpoint of the sugar metabolism function, the sugar metabolism function of patients with type 1 diabetes and type 2 diabetes can be regarded as a linear system represented by a second-order lag element and modeled.

  Then, the amount of meal that is decomposed by the glucose metabolism function of a diabetic patient is set as a positive input, the amount of exercise is set as a negative input, and these positive and negative inputs are added at the input unit and input to this linear system, while discretely It can be considered that the linear system outputs the absolute value of the continuously sampled blood glucose level measurement data D1.

  Therefore, the rise time constant in the measurement data D1 of the blood glucose level of the type 1 diabetic patient and the type 2 diabetes patient, which is the output of the linear system, is much higher than the rise time constant in the blood glucose measurement data D1 of the healthy person. It can be interpreted as long. However, since the time constant when the blood glucose level measurement data D1 rises and the time constant when the blood glucose level measurement data D1 falls are not necessarily the same, the degree of freedom of the time constant is considered. There is a need.

  The reason for approximating the glucose metabolism function of a diabetic patient as a second-order lag element is that the measurement data D1 of the blood glucose level of the diabetic patient continues to rise despite taking 75 g of sugar at a time. This is because a delay occurs. In other words, if the glucose metabolism function of a diabetic patient is approximated as an impulse response with a first-order lag element, the output of the linear system instantaneously reaches the maximum value and then decays over time. D1 cannot be approximated.

  Furthermore, if the glucose metabolism function of a diabetic patient is approximated as a second-order lag element, it is possible to obtain a merit that can be simply described with only the gain constant K and the two types of time constants T1 and T2. Then, if the time constants T1 and T2 are obtained in a linear system approximated by a second-order lag element, the obtained time constants T1 and T2 directly reflect the sugar metabolism function of the diabetic patient, or at least the sugar of the diabetic patient. The time constants T1 and T2 can be used for diagnosis on the assumption that the parameters are closely related to the metabolic function.

  In such a linear system, when the amount of meal as an input is examined, the time required for taking 75 g of sugar or the normal meal time is about 30 minutes, so that the blood glucose level measurement data D1 increases. Is considered to be sufficiently shorter than the time constants T1 and T2 at the time of descending or descending. Furthermore, it is known that the average meal time spent per day is 1 hour and 35 minutes. Similarly, when examining the momentum that is the input of the linear system, the amount of exercise per day is usually less than one hour.

  From these facts, if the amount of meal at time t is the input x1 (t) of the linear system and the amount of exercise is the input x2 (t) of the linear system, the inputs x1 (t) and x2 (t) are impulses in the control theory. It can be considered. That is, the time change pattern of the blood glucose level measurement data D1 in a diabetic patient can be approximately regarded as an impulse response. Similarly, the temporal change pattern of the blood glucose level measurement data D1 of a healthy person can be regarded as an impulse response of a linear system represented by a second-order lag element.

  FIG. 3 is a block diagram of a linear system used when the blood sugar level data analysis system 1 shown in FIG.

If the glucose metabolism function of a diabetic patient or a healthy person is regarded as a linear system represented by a second-order lag element, the glucose metabolism function is represented by a block diagram as shown in FIG. That, X ₁ Laplace transform of the input positive a meal size that is decomposed into sugars x1 _(t), the Laplace transform of the negative x2 (t) is a momentum that leads to sugar consumption and X _2, the linear system input is the sum of _{X 1} and _{X 2.} However, t indicates the time in the real time region, x1 (t) is the amount of meal converted into the amount of sugar, and x2 (t) is the amount of exercise load converted into the amount of sugar.

The output of the linear system is a Laplace transform Y of y (t), where y (t) is the absolute value of the blood glucose level measurement data D1. Precisely, y (t) is a value obtained by subtracting the minimum absolute value of the measurement data D1 from the absolute value of the measurement data D1 of the blood glucose level measured at time t as shown in the equation (1).
[Equation 1]
y (t) = (absolute value of blood glucose level measurement data at time t) − (minimum absolute value of blood glucose level measurement data) (1)
Further, as shown in FIG. 3, the linear system is represented by a second-order lag element, and thus is represented by an expression having a quadratic expression having a frequency s using constants α, β, and K as a denominator. If the time constants of the linear system are T1 and T2, the constants α and β are defined by equation (2).
[Equation 2]
α = 1 / T1
β = 1 / T2
... (2)
When the linear system shown in the block diagram of FIG. 3 is mathematically expressed, the relationship of Expression (3) is established.

  That is, Expression (3) is an expression that represents the block diagram input / output relationship shown in FIG. 3 in the frequency s region.

On the other hand, it is known that the impulse response in the real time domain of the linear system shown by the block diagram shown in FIG.

  However, the delta function δ (t) indicates an impulse input, and the value obtained by integrating the impulse δ (t) with −∞ <t <∞ is equal to 1.

  That is, the equation (4) is an equation representing an impulse response in the real-time region for the second-order lag element, where time is t, gain constants are K, α, β, and input is impulse δ (t).

