JPH0656332B2 - Electronic thermometer - Google Patents

Description

TECHNICAL FIELD The present invention relates to an electronic clinical thermometer, and more particularly to a temperature predicting type electronic clinical thermometer.

[Prior Art] Conventionally, this type of electronic thermometer has a predictive formula that completely prescribes a temperature rise curve in advance, and
The equilibrium temperature was displayed in advance by adding the additional amount obtained by the prediction formula. Therefore, each coefficient (parameter) in the prediction formula has the statistically most predictive error by performing statistical processing of the actual measurement value by the temperature probe used for actual measurement in the manufacturing process of each electronic thermometer, for example. It had to be set to a small value.

By the way, it is known that there are individual differences in the temperature rise curve, and that there are considerable differences between the armpit and the oral temperature measurement even for the same person. In such a case, even if a single prediction formula that corrects for variations in the thermal characteristics of the probe is provided, it is not possible to actually display the accurate equilibrium temperature in advance.

The electronic thermometer disclosed in JP-A-58-225326 solves this problem by providing a plurality of prediction formulas. That is, a plurality of prediction formulas statistically determined based on a large amount of measurement results are provided in advance, and at the time of measurement, condition setting (prediction) is performed by comparing the rising curve of the measured temperature with the selected prediction formula. The parameter of the formula) is changed by trial and error to solve the problem. However, since a plurality of prediction formulas must be defined in advance, as a practical problem, the complexity of adjustment due to variations in the thermal characteristics of the temperature probe that occurs during mass production cannot be avoided. Further, in order to improve the prediction accuracy, a large number of prediction formulas having different characteristics must be provided. Also,
If the selection of the prediction formula is not appropriate near the rise of temperature, the predicted value may overshoot.

The electronic thermometer disclosed in Japanese Patent Laid-Open No. 59-187233 solves the above-mentioned problem by assembling a predictive formula that fits the actual rising curve of the measured temperature. That is, noting that there is a linear relationship (T _L = A−τ′t) between the logarithmic value _TL of the time derivative of the measured body temperature and the sampling time t, the constants A and τ ′ are regressed. Seeking by law. However, since the logarithmic value T _L is not the measured body temperature data itself, an error due to the differential calculation and the logarithmic calculation is introduced, and the error affects the determination of the constants A and τ ′ at a large rate. Also, when the measured body temperature data contains a noise component,
As a result, the noise component exponentially affects the prediction result, giving a very unstable prediction result.

[Problems to be Solved by the Invention] The present invention eliminates the above-mentioned drawbacks of the prior art. The object of the present invention is to provide a temperature rise curve due to variations in the thermal characteristics of the probe or differences in individuals and measurement sites. The object is to provide an electronic clinical thermometer that can always display an accurate preceding display even if there is a difference.

Another object of the present invention is to provide an electronic clinical thermometer that can obtain a stable predicted transition even if the detected temperature includes a noise component.

Another object of the present invention is to provide an electronic thermometer that accurately predicts a temperature measurement value at an arbitrary future time.

Another object of the present invention is to provide an electronic thermometer that accurately predicts a thermal equilibrium temperature value after a very long time in the future.

[Means for Solving Problems] In order to achieve the above object, the electronic thermometer according to the present invention is
Temperature detection means for detecting the temperature and generating temperature data indicating the temperature, time signal generation means for measuring the elapsed time after the start of measurement and generating time data ti indicating the elapsed time, and the detected temperature Storage means for storing data in correlation with the time data t i at the time of detection, extraction means for extracting a plurality of temperature data from the storage means, and the extracted temperature data as an objective variable and correlated with this Functions 1 / t i, 1 / ti ² , of time data t i,
1 / ti ^3, ..., by regression analysis of the function to be described variable 1 / ti ^n, the prediction calculation and signal analyzing means for determining the parameters of the prediction function, the temperature at a future time by the prediction function specified by the determined parameters And a prediction calculation means for performing the calculation.

Further preferably, the extracting means, when extracting the plurality of temperature data, the time interval of the time data correlated with the plurality of temperature data after the time interval of the time data correlated with the plurality of previous temperature data. One of the modes is to take out so as to include the larger one.

