JPH0792407B2 - Electronic thermometer - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (a) Field of Industrial Application The present invention is an electronic thermometer of a speculative type, in which the response value of a response curve is reflected in the calculation of the guess value so as to be able to cope with individual differences. For electronic thermometers.

(B) Conventional technology A guess-type electronic clinical thermometer usually measures temperature for a certain period of time, and based on changes in a plurality of obtained temperature data, a convergence temperature,
That is, the body temperature estimation value is calculated, and the estimation value is sequentially updated and displayed on the display to reduce the measurement time.

This guess value is calculated by a guess formula preset in the CPU (Central Processing Unit).

For this estimation formula, for example, the following formula is used.

S (t) = T + K · dT / dt In this equation, T is the detected temperature at the time when t hours have elapsed from the start of measurement, and the detected temperature T, the measurement time t, the increase rate dT / dt of the detected temperature and the multiplier. K is a calculation factor.

By the way, it is known that the response curve of the detected temperature with respect to time varies depending on the person to be measured. For example, even if the convergence temperatures are almost the same, the response curves may be completely different. However, in the conventional estimation calculation method, the multiplier (K) is calculated as a fixed fixed value. That is, the response values of different response curves are uniformly captured depending on the person to be measured, and the estimation is performed only by the rate of increase at that point in addition to the measurement time and the detected temperature at that point.

Therefore, when the response curve is gentle, the temperature rise rate is small, so the guess value becomes lower than the convergence temperature, that is, the body temperature, and there is a problem that the guess accuracy is not appropriate.

Therefore, an electronic clinical thermometer has been proposed in which the idea of the response curve is incorporated into the calculation element for the estimation value calculation, and the change characteristic of the response curve is reflected in the estimation value (Japanese Patent Application Laid-Open No. 57-57.
-28225 publication).

This thermometer uses the following formula to calculate the guess (Tm).

Tm = Tq + K · dT / dt In this equation, Tq is the measured temperature, K is the measured temperature multiplier,
dT / dt is the rate of temperature rise.

The feature of this estimation method is that the temperature measurement multiplier K, which was conventionally fixedly and uniformly regarded as a constant multiplier, has different values in advance.
Set the type and select from three types according to the response curve of the person to be measured.
The addition value (correction value) is determined by selecting one of the multipliers K from K ₁ , K ₂ and K ₃ . According to this method, it is possible to reflect the difference in the response curve, which differs depending on the person to be measured, in the estimated value, as compared with the method in which the multiplier K is a fixed constant.

(C) Problems to be Solved by the Invention In the estimation method of the electronic thermometer proposed above, three different temperature measurement multipliers are set in advance, and the response among the three multipliers is set according to the response curve. It selects the multiplier corresponding to the value.

However, since this estimation method is a method in which the multipliers of K ₁ , K ₂ , and K ₃ are selected by the absolute value of dT / dt, there is a disadvantage that it is easily affected by the ambient temperature. Also, fixed constants K ₁ , K ₂ ,
With K ₃ , there is a problem that the response curve changes with the passage of time when performing continuous estimation, and the value is not an appropriate value. Further, there are various disadvantages such that the three types of fixed constants cannot sufficiently cope with variations in individual differences.

The present invention solves the above problems of the conventional one, reflects the change characteristic of the response curve in the guess value, and provides an electronic thermometer that can obtain a guess value with high measurement accuracy corresponding to individual differences. The purpose is to do.

(D) Means and Actions for Solving Problems In order to achieve this object, the electronic thermometer of the present invention has the following configuration.

The electronic clinical thermometer includes temperature measuring means for measuring the temperature, a memory for storing the measured values detected by the temperature measuring means for a plurality of samplings, a correction function for approximating with a quadratic equation of time, and a time derivative of the temperature. An inference value calculation means for obtaining a body temperature estimation value by adding an actual measurement value of the body temperature to the product of, and a display means for displaying the latest measurement value in the memory or the estimation value obtained by the estimation value calculation means, Element determining means for determining the coefficient of the correction function based on a plurality of actual measurement values including the latest actual measurement value stored in the memory, and a plurality of actual measurements including the latest actual measurement value stored in the memory Characteristic estimation means for obtaining the time differential value based on a value, and the guess value computing means includes the coefficient obtained by the element determining means, the time differential value of the temperature obtained by the characteristic extracting means, and the memory. And so as to obtain the inferred value and a measured value of Li.