Furthermore, since the input x1 (t) as the amount of meal and the input x2 (t) as the amount of exercise can be regarded as impulses as described above, they can be expressed using the delta function δ (t). However, since there should be a difference in the amount of meal and the amount of exercise, the inputs x1 (t) and x2 (t) are expressed by Expression (5), where c1 and c2 are certain constants.
[Equation 5]
x1 (t) = c1 · δ (t)
x2 (t) = c2 · δ (t)
... (5)
Therefore, the added value x (t) of the amount of meal x1 (t) and the amount of exercise x2 (t) input to the linear system shown in FIG.
[Equation 6]
x (t) = x1 (t) + x2 (t) = c1 · δ (t) + c2 · δ (t) (6)
Therefore, in the time domain when the addition value x (t) of the amount of meal x1 (t) and the amount of exercise x2 (t) is input as an impulse to the linear system shown in FIG. The impulse response of is expressed by equation (7).

  FIG. 4 is a graph of the impulse response of the linear system used when the blood glucose level data analysis system 1 shown in FIG.

  In FIG. 4, the horizontal axis represents time t, and the vertical axis represents the function value of the input x (t) and output y (t) to the linear system.

  In Expression (7), when the constants K, α, β, c1, and c2 are constant, the absolute value y (t) of the blood glucose level measurement data that is the output of the linear system is graphed in the time domain as shown in FIG. It becomes a curve as shown by a solid line. Also, the input x (t) of the linear system is indicated by a dotted line.

According to FIG. 4, it can be seen that the absolute value y (t) of the blood glucose level measurement data, which is the output of the linear system, rises and falls with the maximum value as much as possible. Therefore, if the time when the output y (t) of the linear system takes the maximum value is t ₀ , the time t ₀ is obtained by solving the first derivative y ′ (t) = 0 of the output y (t). (8) can be obtained as follows.

That is, Expression (8) is an expression for obtaining the time t ₀ when the impulse response y (t) in the real time region with respect to the second-order lag element takes a maximum value. According to the equation (8), it can be seen that the smaller the difference between α and β is, the closer the value of (α−β) is closer to 0, the smaller the denominator of the equation (8) becomes, so the time t ₀ becomes larger. That is, as the value of (α−β) is closer to 0, the time t ₀ until the absolute value y (t) of the blood glucose level measurement data reaches the maximum value becomes longer, and blood glucose commonly seen in diabetic patients It turns out that it resembles the time change pattern of the measured value data.

  For this reason, it can be seen that whether α and β and (α−β) are close to 0 can be used as important judgment criteria in diabetes diagnosis. Therefore, it can be seen that it is important to determine the values of α, β, and (α−β) in order to appropriately support diabetes diagnosis.

  Incidentally, in the block diagram of FIG. 3 and the linear system represented by the equations (3) and (7), the input x (t) (= x1 (t) + x2 (t)) can be regarded as an intermittently generated impulse. It is only known that the values of constants c1 and c2, which are information indicating the magnitude of the input x1 (t) as the amount of meal and the input x2 (t) as the amount of exercise, are detected by a diabetic patient or a healthy person. It is an unknown because it can only be measured by another means other than the instrument. Each constant K, α, β is also an unknown number.

  As described above, the constants K, α, β and the constants c1 and c2 representing the amount of meal and exercise amount are unknown numbers, while the absolute value y (t) of the blood glucose level measurement data that is the output of the linear system. Assuming that only the change data is known, if the glucose metabolism function index of a diabetic or healthy person can be estimated as a numerical value, the estimated index is universal biological information that does not depend on the amount of meal or exercise. It is expected to be very effective for diagnosis support of diabetes.

  Therefore, a method for estimating the values of the constants α and β from the measurement data y (t) will be described. First, the values of the constants α and β are estimated by dividing each blood glucose level measurement data y (t) from a minimum value to a maximum value and then to a series of data until it reaches a minimum value. To do. That is, in the blood glucose level measurement data y (t), the start time t = 0 is the time when the segmented measurement data y (t) first becomes the minimum value.

Further, blood glucose level measurement data is approximated by a smooth function capable of up to second order differentiation. For example, a spline function can be used as a function capable of up to second order differentiation. Then, only the blood glucose level measurement data y (t) itself can be measured in the real time region, but the additional information can be extracted by differentiating the blood glucose level measurement data y (t). That is, from the blood glucose level measurement data y (t), a function y ′ (t) (= dy (t) / dt) obtained by first-order differentiation of the measurement data y (t) with respect to time t and the measurement data y (t) are obtained. A function function y ″ (t) (= d ² y (t) / dt ² ) obtained by second-order differentiation at time t can be obtained.

Then, equations (9) and (10) are obtained by taking respective ratios of the obtained functions y ′ (t), y ″ (t) and measurement data y (t).