The electronic thermometer according to the present invention detects the temperature and generates temperature data T ₀ indicating the temperature, and the time data ti indicating the elapsed time by measuring the elapsed time after the start of measurement. And a temperature signal generating means for controlling the detected temperature data T ₀
storage means for storing correlation with i, extraction means for extracting a plurality of temperature data T ₀ (ti) from the storage means, and the extracted temperature data T ₀ (ti) as an objective variable, and correlation with this Function of time data ti 1 / ti, 1
/ Ti ² , 1 / ti ³ , ..., 1 / ti ⁿ as an explanatory variable, T ₀ (ti) = A ₀ + A ₁ / ti + A ₂ / ti ² + A ₃ / ti ³ +
... + by regression analysis based on ^{An / ti n (i = 1} ~N), the following equation is a prediction _{function, Tp (tc) = A 0} + A 1 / tc + A 2 / tc 2 + A 3 / tc 3 + ... +
Parameters _A 0 of ^{_{An / tc n, A 1,}} A 2, ..., the prediction calculation of the temperature Tp at a future time tc and Motomeri signal analyzing means _{A n,} the prediction function Tp (tc) identified in the determined parameters And a prediction calculation means for performing the calculation.

Further preferably, when the plurality of temperature data is taken out, the extracting means is a time interval of time data correlated with a plurality of temperature data later than a time interval of time data correlated with a plurality of previous temperature data. One aspect is to take out so as to include the larger one.

[Description of Embodiments] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a block diagram showing the basic configuration of an electronic clinical thermometer according to an embodiment of the present invention. This electronic clinical thermometer is basically a temperature measuring unit 1 that detects the temperature of the measured part, and a predictive formula that is automatically determined based on a plurality of temperature measuring data and the measured elapsed time data at each temperature measuring time. Is composed of a prediction calculation unit 2 for predicting and calculating and a display unit 3 for displaying the calculated predicted temperature.

The temperature setting unit 1 is a unit that detects the temperature of the measurement site at a predetermined cycle and outputs temperature data T indicating the temperature to the line 104. The prediction calculation unit 2 is a calculation unit that internally includes a prediction function in which parameters are undecided. Before the measurement is started, the temperature measurement data T from the temperature measurement unit 1 is monitored to determine a predetermined measurement start condition, and the measurement is started. After that, in addition to the temperature measurement data T from the temperature measurement unit 1, the measurement elapsed time data t at each temperature detection point from the measurement elapsed time measurement function provided inside is sequentially stored, and a predetermined number of data is accumulated and stored. In addition, by using the data extracted from these, an undetermined parameter of the prediction function is determined by regression analysis, the predictive calculation of the temperature measurement value at the future time is performed by the determined predictive function, and the predicted temperature T _p of the calculation result is calculated by the line 117. Is the part to be output to. The display unit 3 is a unit for numerically displaying the predicted temperatures that are sequentially obtained.

FIG. 2 is a block diagram showing a specific configuration of the electronic thermometer of the embodiment. In the figure, the temperature measurement unit 1 includes a temperature sensitive element 4 such as a thermistor and a temperature measurement circuit 5, and the temperature measurement circuit 5 follows a data sampling command of a predetermined cycle sent from the prediction calculation unit 2 via a line 102. The analog electric signal 101 corresponding to the temperature detected by the temperature sensing element 4 is sampled and converted into a digital signal, and the lines 103 and 1
The circuit 04 outputs the temperature data T in real time.

The prediction calculation unit 2 includes a data reading unit 6, a time measuring unit 7, a measurement control unit 8, a storage unit (MEM) 9, and a storage control unit 1.
0, a data analysis unit 11, a prediction calculation unit 12, and a selection unit 13, each of which has a block configuration shown in FIG. 3 (a) and FIG. 3 (b) stored in a memory (ROM or RAM). Microcomputer (CP
U) can do this.

The measurement control means 8 is a means for integrally controlling the operation of the entire electronic thermometer. Before the measurement is started, the temperature measurement circuit 5 is caused to generate the temperature data T at a rate of, for example, once every 5 seconds, and the temperature data T is generated via the line 104. When the predetermined measurement start condition is determined, that is, when the temperature data T indicates a certain temperature or more and is accompanied by a temperature change of a certain value or more, the line 105 is displayed. And 109, the respective functions of the data reading means 6, the time measuring means 7, the storage control means 10, etc. are activated to start the measurement. After the measurement is started, the clock signal CLOCK generated in the prediction calculation unit 2 via the line 100 is received, for example, with a cycle of 1 second,
In response to this, the operation of each block configuration proceeds in the timer interrupt processing.

The time measuring means 7 uses the line 107 to measure elapsed time data t.
is a means for outputting _i , and the measurement control means 8 after the start of measurement
Counts the signal of 1 second period sent via line 105 to measure the elapsed measurement time from the start of measurement.

The data reading unit 6 is a unit that reads the temperature measurement data T on the line 103 into the predictive calculation unit 2 by the sampling signal sent by the measurement control unit 8 via the line 105 at a cycle of 1 second. Further, the data reading means 6 may be, for example, capable of accumulating several continuous temperature measurement data while updating the old and new, and may output the moving average value T ₀ of the several temperature measurement data to its output terminal. By doing this, the temperature measurement data becomes stable.