In the electronic thermometer having such a configuration, the actual measurement value is detected by the temperature measuring means, and the actual measurement value is sequentially stored in the memory, while the response curve is analyzed and the change characteristic of the parameter P, that is, the detected temperature with respect to time, is calculated. It is extracted as a ratio of change rate or a ratio of change of change rate. Then, based on this extracted value (parameter P), conventionally, a fixed multiplier (a, b,
The estimation element S (t) is calculated for the first time by individually determining the calculation element described as c) and calculating the correction function value h (t) by this calculation element.

Thus, since the guess value is calculated based on the response curve, the response curve that differs depending on the measurement subject is directly reflected in the calculated guess value, and an extremely accurate guess value can be calculated.

(E) Embodiment FIG. 2 is a block diagram showing a specific circuit configuration example of the electronic thermometer according to the present invention.

A temperature sensor 1 such as a thermistor is exposed at the tapered tip of the instrument, and the temperature sensor 1 detects the temperature of the under tongue or the armpit which comes into contact with the temperature sensor 1 and takes out an electric signal according to the temperature.
The temperature information (analog amount) detected at any time is A /
The digital value is converted by the D converter 2 into a digital value that can be easily processed by the CPU 3, and is taken into the CPU 3.

When the power switch 4 is turned on, the CPU 3 activates the temperature sensor 1 to store temperature information sent from time to time in the memory 5 and display it on the display 6. Further, the CPU 3 has a parameter P calculation function for calculating a parameter P by analyzing a temperature rise curve (response curve) for calculating an estimated value based on this temperature information, and an estimated value calculation based on this parameter P. A calculation element determination function for determining the calculation element (multiplier) to be used, a correction function calculation function for obtaining the correction function value h (t) of the response curve based on this calculation element, and the calculation element is substituted into a certain mathematical expression. However, a guess value calculation function for calculating the guess value is provided.

The calculation of the parameter P, the calculation of the calculation element determined by the parameter P, the calculation of the correction function calculated based on the calculation element, and the estimation value calculation formula are explained by the following mathematical formula theory.

Usually, the measurement time and the temperature rise curve when measured with a thermometer are represented by the following equation.

% Temp In this equation, τ _a is a constant during the thermal response of the thermometer's temperature sensitive part, t ₀₀ , t ₁₁ are parameters related to heat conduction between the body surface and the thermometer, heat capacity, etc., and τ _a , t _co are deep human body parts. It is a parameter related to heat conduction between the body and the body surface.

Therefore, for example, for each response example shown in FIG. 3, when each parameter of the equation (1) is obtained, τ _a = 1.83, t ₀₀ = −5.77, t ₁₁ = −3.14, τ _c = 464.7, t _co = −1147.9, and as shown in FIG. 4, τ _a is dominant when 0 ≦ t ≦ 10 seconds, and t ₀₀ and t ₁₁ are dominant when 10 ≦ t ≦ 240 seconds. It can be seen that τ _c and t _co are dominant when 240 ≦ t.

Here, in the prior art, when the estimated value S (t) is obtained,
It was calculated by S (t) = T _q + K ₁ · dT _q / dt. In this equation, T _q is the measured temperature and K ₁ is the temperature measurement constant.

Therefore, the estimated value S (t) is changed to S (t) based on the equation (1).
= T + h (t) · dT / dt Dashi That is, the estimated value S (t) is expressed as S (t) = T + h (t) · dT / dt (3).

In the formula (2), when t> 10 However, Becomes

When this equation (4) is approximated by a polynomial, K ₀₀ = 100. Becomes In this formula, a _{0 to} a ₂ are constants. t ₀₀
Is a parameter related to heat conduction and heat capacity of the body surface and the thermometer, and is characteristically determined by individual differences.

FIG. 5 is an example of calculating the correction function of the equation (2) with respect to the response example of FIG. Here, when the approximation by the quadratic equation is performed with t ≦ 60 seconds, when 60 ≦ t ≦ 200, h ₁ (t) = − 0.0054 (t-200) ² +255 (6) and 200
When ≤t≤600, h ₂ (t) =-0.0014 (t-200) ² +255 (7).
However, when h ₁ (t) and h ₂ (t) <0, h ₁ (t) = 0, h ₂
(T) = 0.

It should be noted that the approximation may use a polynomial of second or higher degree or a table function.

Further, when the estimated value S (t) is calculated by the equation (3) using the correction function approximate values of the equations (6) and (7), the following table is obtained.