  Equation (9) is an equation representing data (y ′ (t) / y (t)) obtained from blood glucose measurement data y (t) and first-order differentiated function y ′ (t). (10) is an expression representing data (y ″ (t) / y (t)) obtained from the blood glucose level measurement data y (t) and the second-order differentiated function y ″ (t).

  Therefore, as shown in the equation (7), the constant K is a product of (c1 + c2) and cannot be estimated by separation, but the nonlinear simultaneous equations consisting of the equations (9) and (10) are By solving numerically, the constants α and β at time t can be estimated as parameters. That is, the parameters α, β, and (α−β) can be obtained as a function of time t as an index of the sugar metabolism function.

  The diabetes index value calculation means 6 of the blood sugar level data analysis system 1 reads the blood sugar level measurement data y (t) measured and stored in the blood sugar level measurement data database 5 in advance, and displays the glucose metabolism function index. As at least one of the parameters α, β and (α-β), preferably all as a function of time t.

  FIG. 5 shows blood glucose level measurement data derived from the linear system used when the blood sugar level data analysis system 1 shown in FIG. FIG. 4 is a graph of parameters α, β and (α−β) obtained as an index of a sugar metabolism function from y (t) as a function of time t.

  FIG. 5 (a) shows blood glucose level measurement data y (t) of a healthy person used to determine parameters α, β and (α-β), which are indicators of glucose metabolism function, and the horizontal axis represents time. t represents the value of y (t) on the vertical axis. The blood glucose level measurement data y (t) is divided into a series of data from the minimum value to the maximum value until it reaches the minimum value, as indicated by the dotted frame in FIG. 5A. The start time t = 0 of each measurement data y (t) is the time at which the measurement data y (t) in each region first has a minimum value.

  FIG. 5B shows the parameter α obtained from the measurement data y (t) shown in FIG. 5A, the horizontal axis shows the time t, and the vertical axis shows the value of the parameter α. According to FIG. 5B, it can be seen that the parameter α of the healthy person repeatedly increases and decreases with time t, and shows a tendency to have a maximum value and a minimum value in each region. It can be seen that there is a time when the parameter α exceeds a certain value in at least each region.

  FIG. 5C shows the parameter β obtained from the measurement data y (t) shown in FIG. 5A, the horizontal axis shows time t, and the vertical axis shows the value of parameter β. According to FIG. 5C, it can be seen that the parameter β of the healthy person has a small temporal variation compared to the parameter α, and tends to always be less than or less than a certain value.

  FIG. 5 (d) shows the parameter (α−β) obtained from the measurement data y (t) shown in FIG. 5 (a), the horizontal axis represents time t, and the vertical axis represents the parameter (α-β). Each value is shown. According to FIG. 5 (d), it can be seen that the parameter (α−β) of a healthy person repeatedly increases and decreases with time t, and shows a tendency to have a maximum value and a minimum value in each region. Moreover, it turns out that the parameter ((alpha)-(beta)) of a healthy person shows the tendency which becomes always more than a fixed value.

  The parameters α, β and (α−β) thus obtained by the diabetes index value calculation means 6 are given to the diabetes index value determination means 7 as an index of the glucose metabolism function.

  The diabetes index value determination means 7 compares the values of the parameters α, β and (α−β) received as an index of the glucose metabolism function from the diabetes index value calculation means 6 with a preset threshold value, so that the parameters α, a function of determining whether blood glucose measurement data y (t) used for obtaining β and (α-β) is that of a healthy person, or of a diabetic patient or a borderline patient, The determination result is given to the monitor 2 and displayed.

By the way, when eating continuously over a longer time than usual, the input x (t) to the linear system continues, so that the apparent sugar metabolism function behaves as if it has been reduced. That is, the time constants T ₁ (= 1 / α) and T ₂ (= 1 / β) tend to be calculated larger. Even when no movement is performed, the input x (t) increases and the apparent time constants T ₁ and T ₂ tend to be calculated larger. When determining whether a healthy person, whether the time constant T ₁ and T ₂ is small is an important criterion.

Therefore, there is a case for the parameter alpha is that, as shown in a portion surrounded by a dotted line in FIG. 5 (b), a case or a threshold epsilon _alpha or exceeds the parameter alpha preset threshold is epsilon _alpha in each area It can be used as a criterion for determining whether the measurement data y (t) belongs to a healthy person. As for parameter β, as shown in FIG. 5C, whether parameter β is always less than or less than a certain threshold value ε _β, whether measurement data y (t) is that of a healthy person or not. It can be a criterion. Further, as shown in FIG. 5 (d), the parameter (α-β) indicates whether the parameter (α-β) is always greater than or equal to a certain threshold value ε _α-β. It can be used as a criterion for determining whether or not it is.

  Then, the diabetes index value determination means 7 determines whether the blood glucose level measurement data y (t) is that of a healthy person by comparing each of the parameters α, β, (α−β) with each threshold value. Alternatively, it can be determined whether the patient is a diabetic patient or a borderline patient, and the determination result is given to the monitor 2 together with the values of the parameters α, β, and (α−β) as necessary. Is done.