The storage means 9 is a means for sequentially correlating the temperature measurement data T ₀ read by the data reading means 6 and the measured elapsed time data t _i measured by the time measuring means 7 at the temperature measurement time thereof and storing them.

The storage control means 10 controls the above-mentioned data writing to the storage means 9, and sequentially extracts and provides the data analysis means 11 with a number of data necessary for data analysis as the measurement progresses. Is. Further, the storage control means 10
In ND, ND is a counter for writing a predetermined number of data in the storage means 9 or a counter for extracting a predetermined number of data from the storage means 9. SD is a counter that holds the address of the data to be read out first from the storage means 9, and DT is a register that holds the time interval (address interval) of a series of data to be extracted from the storage means 9.

FIGS. 4 (a) and 4 (b) show examples of modes of data writing control and data fetching control by the storage control means 10.
In FIG. 4 (a), the horizontal axis represents the measurement elapsed time t ₁ (seconds).
Is shown. The temperature measurement data T ₀ (t _i ) is stored in such a manner that it can be correlated with the time data t _i at the time of the temperature measurement.
If the temperature measurement data T ₀ (t _i ) of the second cycle is sequentially stored, the measurement elapsed time t _i on the horizontal axis is stored as it is in the storage means 9.
The address may be stored so that it can be replaced with the address ADD. This arrangement can be omitted storage area of the time data t _i. Thus, for the first 10 seconds, from ADD1 A
Temperature data T ₀ t ₁ ) to T ₀ (t ₁₀ ) are stored up to address DD10, and temperature data T ₀ (t ₁₁ ) to T ₀ (t ₂₀ ) from address ADD11 to address ADD20 for the next 10 seconds. Memorize
Even in this form, the temperature measurement data T ₀ (t _i ) can be correlated with the time data t _i at the temperature measurement time. Of course, as in the present embodiment, the temperature measurement data T ₀ (t _i ) and the time data t _i at the temperature measurement time may be stored as a pair.

On the other hand, assuming that 10 sets of data are required for each analysis of the temperature rise curve by the data analysis means 11, the storage control means 10 causes the data analysis means 11 to add addresses ADD1 to ADD10 when the first 10 seconds have elapsed. Temperature data T ₀ (t ₁ ) to T ₀ (t ₁₀ ) can be provided. Figure 4 (a)
The vertical axis indicates the data time interval (read interval) DT in each read cycle, and in the first read cycle, DT holds the contents of 1 second intervals. At this time, SD also holds the contents of the first second. Next 10
At the time when a second has passed, the storage control means 10 has SD and D
When the contents of T are 2 respectively, the temperature data T ₀ (t
₂ ), temperature data T ₀ (t ₄ ) at address ADD ₄ , temperature data T ₀ (t ₆ ) at address ADD ₆ , ..., And temperature data T ₀ (t ₂₀ ) at address ADD ₂₀ can be extracted and provided. In this way, as the measurement progresses, it is possible to analyze a temperature rise curve in a longer range even with a limited number of data, and the storage control is easy. Thereafter, the contents of SD and DT are increased as shown in the figure.

FIG. 4 (b) shows a case in which the contents of SD are fixed at the first second, and the others are the same as in FIG. 4 (a). In this way, the first second data will always be the analysis target. Furthermore, it is easy to extend this idea to various extraction methods.

Further, since the temperature measurement data T ₀ (t _i ) is obtained in a predetermined cycle, if the temperature measurement data T ₀ (t _i ) are stored in a predetermined sequence in the order of occurrence, the temperature measurement data T _{0 for} each sampling will be stored. It is not necessary to correlate (t _i ) with the current time data t _i each time. That is, for example, the storage control means 10 reads out the time data t _k of the time measurement means 7 via the line 118 and the measurement control means 8 only when the _kth temperature measurement data T ₀ (t _k ) is stored. Energize and line 11
When it is possible to read this through 9, all thermometer preceding said temperature measurement data T _{0 (t k)} data T ₀
(t _i ) is correlated with the time data t _i obtained by calculating t _i = i × t _k / k (where i = 1 to k−1).