FIG. 6 shows an example of calculating the guess value of this response curve.
As shown in the table above, when the guess value is calculated using h ₁ (t) and h ₂ (t), it is possible to calculate with an error of ± 0.01 ° C for the value of 600 seconds, which is practically It can be seen that there is no problem even in the approximation using the quadratic formula. Also, for simplification, h ₁ (t)
Even if only used, accuracy does not matter up to about 240 seconds. As can be seen from the above table, when only h ₁ (t) is used, the temperature shows a gradual decrease from 240 seconds to 420 seconds, and the temperature rises again after 420 seconds and then increases at 600 seconds.
_The temperature rises to 37.27 ° C, which is the same as when using ₁ (t) and h ₂ (t) (see Fig. 6). Therefore, for up to 240 seconds, sufficient accuracy can be obtained with a simple method using only h ₁ (t). Originally, in a guess-type electronic thermometer that knows the saturation value in a short time, it does not make sense to measure for a long time, and when measuring for a long time, actual measurement is more reliable, so it is preferable in practice I can say.

Next, the theory of analyzing the response curve and calculating the response value (parameter P) will be described. The approximation of the correction function of equation (5) is t
It is determined by the parameter of ₀₀ .

Therefore, considering the means for extracting the parameter t ₀₀ from the equation (1) next, when t> 10 from the equation (1), However, Becomes Furthermore, when this equation (8) is differentiated by t, Becomes Here, considering an approximate value of three points of t = t _a , t _b , and t _c (t _a <t _b <t _c ), from τ _c > 400, And the above equation (9) is Can be approximated by

Thus, from (Ta-Tb) ÷ (Tb-Tc), Is obtained. Here, t _a, t _b, taking the t _c at equal intervals of _{Δt, t a ...... (15)} t b = t a + Δt ...... (16) t c = t a + 2Δt ...... (17) becomes , Substituting these into equation (14), Is obtained. This equation (18) is approximately And by this, Becomes Here, it can be seen that t ₀₀ is related to the ratio of the difference in the temperature change rates at the three points.

By the way, this equation can be approximated as the following equation, because the value that can be taken in actual data is limited.

Here, T _ab / T _bc can be calculated if there are temperature data of 3 points at equal intervals and at least 1 point of each point.
For example, the three points a, b, and c shown in FIG.
From -1, a + 1, b-1, b + 1, c-1, c + 1, Then, if we take a ₋₁ −a ₋₁ = b ₋₁ −b ₋₁ = c ₋₁ −c ₋₁ = Δt ′, then Becomes Further, Δt ′ is set so that a ₊₁ = b ₋₁ and b ₊₁ = c _−1.
If you take, it becomes the data of four points T ₀ , T ₁ , T ₂ , T ₃ , Further, for simplification, if T ₂ = T ₁ = T ₁₂ is taken, then 3 point data are obtained. Here, as a result of analyzing many actual data, (2
The correlation as shown in the following equation was obtained between the parameters of equation (7) and the correction function equation.

Temperature data of 3 points of t = 10, 20, 30 seconds are T ₁₀ , T ₂₀ , T ₃₀
Then, the parameter P is given by the following equation.

Further, the correction function h (t) is expressed by the following equation.

h (t) = a (t + b) ² + c (29) Further, multipliers a, b, and c, which are calculation elements, are obtained by the following equation.

a = -0.00282 + 0.0024 · P …… (30) b = −320 + 110 · P …… (31) c = 192−120 · P …… (32) And the estimated value S (t) is .

Based on the above mathematical formula theory, the above-mentioned estimation value calculation formula was obtained.

Thus, the analysis of the temperature curve (response curve) performed to calculate the guess value is performed after the response of the temperature sensing part is saturated, so it is less affected by the ambient temperature, and the parameter of equation (28) is , It is peculiar to each individual, and by introducing this parameter to the estimation value calculation, it is possible to deal with individual differences.

A buzzer 7 and a power source 8 are connected to the CPU 3, respectively, and the buzzer 7 is activated when the estimated value S (t) reaches a substantially constant value, and is set to give a notification to that effect.

However, the hardware configuration shown in FIG. 2 is the same as that of a general electronic thermometer known in the related art. The electronic thermometer of this embodiment is characterized by the functional configuration of the CPU 3.

Therefore, the software configuration and operation of the electronic thermometer of this embodiment will be described below with reference to the flow shown in FIG.