On the other hand, a case where the parameter α (= 1 / T ₁ ) related to the time constant T ₁ in the linear system can be estimated more easily is a case where the time constant T ₁ and the time constant T ₂ are extremely different. It is done. For example, when the time constant T ₁ is sufficiently larger than the time constant T ₂ and there is a relationship of T ₁ >> T ₂ , for the delta function δ (t) in the real time domain of the linear system of Equation (4) The exponential term including the parameter β can be ignored in the equation indicating the impulse response y (t). Therefore, Equation (4) can be approximated as Equation (11) using the parameter α.

That is, Expression (11) is an expression approximated by ignoring the exponent term including the parameter β in Expression (4) when there is a relationship of T ₁ >> T ₂ .

Furthermore, the calculation formula for obtaining the parameter α from the approximate expression (11) and the blood glucose level measurement data y (t) is the expression shown in the expression (12).

  As described above, when the exponential term including one of the parameter α or the parameter β is ignored and the approximate expression equivalent to the expression (11) can be regarded as being established, the diabetes index value calculation means 6 can calculate the expression (12 ) Can be obtained from the blood glucose level measurement data y (t) by simpler calculation.

  FIG. 6 shows the expression (12) derived from the linear system used when the glucose metabolism function analysis system 1 shown in FIG. It is the figure which graphed the parameter (alpha) calculated | required as a parameter | index of glucose metabolism function from the measurement data y (t) of value, and the estimated input x (t) as a function of time t.

  FIG. 6A shows measurement data y (t) of a healthy person's blood glucose level used to obtain the parameter α which is an index of the glucose metabolism function, the horizontal axis represents time t, and the vertical axis represents y ( The values of t) are shown respectively. The blood glucose level measurement data y (t) is divided into a series of data until it reaches the minimum value after changing from the minimum value to the maximum value as indicated by the dotted frame in FIG. 6A. The start time t = 0 of each measurement data y (t) is the time at which the measurement data y (t) in each region first has a minimum value.

  FIG. 6B shows the logarithmic function ln {y (t)} of the measurement data y (t) shown in FIG. 6A, where the horizontal axis represents time t and the vertical axis represents the logarithmic function ln {y (t (t)). )} Values are shown respectively. According to FIG. 6B, it can be seen that the logarithmic function ln {y (t)} has a maximum value so as to be approximately synchronized with the measurement data y (t).

  FIG. 6C shows the parameter α obtained by differentiating the logarithmic function ln {y (t)} shown in FIG. 6B, the horizontal axis represents time t, and the vertical axis represents the value of the parameter α. Respectively. According to FIG. 6C, it can be seen that the parameter α of the healthy person repeatedly increases and decreases with time t, and shows a tendency to have a maximum value and a minimum value in each region. It can be seen that at least in each region, there are cases where the parameter α exceeds a certain value and cases where the parameter α falls below or below a certain value.

  FIG. 6D shows an input x (t) (= x1 (t) + x2 (t) estimated from the measurement data y (t) shown in FIG. 6A and the impulse response equation of the linear system obtained by approximation. )), The horizontal axis represents the time t, and the vertical axis represents the value of the input x (t). As shown in FIG. 6D, when the exponential term including one of the parameter α or the parameter β can be ignored in the equation (4), the value of the input x (t) can be estimated. For this reason, the diabetes index value calculation means 6 has a function of estimating an input x (t) that is an added value of the amount of meal and the amount of exercise, and the obtained input x (t) is used for advice on the amount of meal and the amount of exercise in the diagnosis of diabetes. It is also possible to make it possible to refer to such information for diagnosis support.

  Then, the parameter α and the input x (t) obtained by the diabetes index value calculation means 6 in this way are given to the diabetes index value determination means 7 as an index of the glucose metabolism function and useful information for diagnosis support. .

Further, when the parameter α is obtained by ignoring the exponent term including the parameter β in the equation (4), as shown in the part surrounded by the dotted line in FIG. in the received parameter α is the region, and there preset there may be the case, or the threshold epsilon _{[alpha] 1} or more than the threshold epsilon _{[alpha] 1,} and less than a certain preset threshold epsilon _{[alpha] 2} or the threshold epsilon _{[alpha] 2} or less Whether or not a case exists can be used as a criterion for determining whether the diabetes index value determination unit 7 determines that the measurement data y (t) belongs to a healthy person.

  Next, the operation of the blood sugar level data analysis system 1 will be described.

  7 obtains an index of the glucose metabolism function from the blood glucose level measurement data y (t) by the blood glucose level data analysis system 1 shown in FIG. 1 and determines whether the measurement data y (t) belongs to a healthy person. It is a flowchart which shows a procedure, The code | symbol which attached | subjected the number to S in the figure shows each step of a flowchart.