The data analysis unit 11 is a unit that determines an undetermined parameter of the built-in prediction function, and is generally the storage control unit 10.
Reads each N pieces of temperature data T ₀ (t _i ) and time data t _i , the temperature data T ₀ (t _i ) is used as an objective variable and the function of the time data is explained as explanatory variables 1 / t _i , 1 / t _i ² , 1 / t _i ³ ,…, 1
The following equation is given as / t _i ⁿ , T ₀ (t _i ) = A ₀ + A ₁ / t _i + A ₂ / t _i ² + A ₃ / t _i ³ … + A _n / t _i ⁿ (i = 1 by regression analysis based on to n), the parameters of a prediction function (regression _{coefficient) a 0, a 1, a} 2, ..., is a means for obtaining the _{a n.} Thus undetermined parameter A ₀ to A _n of the prediction function is determined, it can be identified prediction equation.

The prediction calculation means 12 is a means for predicting calculation of a temperature measurement value in the future, preferably at an arbitrary time, using the prediction formula specified by the data analysis means 11, and the data analysis means 11 described above.
By the parameters A _{0 to} A _n determined by: T _p (t _c ) = A ₀ + A ₁ / t _c + A ₂ / t _c ² + A ₃ / t _c ³ … + A _n / t _c ⁿ Accordingly, the temperature T _p (t _c ) at the future time t _c is predicted and calculated, and the predicted temperature data T _p (t _c ) is output to the line 116.

The selecting unit 13 predicts the predicted temperature T _p , for example, after a sufficient time has passed that it does not make sense to continue the predictive calculation.
Is a means for switching from the preceding display by the measurement temperature data T _{0 to} a direct display, and the selection means 13 is connected to the prediction calculation means 12 side until the predetermined prediction stop condition after the start of measurement is determined. The display unit 3 displays the predicted temperature.

The operation principle of the temperature prediction according to the present invention in such a configuration is as follows.

The present inventor estimated the shape of the temperature rise curve of the temperature probe during body temperature measurement by performing theoretical analysis of heat conduction in the body temperature measurement system. That is, in this analysis, for example, the model of the body temperature measurement system shown in FIG. 5 is used, and the measurement system is divided into three regions of probe, skin, and subcutaneous tissue, and the temperature distribution of each region is measured in the process of body temperature measurement. I assume that. In other words, each region is treated by the concept of a minute volume, but the heat capacity of the subcutaneous tissue is assumed to be infinite. However, the names of the skin and the subcutaneous tissue are given for the sake of convenience to assume that the human body side is a two-layer model, and do not correspond exactly to the actual human body structure.
Also, if necessary, it is possible to improve the approximation model by dividing it into multiple regions according to future development.

In the measurement system model of FIG. 5, the heat transfer coefficient between the probe and the skin is h ₁ , its boundary area is A ₁ , and the heat transfer coefficient between the skin and the subcutaneous tissue is h ₂ , and its boundary area is A ₂ . . Also, the temperature of the subcutaneous tissue is assumed to have infinite heat capacity,
It is assumed that a constant value T _sat is taken with respect to time. Thus, since the amount of heat absorbed by the probe from the skin after the probe is attached to the measurement site is equal to the increase in the internal energy of the probe, the following equation (1) is established.

Similarly, since the amount of heat absorbed by the skin from the subcutaneous tissue and the probe is equal to the increase in the internal energy of the skin, the following expression (2) is established.

Where T _p , ρ _p , C _p , V _p : probe temperature, density, specific heat, volume T _s , ρ _s , C _s , V _s : skin temperature, density, specific heat, volume T _sat {= T _p (∞)}: temperature of subcutaneous tissue = equilibrium temperature.

Next, by solving the simultaneous linear differential equations consisting of the equations (1) and (2), the following equation (3) is obtained.

However, Is.

Since the equation (3) is a higher-order linear differential equation, it can be solved using the Laplace transform. That is, And calculate each term, And here, Is.

Further this Solving for Is obtained.

Here, if the solution of s ² + (K ₁ -K ₂ + K ₃ ) s + K ₁ · K ₃ = 0 is m ₁ , m ₂ , Becomes

Now, when m ₁ ≠ m ₂ , from equation (4), Becomes

here, Is known, the formula for T _p (t) is obtained as follows.

T _p (t) = T _sat + M ₁ e ^m1t + M ₂ e ^m2t (5) Is.

Also, when m ₁ = m ₂ , from equation (4), Is.

Again here Is known, the T _p (t) equation is obtained as follows.

T _p (t) = T _sat + M ₃ e ^m1t + M ₄ e ^m1t (6) However, M ₃ = C ₀ -T _sat M ₄ = C ₁ -m ₁ C ₀ + 3m ₁ T _sat .

Thus, the theoretical equation of the temperature rise curve of the probe is given by equations (5) and (6).