When the power switch 4 is turned on, the start flag, the state flag and the timer are initialized, and the instrument is initialized [Step (hereinafter referred to as “ST”)
1].

The start flag is a flag for determining whether or not the sensor temperature has risen after the power switch 4 is turned on and a temperature detection start state is detected.

The state flag is a flag for selecting whether to execute the calculation process of the parameter P or the calculation process of the estimated value after the actual measurement is started.

Further, the timer is used to measure the time from the start of measurement when the power switch 4 is turned on to the end of measurement.

After the power switch 4 is turned on, in the next ST2, it is determined whether or not it is the sampling time.

In the embodiment, the measured temperature is measured every 1 second. Therefore, the sensor 1 sends the current temperature information to CP every sampling time.
While being sent to U3, the measured values that gradually rise are updated and stored in the memory 5, while the following operations are performed at each sampling time, and the process returns to ST2.

If the sampling time has come, this ST2
The determination is “YES”, and the time t from the start of measurement is incremented by 1 for each sampling time (ST3).
Then, the current temperature Tt is detected (ST4).

In ST5, it is determined whether or not the current time count is an odd number. If it is not an odd number of seconds, this ST5 judgment is "N
It becomes "O" and proceeds to ST7. Conversely, if it is an odd number of seconds, the judgment is "YE
S ”, and the measured value is updated and recorded in the memory 5 in the next ST6 (ST6). That is, in the embodiment, the recording of the temperature data is set to be executed only for an odd number of seconds.

Further, in the embodiment, the measured values (T ₀ , T ₁ , T ₂ ,
T ₃₎ is adapted to always be stored and held in the register (ST6). Therefore, in ST7, the actual measured value Tt this time is
It is determined whether or not it is higher than T ₃ stored in the memory 5, and when it is higher, T ₃ is set as T _2, and the actual measurement value Tt this time is stored as T ₃ (ST8). This process is executed every odd seconds.

In ST9, it is determined what state the start flag is in. Now, if the start flag is 0, that is, the power switch 4 is only turned on, the detected temperature is still at room temperature, and the start of temperature detection is not detected, the next ST
At 10, it is determined whether or not the difference between the current measured value T ₃ and the previous measured value T ₂ is higher than a predetermined value (for example, a constant value of 1 ° C. or a temperature rise rate corresponding to 1 ° C.). Of course, in this case, the temperature detection is not started, so the determination in ST10 is "NO", and the process returns to ST2.

Now, thermometry is started actually measured value Tt rises, the timer count of the rise value is assumed to be odd seconds, any determination of ST5 and ST7 be "YES", and the memory the measured value as the peak value T ₃ It is stored in 5 (ST7). Here, the start flag is still 0. Therefore, S via ST9
The process proceeds to T10, and the determination in ST10 is "YES", and the start flag 0 is set to 1 in ST11. That is, it is determined here that the temperature measurement has started, the timer starting from the time point of the temperature measurement is set to 0 again (ST11), and thereafter, the same temperature measurement is repeatedly performed as the timer measures.

Then, at the subsequent sampling time, the start flag is 1, so the process proceeds from ST9 to ST13. ST1
At 3, it is determined what state the state flag is in. The state flag is 0 now. That is, since 40 seconds have not elapsed since the start of actual measurement, the temperature data for calculating the estimated value is not detected. Therefore,
Move to ST14.

ST14 to ST16 analyze the response curve as shown in FIG. 8 and execute the processing operation for calculating the parameter P. In ST14, it is determined whether or not the time is t _a , that is, for example, 20 seconds, and if so, the determination in ST14 is “YE
S ", the detected temperature T ₃ at this point (36.31 ℃ in Fig. 8)
Is Ta (ST17). Then, this Ta (36.31 ° C.) is displayed on the display 6 (ST12), and the process returns to ST2. Then, similarly, when the time t _b, that is, 30 seconds elapses (10 seconds elapses from the time point t _a ), the determination in ST15 is “YES”, and the detected temperature at this point is the same.
T ₃ a (36.49 ℃) and Tb (ST18). Furthermore, if time t _c, that is, 40 seconds has elapsed, the determination in ST16 is “YES”, and the detected temperature T ₃ (36.60 ° C.) at this point is T c (S
T19).

Here, the temperature information for calculating the parameter P is obtained. Therefore, in ST20, the parameter P is calculated to be 0.61 by substituting this data (Ta, Tb, Tc) into the above-described formula for calculating the parameter P (ST20).
That is, since the temperature curve is analyzed by the multipoint ratio, the response curve of the individual difference can be dealt with, and the multipoint ratio can prevent the adverse effect of the ambient temperature.