  First, in step S1, measurement data y (t) at each time t of the blood glucose level of a subject such as a diabetic subject or a person suspected of having diabetes is measured in advance, and the blood glucose level measurement data database is input from the input device 3. 5 is stored. The method of measuring the blood glucose level is arbitrary. For example, the blood glucose level measurement device and the blood glucose level data analysis system 1 are integrated, and the output of the blood glucose level measurement device is written and recorded in the blood glucose level measurement data database 5 as needed. You may make it do.

  Next, in step S 2, the measurement data y (t) at each time t of the blood glucose level of the subject stored in the blood glucose level measurement data database 5 is read by the diabetes index value calculation means 6.

  Next, in step S3, the diabetes index value calculation means 6 is derived from the linear system represented by the second-order lag element based on the blood glucose level measurement data y (t) read from the blood glucose level measurement data database 5. By numerically solving the nonlinear simultaneous equations composed of various calculation formulas, for example, formula (9) and formula (10), the parameters α, β and (α−β) at time t are obtained as indices of the sugar metabolism function. Then, the diabetes index value calculation means 6 gives the obtained parameters α, β and (α−β) to the diabetes index value determination means 7.

Next, in step S4, the diabetes index value determination means 7 uses the values of the parameters α, β and (α−β) received from the diabetes index value calculation means 6 as threshold values ε _α and ε that are set in advance. Compare with _β and ε _α-β . Then, diabetes index value determination means 7, the parameters alpha determines whether may be the case, or the threshold epsilon _alpha or exceeds the threshold epsilon _alpha exists. Also, the parameters beta determines whether the threshold epsilon _beta or less or less than that at all times. Further, it is determined whether or not the parameter (α−β) is _{equal to} or _larger than a certain threshold value εα _−β .

In step S4, the diabetes index value judging means 7 obtains the judgment results by comparing the parameters α, β and (α−β) with the thresholds ε _α , ε _β and ε _α-β, and sets the parameters α, β and ( The value is given to the monitor 2 together with the value of α−β) for display.

As a result, the diagnostician refers to the determination result by comparing each parameter α, β and (α−β) with the threshold values ε _α , ε _β and ε _α-β , so that the subject is diabetic or borderline. You can know if there is. For example, if each parameter α, β and (α−β) all satisfy the conditions indicated by the threshold values ε _α , ε _β and ε _α-β , the subject may be diagnosed as a healthy person. it can. In addition, when all or a part of each parameter α, β and (α−β) does not satisfy the conditions indicated by the threshold values ε _α , ε _β and ε _α-β , the subject is a diabetic patient or a boundary. It can be diagnosed as a type. In addition, the reference | standard whether it is a diabetic patient or a boundary type can be separately determined beforehand, for example empirically.

  Furthermore, the diagnostician can grasp the subject's glucose metabolism function numerically and specifically as parameters by confirming the values of the parameters α, β and (α−β).

  According to the blood glucose level data analysis system 1 as described above, a parameter that directly reflects the glucose metabolism function is specified by performing modeling assuming that the glucose metabolism function of a subject such as a diabetic patient is a linear system. It can be estimated as a numerical index. For this reason, it is possible to display medically-founded parameters, and the effectiveness of diabetes diagnosis support can be improved.

  In addition, a method for determining the severity of diabetes by calculating the rate of time change in blood glucose level immediately after a meal has been proposed. However, this method has a problem that the influence of the amount of meal and exercise amount increases. was there. In other words, if you eat a lot of food or if you exercise actively, there is a risk of an error that the time change rate of the blood glucose level becomes large, and if such an error occurs, There is a drawback that data processing such as measuring and correcting the amount of meal and the amount of exercise by another means is necessary. In addition, in the conventional diabetes diagnosis support method, it is difficult to obtain and display a parameter that reasonably reflects the biological information of the diabetic patient for diagnosis support.

  On the other hand, according to the blood sugar level data analysis system 1, since the amount of meal and the amount of exercise are not required when obtaining a parameter that is an index for diabetes diagnosis support, the amount of exercise and the amount of exercise are not greatly affected. For this reason, a more accurate index can be obtained as a parameter reflecting the subject's glucose metabolism function.

  Note that, in the blood glucose level data analysis system 1, the determination as to whether or not the subject is a diabetic patient may be made by the user only by obtaining a parameter that reflects the glucose metabolism function. Therefore, the blood glucose level data analysis system 1 may be configured by omitting the diabetes index value determination means 7.

  FIG. 8 is a functional block diagram showing a second embodiment of the blood sugar level data analysis system according to the present invention.

  In the blood sugar level data analyzing system 1A shown in FIG. 8, the sugar load test data database 10 and the system input estimating means 11 are provided, and the detailed functions of each component are the same as the blood sugar level data analyzing system 1 shown in FIG. Is different. Since other configurations and operations including a linear system represented by a second-order lag used for processing are not substantially different from the blood glucose level data analysis system 1 shown in FIG. 1, the same components are denoted by the same reference numerals. Description of similar functions is omitted.