Now, in the above equations (5) and (6), m ₁ , m
₂ and M _{1 to} M ₄ are given as a function of various physical quantities contained in the body temperature measuring system including physical properties (density, specific heat, volume, etc.) of the probe and skin, and these values are thermometers. It changes from time to time and from measurement to measurement. Therefore, at the time of measurement, m ₁ , based on the temperature data detected by the probe,
m ₂ and M _{1 to} M ₄ need to be determined.

In addition, some electronic thermometers do not start reading temperature data after attaching the probe to the measurement site by some method, for example, until the probe detects a predetermined temperature. It is convenient to further transform equations (5) and (6) into the following equation.

T _p (t) ＝ T _sat + Pe ^m1t + Qe ^m2t (7) T _p (t) ＝ T _sat + Re ^m1t + Ste ^m1t (8) where P ＝ M ₁ e ^{m1 ′ Δ t} Q ＝ M ₂ e ^{m2 · Δ t} R = M ₃ e ^{m1 · Δ t} + M ₄ Δte ^{m1 · Δ t} S = M ₄ e ^{m1 · Δ t} , where Δt: time from probe attachment to measurement start t: measurement It is the time when the start time is t = 0.

In the above formula (7), if m ₁ and m ₂ can be set to fixed values, the regression analysis method or the simultaneous equations are solved by using the temperature data detected in time series during measurement.
T _sat , P, Q can be obtained relatively easily. However, m ₁ and m ₂ are values that change for each measurement depending on the person being measured or the measurement conditions, and moreover, all such variable elements are incorporated for each measurement to find the optimum prediction function, which is universal. It is an object of the present invention to perform highly predictable temperature prediction. In this case, there is a mathematical method for obtaining m ₁ , m ₂ and T _sat , P, Q by solving the simultaneous equation (7) using the temperature data detected during measurement. , The detected temperature data contains a noise component and the expression (7) contains an exponent part, which gives a very unstable result.

Then, Taylor expansion of the above equation (7) is performed to obtain the following (9)
Get the expression.

T _p (t) ＝ A ₀ + A ₁ / t + A ₂ / t ² + A ₃ / t ³ + ・ ・ ・ + A _i / t ⁱ + ・ ・ ・ ＝ T _sat + A ₁ / t + A ₂ / t ² + A ₃ / t ³ + ・ ・ ・ + A _i / t ⁱ + ・ ・ ・ {∵T _p (∞ (= T _sat } (9) Then, in this embodiment, for example, the fourth and subsequent orders are omitted. Then, the following equation (10) is obtained.

T _p (t) = A ₀ + A ₁ / t + A ₂ / t ² + A ₃ / t ³ (10) The above is the same for the above formula (8).

Next, according to the above equation (10), using temperature data detected in time series and time data at each time point, for example, N pieces of temperature data are set as the target variable T ₀ (t _i ) and the function of the time data is explained. variable _{1 / t i, 1 / t} i 2, the following equation to ^{_{1 / t i 3, T o}} (t i) = a 0 + a 1 / t i + a 2 / t i 2 + a 3 / t i By regression analysis based on ³ (i = 1 to N), the parameters (regression coefficients) A _{0 to} A ₃ of the prediction function can be determined.

The method of obtaining the parameters A _{0 to} A ₃ is as follows, for example.

The objective variable is T ₀ (t _i ), and the explanatory variables are 1 / t _i , 1 / t _i ² and 1 / t _i
³ and i = 1 to 10, for simplicity, T ₀ (t _i ) = A ₀ + A ₁ / t _i + A ₂ / t _i ² + A ₃ / t _i ³ is represented by the following equation.

The variance / covariance matrix of y _i = A ₀ + A ₁ · x _1i + A ₂ · x _2i + A ₃ · x _3i x _{1 to} x ₃ is As Can be expressed as

Also, And the covariance matrix of y and x ₁ to x ₃ is Can be expressed as

Then, the following simultaneous equations are solved to obtain A ₁ , A ₂ , and A ₃ .

The algorithm for this simultaneous equation is sweep out meth.
od) etc.

Next, A ₀ is calculated from the following equation.

In this way, the preceding temperature data includes all separable conditions, the parameters of the prediction function are determined from the correlation between these temperature data and time data, and the optimum prediction function for the time being is specified. That is why.

Therefore, using this identified prediction function, the future time t
The temperature at _c can be calculated according to the following equation: T _P (t _c ) = A ₀ + A ₁ / t _c + A ₂ / t _c ² + A ₃ / t _c ³ .

As described above, according to the present invention, first, the accumulation of the preceding temperature data is started, the parameter of the prediction function is determined by the temperature data, the prediction calculation of the temperature at the future time is performed by the determined prediction function, and the calculation is performed. By repeating the cycle of displaying the predicted temperature, a precedent display of the body temperature that is universal and reaches the equilibrium temperature smoothly in a short time can be obtained.