After that, the state flag is set to 1 (ST21), the peak value is displayed on the display 6 (ST12), and the process returns to ST2.

At the next sampling time, 40 seconds have passed since the actual measurement started, and it is possible to estimate. That is, the state flag is 1.

Therefore, the determination in ST13 is "YES", and the process proceeds to ST22. ST22
Then, it is determined whether or not the current sampling time is an odd number of seconds.

In this embodiment, similarly to ST5, the estimation value is set to be calculated only for the measured value (peak value) of odd seconds.

Now assuming that the sampling time is an odd second, this
ST22 of the determination is "YES", and the process proceeds to ST23, firstly wherein the temperature increase rate dT / dt is calculated by T ₃ -T _0/6.

After that, a multiplier a, which is a calculation element for the estimation value calculation,
b and c are determined from the parameter P (ST24). That is, it is determined by the above equations (30) to (32). When the multipliers a, b, and c are obtained, the correction function value h
Obtain (t) (ST25). The correction function h (t) is calculated by substituting a, b, and c in the above equation (29). That is, in the embodiment, since a multi-order time function is used as the correction coefficient, an appropriate value can be obtained even when the continuous estimation is used.

Here, all the calculation factors for the calculation of the guess value have been decided. Thus, the estimated value S (t) is calculated based on the above equation (33) (ST26), and this estimated value is displayed on the display 6 (ST27).

The above-mentioned estimation calculation is executed at every odd sampling time. Although not shown in this flowchart, the buzzer 7 notifies that the estimated value displayed here is highly accurate and can be trusted. After that, the power supply switch 4 is turned off to complete the measurement.

(F) Effect of the Invention In the present invention, as described above, the response curve is analyzed to calculate the parameter P, the calculation element for calculating the estimation value is determined from the parameter P, and the correction function is calculated from this calculation element. It was decided to obtain the value and calculate the guess value.

According to the present invention, the factor for calculating the estimated value is determined corresponding to the change characteristic of the detected temperature with respect to time. Therefore, when the change characteristics are different, the calculation elements are different, so that the change characteristics are reflected in the guess value, and more accurate guess calculation corresponding to the individual difference can be performed.

Further, in the present invention, the multipliers a, b, and
Since c is calculated and the correction function h (t) of time is calculated from this, it is possible to obtain an appropriate value by continuous estimation. Further, in the present invention, since the multipoint ratio is used for the analysis of the response curve, there is little effect of the ambient temperature, and more accurate measurement can be performed. .

[Brief description of drawings]

FIG. 1 is a flow chart showing the processing operation of the electronic thermometer of the embodiment, FIG. 2 is a block diagram showing an example of the circuit configuration of the electronic thermometer, FIG. 3 is a response curve diagram of the detected temperature with respect to time, and FIG. Is a response curve diagram based on 100% ratio in FIG. 3, FIG. 5 is an explanatory diagram showing correction function values for the response curve of FIG. 3, and FIG. 6 is a guess value for the response curve of FIG. FIG. 7 is an explanatory view showing a calculation example, FIG. 7 is an explanatory view showing a calculation example for obtaining the parameter P, and FIG. 8 is an explanatory view showing a concrete calculation example of the parameter P. 1: Temperature sensor, 3: CPU, 5: Memory, 6: Display.

Continuation of front page (56) Reference JP-A-60-157031 (JP, A) JP-A-60-147621 (JP, A) JP-A-56-46440 (JP, A)

Claims (1)

[Claims] 1. A temperature measuring means for measuring a temperature, a memory for storing measured values detected by the temperature measuring means for a plurality of samplings, a correction function for approximating a quadratic expression of time, and a temperature time. Estimated value calculation means for obtaining a body temperature estimated value by adding the measured value of the body temperature to the product of the differential,
Display means for displaying the latest measured value in the memory or the guess value obtained by the guess value calculation means, and the correction function based on a plurality of measured values including the latest measured value stored in the memory Element determining means for determining the coefficient of
And a characteristic extraction unit that obtains the time differential value based on a plurality of actual measurement values including the latest actual measurement value stored in the memory, and the guess value calculation unit is a coefficient determined by the element determination unit. An electronic thermometer, wherein the estimated value is obtained from the time differential value of the temperature obtained by the characteristic extracting means and the actual measurement value of the memory.