  The blood sugar level data analysis system 1A allows the computer 4 equipped with the monitor 2 and the input device 3 to read the blood sugar level data analysis program, so that the glucose tolerance test data database 10, the blood sugar level measurement data database 5, the diabetes index value calculation means 6. Functions as diabetes index value judging means 7 and system input estimating means 11.

  In the glucose tolerance test data database 10, blood glucose level data at each time of a subject such as a patient or a examinee obtained by conducting a glucose tolerance test in advance is stored as glucose tolerance test data.

  In the blood glucose level measurement data database 5, blood glucose level measurement data at each time of a subject such as a patient or a examinee when a glucose tolerance test is not performed in advance is measured and stored by an arbitrary method.

  The diabetes index value calculation means 6 reads glucose tolerance test data, which is blood glucose level data obtained as a result of the glucose tolerance test, from the glucose tolerance test data database 10, and expresses the subject's glucose metabolism function as a second-order lag element. By fitting the sugar tolerance test data to a mathematical formula obtained by modeling as a linear system, a function of obtaining an index of the glucose metabolism function as a numerical value is provided. That is, the amount of sugar administered to the subject in the glucose tolerance test is defined as a constant amount (75 g), and the dietary amount x1 (t) that is input to the linear system is known as a numerical value.

  Accordingly, if the glucose tolerance test data is curve-fitted to the equation (7) of the impulse response y (t) of the linear system, that is, the curve of the impulse response y (t) shown in FIG. 4, the constants K, α, and β It can be estimated as an index of glucose metabolism function.

  The diabetes index value determining means 7 is an index of the glucose metabolism function estimated by the diabetes index value calculating means 6, for example, indices such as constants α, β, (α−β), etc. A function for determining whether the glucose tolerance test data is for a diabetic patient, a boundary type or a healthy person, and a determination result are given to the monitor 2 and displayed by a comparison method using the same threshold as the diabetes index value determination means 7 It has a function to make it. At this time, if necessary, the indicator of the sugar metabolism function itself can be given as a numerical value to the monitor 2 for display.

  The system input estimation means 11 is stored in the blood glucose level measurement data database 5 with a function for obtaining the transfer function of the linear system in the frequency domain using the constants K, α, and β estimated by the diabetes index value calculation means 6. It has a function of estimating the input x (t) of the linear system from the output of the linear system, which is blood glucose level measurement data in the real-time region, and the transfer function information.

That is, if the transfer function of the linear system is G, and the result of Laplace transform of each of the input x (t) and the output y (t) of the linear system is X and Y, the impulse response of the linear system is expressed by Equation (13). It is.
[Equation 13]
Y = GX (13)
Therefore, the input x (t) of the linear system is expressed by equation (14) from equation (13).
[Formula 14]
x (t) = L ⁻¹ (X) = L ⁻¹ (Y / G) (14)
However, L ⁻¹ indicates inverse Laplace transform.

  Therefore, if the product Y / G of the inverse 1 / G of the transfer function G and the Laplace transform Y of the blood glucose level measurement data y (t) is obtained, and Y / G is subjected to the inverse Laplace transform, the real time input x (t ) Can be estimated retroactively.

  Furthermore, the input x (t) is an added value of the amount of meal x1 (t) and the amount of exercise x2 (t). For this reason, the conversion coefficient between the amount of meal x1 (t) converted into sugar and the amount of calories in the meal, and the conversion coefficient between the amount of exercise load x2 (t) converted into the amount of sugar and the amount of calories of exercise are known. In some cases, the meal amount x1 (t) and the exercise amount x2 (t) can be expressed as familiar amounts such as the calorie amount of the meal and the calorie amount of the exercise.

  The system input estimation unit 11 is provided with a function for estimating all or part of the meal amount x1 (t), the exercise amount x2 (t), the calorie amount of the meal, and the calorie amount of the exercise as necessary. Information such as the input x (t) of the linear system, the amount of meal x1 (t), the amount of exercise x2 (t), the amount of calorie of the meal, and the amount of calorie of the exercise estimated by the system input estimating means 11 is displayed on the monitor 2. Configured to be given and displayable.

  Next, the operation of the blood sugar level data analysis system 1A will be described.

  FIG. 9 shows a procedure for obtaining an index of the glucose metabolism function from the glucose tolerance test data and estimating the input x (t) from the measurement data y (t) of the blood glucose level by the blood glucose level data analysis system 1A shown in FIG. It is a flowchart, The code | symbol which attached | subjected the number to S in the figure shows each step of a flowchart.

  First, in step S10, a subject's glucose tolerance test is performed in advance, and blood glucose level data obtained as a result of the glucose tolerance test is input to the blood glucose level data analysis system 1A via the input device 3. And it is preserve | saved in the glucose tolerance test data database 10 as glucose tolerance test data.