FIGS. 3 (a) and 3 (b) are flow charts showing the process of temperature measurement by the electronic thermometer of FIG. In FIG. 3 (a), when the power is turned on, the input is made to the starting step S100, and first, the temperature measuring unit 1 and the measurement control means 8 work to make the input to the temperature measuring step S101. Temperature measurement step S101
Then, the measurement control means 8 causes the temperature measuring circuit 5 to detect the temperature, for example, once every 5 seconds, and monitors the temperature detection data T. Judgment steps S102 and S103 are parts for deciding whether or not to start the body temperature measurement. In step S102, it is judged whether or not a predetermined temperature T _h , for example, 30 ° C. is exceeded, and in step S103, for example, for one second. Converted to 0.1
It is judged whether or not a temperature rise of ℃ or more is recognized. Then, if all of the above conditions are satisfied, step S1
In step 04, the time measuring means 7 is cleared and started via the line 105. That is, the time measuring counter of the time measuring means 7 is reset and the measurement of the measurement elapsed time is started. Further, in step S105, the data reading function of the data reading means 6 is activated via the line 105, and in step S106, the data reading / writing function of the storage control means 10 is activated via the line 109.

In steps S107 to S110, preprocessing for reading / writing the storage unit 9 is performed. That is, the counter SD is cleared in step S107, and the counter ND is set to 10 in step S108.
To set. In step S109, the content of the counter SD is incremented by 1, and in step S110, the content of the counter SD is set in the register DT. Then, in step S111, the timer interrupt can be performed once per second, and in step S112, the CPU executes an idle routine to wait for the occurrence of the timer interrupt.

In FIG. 3 (b), if a timer interrupt occurs, step S
Enter in 200. In step S201, a timer interrupt is added. In step S202, the temperature data T ₀ read by the data reading means 6 is written in the storage means 9. Step S20
In 3, the measured elapsed time data t _i generated by the time measuring means 7 at the temperature measuring time is written in the storage means 9. Step S20
At 4, the counter ND is decremented by -1. Step S20
At 5, it is determined whether or not the content of the counter ND is "0".
If not "0", the process returns to step S111 to wait for the next timer interrupt. In this way, the temperature data T ₀ and the measured elapsed time data t _i are sequentially correlated and stored in the storage means 9.

Eventually, when the content of the counter ND becomes "0", the step S2
Proceed to 06. Step storage control unit 10 in S206 reads the temperature data T ₀ and the measured elapsed time data t _i accumulated in the storage means 9 in the data analysis means 11. In this case, the 10 sets of data that the storage control means 10 first reads are temperature data T ₀ (t ₁ ) and time data t ₁ (= 1 second), temperature data T ₀ (t ₂ ) and time data t ₂ ( , And temperature data T ₀ (t ₁₀ ) and time data t ₁₀ (= 10 seconds). In step S207, the data analysis means 11 uses [T ₀ (t ₁ ), 1 / t ₁ , 1 / t ₁ ² , 1 / t ₁ ³ ] [T ₀ (t ₂ ), 1 based on these data. / t ₂ , 1 / t ₂ ² , 1 / t ₂ ³ ] ・ [T ₀ (t ₁₀ ), 1 / t ₁₀ , 1 / t ₁₀ ² , 1 / t ₁₀ ³ ] table is created and temperature data is created. T ₀ (t _i ) is the objective variable,
And the function of time data correlated with this 1 / t _i , 1 / t _i ² , 1 / t
By regression analysis using the following equation with _i ³ as an explanatory variable, T ₀ (t _i ) = A ₀ + A ₁ / t _i + A ₂ / t _i ² + A ₃ / t _i ³ (i = 1 to 10) The parameters (regression coefficient) A _{0 to} A ₃ of the prediction function are obtained. In step S208, the regression coefficients A _{0 to} A ₃ obtained by the data analysis means 11 are sent to the prediction calculation means 12, and the prediction calculation means 12 uses the following equation: T _p (t _c ) = A ₀ + A ₁ / t _c + A _{According to 2} / t _c ² + A ₃ / t _c ³ , the predicted temperature t of the measurement site at the future time t _{c
Find p} (t _i ). According to the present invention, since the measured elapsed time data t _c can be set arbitrarily, it is possible to accurately predict the temperature measurement value when an arbitrary time in the future has passed. That is, when predicting the temperature at the time of thermal equilibrium, t _c is ideally infinite, but when considering the convention of general thermometry, for example, t _c
c can be set to a predetermined value within the range of 300 to 600 seconds.