  Next, in step S <b> 11, the diabetes index value calculation means 6 reads glucose tolerance test data from the glucose tolerance test data database 10.

  Next, in step S12, the diabetes index value calculation means 6 performs the curve fitting of the glucose tolerance test data to the curve of the impulse response y (t) of the linear system shown in FIG. , Α, and β are estimated as indicators of glucose metabolism function. Then, the diabetes index value calculation means 6 gives the estimated constants K, α, and β to the diabetes index value determination means 7 and the system input estimation means 11.

  Next, in step S13, the diabetes index value determination means 7 performs a comparison determination between the parameters α, β and (α−β) and the threshold values from the constants α, β received from the diabetes index value calculation means 6.

  Next, in step S14, the diabetes index value determining means 7 is displayed by giving the monitor 2 the result of comparison determination with the threshold values of the parameters α, β and (α−β). As a result, the diagnostician can know whether or not the subject is diabetic or borderline by referring to the determination result by comparison with the threshold values of the parameters α, β and (α−β).

  In step S15, blood glucose level measurement data y (t) at each time t of the subject is measured at an arbitrary timing and stored in the blood glucose level measurement data database 5 from the input device 3.

  Next, in step S <b> 16, the system input estimation unit 11 reads blood glucose level measurement data y (t) from the blood glucose level measurement data database 5.

  Next, in step S17, the system input estimation unit 11 obtains the inverse 1 / G of the transfer function G in the linear system using the constants K, α, and β received from the diabetes index value calculation unit 6. Further, the system input estimation means 11 obtains the product Y / G of the inverse 1 / G of the transfer function G and the Laplace transform Y of the blood glucose level measurement data y (t) read from the blood glucose level measurement data database 5. Then, the system input estimation means 11 estimates the input x (t) in real time retroactively by performing inverse Laplace transform on Y / G as shown in the equation (14).

  Further, if necessary, the whole or a part of the amount of meal x1 (t), the amount of exercise x2 (t), the amount of calorie of the meal, and the amount of calorie of the exercise is estimated from the input x (t).

  Next, in step S18, the system input estimation means 11 obtains all or part of the estimated input x (t), meal amount x1 (t), exercise amount x2 (t), meal calorie amount, and exercise calorie amount. It is displayed by giving to the monitor 2.

  As a result, the diagnostician limits the amount of meals from the information displayed on the monitor 2 to the extent of the meal quantity at time t between the latest meal quantity input x (t) and the next meal quantity input x (t). It is possible to rationally derive a specific numerical target such as how to increase the amount of exercise or how much exercise should be performed from the standpoint of management such as diet therapy or insulin administration.

  According to the blood glucose level data analysis system 1A as described above, in addition to the same effect as the blood glucose level data analysis system 1 shown in FIG. 1, the gain constant K and the time constant T1 (= 1/1 /) in the linear system are obtained from the glucose tolerance test data. Since α) and T2 (= 1 / β) can be estimated, a transfer function (or ordinary differential equation) governing the input / output of the linear system can be obtained. As a result, the past input x (t), that is, the past meal amount x1 (t) and the past exercise amount x2 (t) is estimated from the blood glucose level measurement data y (t) in the past. be able to. Then, by the next meal time, a standard required for the meal amount x1 (t) and the exercise amount x2 (t) can be shown as a numerical value.

1 is a functional block diagram showing a first embodiment of a blood sugar level data analysis system according to the present invention. The figure which shows the time change of the blood glucose level and insulin level in the glucose tolerance test of a healthy person, a type 1 diabetes patient, and a type 2 diabetes patient. The block diagram of the linear system used when calculating | requiring the parameter | index of a glucose metabolism function as a numerical value with the blood glucose level data analysis system shown in FIG. The figure which plotted the impulse response of the linear system used when calculating | requiring the parameter | index of a glucose metabolism function as a numerical value with the blood glucose level data analysis system shown in FIG. Expressions (9) and (10) derived from a linear system used when calculating an index of glucose metabolism function as a numerical value by the blood sugar level data analysis system shown in FIG. 1 and blood sugar level measurement data y (t) FIG. 3 is a graph in which parameters α, β and (α−β) obtained as indices of sugar metabolism function are plotted as a function of time t. The blood glucose level data analysis system shown in FIG. 1 derives from the linear system used when obtaining an index of the glucose metabolism function as a numerical value, and the blood glucose level measurement data y derived from the equation (12) derived from the fact that the time constant difference is sufficiently large. The figure which graphed the parameter (alpha) calculated | required as a parameter | index of glucose metabolism function from (t), and the estimated input x (t) as a function of time t. The flowchart which shows the procedure which calculates | requires whether the measurement data y (t) is a healthy subject while calculating | requiring the parameter | index of glucose metabolism function from the measurement data y (t) of a blood glucose level by the blood glucose level data analysis system shown in FIG. The functional block diagram which shows 2nd Embodiment of the blood glucose level data analysis system which concerns on this invention. The flowchart which shows the procedure which calculates | requires the parameter | index of glucose metabolism function from the glucose tolerance test data, and estimates input x (t) from the measurement data y (t) of a blood glucose level by the blood glucose level data analysis system A shown in FIG. The figure which shows the standard of diabetes diagnosis in the conventional glucose tolerance test.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1, 1A Blood glucose level data analysis system 2 Monitor 3 Input device 4 Computer 5 Blood glucose level measurement data database 6 Diabetes index value calculation means 7 Diabetes index value determination means 10 Glucose tolerance test data database 11 System input estimation means