In step S209, the calculated predicted temperature T _p is digitally displayed on the display unit 3. In step S210, it is determined whether the prediction is completed. As the judgment condition for the prediction end, for example, when the absolute value of the gradient of the predicted temperature is less than or equal to a predetermined value, or when the difference between the predicted temperature and the measured temperature is less than or equal to a predetermined value, or the prediction calculation is continued When a sufficient measurement time has passed so as not to fail, etc. Then, if the prediction end determination condition is not satisfied in step S210, the process returns to step S108. If the prediction end determination condition is satisfied in step S210, the process proceeds to step S211, and the process ends. In this way, the preceding display of the predicted temperature is fixed.

It should be noted that, after about 10 minutes have passed since the start of measurement, the measured temperature generally reaches the equilibrium temperature regardless of the measurement site. Instead of fixing the preceding display of the temperature, the selection of the selection means 13 in FIG. 2 may be reversed to display the measured temperature T ₀ .

FIG. 6 is a graph showing the measurement process of axillary temperature measurement by the electronic thermometer of the embodiment, and FIGS. 7 and 8 are graphs showing the measurement process of oral temperature measurement. According to these figures,
Predicted value of equilibrium temperature T _p (∞) for temperature measurement data T _{0 in the} armpit and mouth, or temperature measurement value T ₀ max after 420 seconds have elapsed
The predicted value T _p (420) of (420) or the predicted value T _p ( ₆₀₀ ) of the temperature measurement value T ₀ max (600) after the elapse of 600 seconds should have a very stable rising curve. I understand. In general, variations in thermal characteristics during mass production of probes have a smaller effect on the shape of the body temperature measurement curve than differences in the site to be measured (armpit, mouth, etc.), so even if the probe is changed, the transition of the predicted value is stable. Draw a rising curve. Also, the time until the predicted value points to the temperature measurement value at the future time is not necessarily shorter than the conventional prediction method, but overshoot near the start of the measurement that is likely to occur in the conventional prediction method,
The instability in which the predicted value deviates extremely due to noise superimposed on the measured temperature curve as shown in FIG. 8 is eliminated.

In the description of this embodiment, the right side of the regression equation is up to the third order, but this order can be changed. However, if the order is reduced, the prediction effect becomes smaller. For example, if the order is reduced from the third order to the second order, the change in the predicted value is as shown in FIG. The area between the curve T ₀ and the curve T ₀ is drawn. Also, it should be taken into consideration that the regression analysis takes time and the memory required for calculation increases as the order increases.

Further, in FIG. 5, if the body temperature measuring system is divided into more regions, it can be seen that the assumption of uniform temperature for each region, that is, the handling as a minute volume element is more appropriate. By the way, when the area is increased, the calculation becomes exponentially complicated, but the shape of T _p (t) can be relatively easily inferred by following the expansion of the equation. For example, when one layer is added between the skin and the subcutaneous tissue, the simultaneous differential equations corresponding to the equations (1) and (2) are changed from two elements to three elements. A cubic equation of dT _p (t) / dt corresponding to is obtained. Further expanding this using the Laplace transform, the denominator on the right side of the expression corresponding to expression (4) becomes a cubic expression of S. Then, assuming that these three solutions are S ₁ , S ₂ , and S ₃ , respectively, the finally obtained form of T _p (t) is as follows.

That is, when S ₁ ≠ S ₂ , S ₂ ≠ S ₃ , S ₃ ≠ S ₁ , T _p (t) = T _sat + Ψ ₁ e ^s1t + Ψ ₂ e ^s2t + Ψ ₃ e ^s3t and S ₁ When S ≠ S ₂ and S ₁ ≠ S ₃ , T _p (t) = T _sat + Ψ ₄ e ^s1t + Ψ ₅ e ^s2t + Ψ ₆ te ^s1t , and when S ₁ = S ₂ = S ₃ , T _p (t) = T _sat + Ψ ₇ e ^s1t + Ψ ₈ te ^s1t + Ψ ₉ t ² e ^s1t .

Thus, the shape of T _p (t) changes with the increase in the number of regions, and it can be inferred from the way of expansion without recalculation. That is, even when the region is divided into multiple regions, the obtained T _p (t) is transformed into the formula (9) by Taylor expansion, so the same result can be obtained according to the above-mentioned reasoning. .

[Effects of the Invention] As described above, according to the present invention, since all the coefficients in the prediction formula are calculated using the real-time temperature data at the time of measurement, there are variations in the thermal characteristics of the probe or individuals or measurement sites. Even if there is a difference in the temperature rise curve due to the difference, the accurate preceding display can always be performed.