Claims (8)

A blood glucose level measurement data database for storing blood glucose level measurement data at each time of the subject;
The subject 's sugar is a linear system in which at least one of the subject's dietary amount and exercise amount is an impulse input, and an impulse response y (t) to the impulse input is represented by a second-order lag element representing a temporal change in blood glucose level. A means for obtaining an index of the subject's glucose metabolism function based on a mathematical expression modeling the metabolic function, the time constant of the mathematical expression obtained based on the measurement data read from the blood glucose level measurement data database Diabetes index value calculating means for outputting a parameter as a numerical value of an index representing the glucose metabolism function;
A blood glucose level data analysis system comprising: A glucose tolerance test data database for storing blood glucose measurement data obtained by the subject's glucose tolerance test as glucose tolerance test data;
At least one of the subject's meal amount and exercise amount is set as an impulse input, and the impulse response y (t) to the impulse input is expressed by a second-order lag element representing a time change of blood glucose level. A means for obtaining an index of the subject's glucose metabolism function based on a mathematical expression modeling the glucose metabolism function, wherein the glucose tolerance test data read from the glucose tolerance test data database is converted into the impulse response y (t). Diabetes index value calculating means for outputting a parameter related to the time constant of the mathematical formula obtained by fitting as a numerical value of an index representing the glucose metabolism function;
A blood glucose level data analysis system comprising: In the above formula, the time change of the blood sugar level when at least one of the meal amount and the exercise amount is an impulse input is represented by the following impulse response y (t):

It is an impulse response equation when assumed to be expressed by
Diabetes index value calculation means , the target from the measurement data and impulse response equation read from the blood glucose level measurement data database, or from the glucose tolerance test data and impulse response equation read from the glucose tolerance test data database Calculating and outputting an index of a person's glucose metabolism function as numerical values α and β in the impulse response ,
The blood glucose level data analysis system according to claim 1 or 2, wherein The blood glucose level data analysis system according to claim 1 or 2, further comprising a diabetes index value determination unit that compares and determines the index obtained by the diabetes index value calculation unit with a preset threshold value. The input of the linear system is estimated from a blood glucose level measurement data database storing blood glucose level measurement data at each time of the subject, the blood glucose level measurement data read from the blood glucose level measurement data database, and the index. The blood sugar level data analysis system according to claim 2, further comprising system input estimation means. Computer
Means for reading the measurement data from a blood glucose level measurement data database for storing blood glucose level measurement data at each time of the subject ;

The subject's sugar is expressed by a linear system in which at least one of the subject's meal amount and exercise amount is an impulse input, and the impulse response y (t) to the impulse input is represented by a second-order lag element representing a temporal change in blood glucose level. A means for obtaining an index of the subject's glucose metabolism function based on a mathematical expression modeling the metabolic function, the time constant of the mathematical expression obtained based on the measurement data read from the blood glucose level measurement data database Diabetes index value calculating means for outputting a parameter as a numerical value of an index representing the glucose metabolism function;
A blood glucose level data analysis program characterized by being made to function as. Computer
Means for reading the glucose tolerance test data from a glucose tolerance test data database for storing blood glucose measurement data obtained by the subject's glucose tolerance test as glucose tolerance test data ;
At least one of the subject's meal amount and exercise amount is set as an impulse input, and the impulse response y (t) to the impulse input is expressed by a second-order lag element representing a time change of blood glucose level. A means for obtaining an index of the subject's glucose metabolism function based on a mathematical expression modeling the glucose metabolism function, wherein the glucose tolerance test data read from the glucose tolerance test data database is converted into the impulse response y (t). Diabetes index value calculating means for outputting a parameter related to the time constant of the mathematical formula obtained by fitting as a numerical value of an index representing the glucose metabolism function;
A blood glucose level data analysis program characterized by being made to function as. In the above formula, the time change of the blood sugar level when at least one of the meal amount and the exercise amount is an impulse input is represented by the following impulse response y (t):

It is an impulse response equation when assumed to be expressed by
Diabetes index value calculation means , the target from the measurement data and impulse response equation read from the blood glucose level measurement data database, or from the glucose tolerance test data and impulse response equation read from the glucose tolerance test data database obtains and outputs an indication of glucose metabolism in persons as a number α and β in the impulse response,
The blood sugar level data analysis program according to claim 6 or 7, wherein

Images