Further, according to the present invention, since the real-time temperature data itself is used as the objective variable, there is no influence due to the calculation error, the parameter determination is stable, and the predicted value largely fluctuates even with noise superimposed on the measured temperature curve. I do not.

Further, according to the present invention, since the temperature data is taken out so as to cover the entire temperature rise curve at the time of each measurement, the transition of the predicted value draws a natural rise curve and may overshoot near the rise of the temperature. Without it, the measurement can be performed without the user being aware of what is predicted.

Further, according to the present invention, an arbitrary future time can be set for the prediction formula, so that the temperature measurement value after the lapse of an arbitrary measurement time in the measurement system can be easily provided.

Further, according to the present invention, an arbitrary future time can be set in the prediction formula, so that the predicted value of the thermal equilibrium temperature after a very long time in the future can be easily provided.

[Brief description of drawings]

FIG. 1 is a block diagram showing a basic configuration of an electronic clinical thermometer of an embodiment according to the present invention, FIG. 2 is a block diagram showing a concrete configuration of an electronic clinical thermometer of an embodiment, and FIGS. 3 (a) and 3 (b) are schematic diagrams. 2 is a flow chart showing the temperature measuring process by the electronic thermometer of FIG. 4, FIGS. 4 (a) and 4 (b) are diagrams showing a mode of data writing and data reading to the storage means, and FIG. 5 is a heat conduction of the body temperature measuring system. The figure which shows a model, FIG. 6 is a graph which shows the measurement process of axillary temperature measurement by the electronic thermometer of an Example, FIGS. 7 and 8 are the graph diagrams which show the measurement process of mouth thermometry by the electronic thermometer of an Example, FIG. 9 is a graph showing the relationship between the order of the prediction formula and the prediction curve. In the figure, 1 ... Temperature measuring unit, 2 ... Prediction calculation unit, 3 ... Display unit, 4 ... Temperature sensing element, 5 ... Temperature measuring circuit, 6 ... Data reading means, 7 ... Time measuring means , 8 ... measurement control means, 9 ... storage means, 10 ... storage control means, 11 ...
Data analysis means, 12 ... Prediction calculation means, 13 ... Selection means.

Claims (4)

[Claims] 1. A temperature detecting means for detecting a temperature and generating temperature data indicating the temperature, and a time signal generating means for measuring an elapsed time after the start of measurement and generating time data ti indicating the elapsed time. Storage means for storing the detected temperature data in correlation with the time data ti at the time of the detection, extraction means for extracting a plurality of temperature data from the storage means, and using the extracted temperature data as an objective variable Functions 1 / ti, 1 / ti ² , of the correlated time data ti,
1 / ti ^3, ..., by regression analysis of the function to be described variable 1 / ti ^n, the prediction calculation and signal analyzing means for determining the parameters of the prediction function, the temperature at a future time by the prediction function specified by the determined parameters An electronic clinical thermometer, comprising: 2. The time period of time data correlated to a plurality of temperature data after the time interval of time data correlated to a plurality of temperature data before, when the plurality of temperature data are retrieved from the plurality of temperature data. The electronic clinical thermometer according to claim 1, wherein the electronic thermometer is taken out so as to include the one with a larger interval. 3. Temperature data T for detecting temperature and showing the temperature
₀ , temperature detection means, time signal generation means for measuring the elapsed time after the start of measurement and generating time data ti indicating the elapsed time, and detected temperature data T ₀ for the time data t at the time of the detection.
storage means for storing correlation with i, extraction means for extracting a plurality of temperature data T ₀ (ti) from the storage means, and the extracted temperature data T ₀ (ti) as an objective variable, and correlation with this Function of time data ti 1 / ti, 1
The following equation using / ti ² , 1 / ti ³ , ..., 1 / ti ⁿ as explanatory variables, T ₀ (ti) = A ₀ + A ₁ / ti + A ₂ / ti ² + A ₃ / ti ³ + ... +
By regression analysis based on ^{An / ti n (i = 1} ~N), the following equation is a prediction _{function, Tp (tc) = A 0} + A 1 / tc + A 2 / tc 2 + A 3 / tc 3 + ... +
A _n / tc parameter _A 0 of _{^{n, A 1, A 2,}} ..., the prediction calculation of the temperature Tp at a future time tc and signal analyzing means for determining the _{A n,} the prediction function Tp (tc) identified in the determined parameters An electronic clinical thermometer, comprising: 4. The time of the time data correlated with the plurality of temperature data after the time interval of the time data correlated with the plurality of temperature data before, when the plurality of temperature data is retrieved. The electronic thermometer according to claim 3, wherein the electronic thermometer is taken out so as to include one having a larger interval.