JPH053873A - Living body rhythm curve evaluating device - Google Patents

Abstract

PURPOSE:To offer the device which can evaluate a living body rhythm curve by taking environmental information and physical condition information into consideration. CONSTITUTION:Based on a living body rhythm curve, a feature parameter, environmental information and physical condition information, a data base 5 is constituted, and when the living body rhythm curve of a certain day, its feature parameter, and the environmental information are inputted, the data base 5 is retrieved by a retrieving means 6, and physical condition information on that day is decided, and outputted by an output means 7. Also, the data base 5 is corrected by a data base correcting means 8 so that a difference between the actual physical condition information of this and that obtained by the output means 7 is converted to zero. In such a way, as the device is used, it conforms to the use environment and the physical condition of a person to be examined, and the quantitative evaluation accuracy of the living body rhythm curve is improved.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a biorhythm curve evaluation device for evaluating a biorhythm curve based on a database composed of a human biorhythm curve and its characteristic parameters, environment information and physical condition information. is there.

[0002]

2. Description of the Related Art When various biological phenomena are expressed in time series, they often show periodicity. Moreover, many of them are considered to be self-excited vibrations, and are collectively called biorhythms. Biological rhythms are divided into several types according to their cycle, and they range from a long one year to a few seconds. Humans become more active in the light period and become active, and in the dark period become less alert and enter rest. This is due to the Circadian rhythm.
m: biological rhythm (biological rhythm) carved by a biological clock called a rhythm with a cycle of about 1 day.
Is one of.

Of the biological rhythms, the one most closely related to human life is the circadian rhythm having a cycle of about one day. Typical circadian rhythms in humans include temperature fluctuations, sleep-wake cycles, and hormone secretion fluctuations. In addition, physiological phenomena associated with human life, such as mental and physical activity, work and exercise ability, sensitivity to drugs, and functions of the autonomic system, can be considered to show circadian fluctuations.

At present, the theory that human circadian rhythm may be divided into two vibrating body groups, a group centering on the core body temperature rhythm and a group centering on the sleep-wake cycle, is predominant. It is said that the core body temperature rhythm is influenced by the light-dark cycle, and the sleep-wake cycle is influenced by the social entrainment factor.
At the laboratory level, there are quite advanced technologies such as polygraphs for monitoring arousal level and biological rhythm, but they do not cause pain to the subjects in daily work situations, and
Under the present circumstances, there is no hindrance to the work behavior, and the activity of the living body cannot be monitored non-invasively.

The method of measuring the biological rhythm based on the body temperature will be described below. The body temperature, especially the core body temperature rhythm, has little influence from the outside, shows a clear circadian rhythm, has a fairly clear relationship with other rhythms, and is capable of continuous measurement. It is regarded as the most important indicator of circadian rhythm. Candidates for deep body temperature measurement include rectal temperature, eardrum temperature, esophageal temperature,
Examples include deep subcutaneous temperature and urine temperature, but the rectal temperature satisfies the condition that continuous measurement is possible for a long time. However, it is a disadvantage that all of them are measurement methods that give pain to the subject.

A general method for measuring rectal temperature is to insert a probe with a thermistor embedded at the tip from the anus for 10 cm or more and fix it with tape so that it will not come off. A device for calculating a temperature from a resistance value of a thermistor and storing it in a memory is commercially available as a portable thermometer. In addition to the method of directly measuring the rectal temperature, a device (core temp) capable of measuring the deep body temperature from the surface of the skin by the convective heat exchange method is commercially available. The larger the sensor diameter, the deeper the body temperature can be measured, approximately 10m from the skin surface.
It is possible to measure body temperature up to m depth. However, in this method, the sensor needs to heat the skin, and when used for a long time like rhythm measurement, there is a risk of low-temperature burns, and care must be taken when handling.

Next, a method of analyzing the biological rhythm will be described. Basically, the analysis of biological rhythm is to find the three elements of rhythm (cycle, phase, shape). In order to analyze the body temperature rhythm, the period and the minimum value phase are calculated, and the amplitude is calculated as the characteristic of the shape. In the body temperature rhythm, the minimum value phase is an especially important factor, but this is generally obtained by visual inspection.

As a method of obtaining the rhythm cycle from the data measured at regular intervals, the autocorrelation method (correlogram), the power spectrum method, the cosigner method, the periodogram method, etc. can be used. For body temperature rhythm, the cosigner method and the periodogram method are often used. The cosigner method is a method for obtaining the period, amplitude, and phase by least-squares approximation to a sine wave.

Next, as a method for exactly obtaining the circadian rhythm, a constant routine (Consta)
nt route) method is known. Since this method is affected by the outside even if the body temperature is deep, the true rhythm is not always expressed as it is. It is a method to measure the rectal temperature for several tens of hours and obtain the cycle, amplitude, and phase of the core body temperature.

[0010]

As described above, as a method of analyzing a biological rhythm by measuring a body temperature, a visual inspection method, a cosinar method, and a constant routine method are generally used. However, a quantitative analysis is performed by the visual inspection method. There is a problem that you can not. In addition, the co-signer method has a problem that the waveform of the body temperature rhythm is not always sinusoidal. Further, the constant routine method can eliminate the influence of disturbance, but it requires the subject to be held in the laboratory for several tens of hours, which makes it impossible to evaluate the rhythm curve in daily life. In addition, in order to know the characteristics of the biorhythm, it is important to qualitatively evaluate the length of the cycle, the magnitude of the amplitude, the advance / retreat of the phase, etc. There is a problem that it cannot be evaluated. For example, there is a problem that it is not possible to quantitatively evaluate whether the biorhythm curve is clean or quantitatively whether the biorhythm curve is sharp. Further, the conventional technique has a problem that it is not possible to make an evaluation in consideration of environmental information and physical condition information.

The present invention has been made in view of the above points, and an object thereof is a biological rhythm curve evaluation device capable of quantitatively evaluating a biological rhythm curve in consideration of environmental information and physical condition information. To provide.

[0012]

In order to solve the above-mentioned problems, in the biorhythm curve evaluation apparatus according to the present invention, as shown in FIG. 1, a rhythm curve input means 1 for inputting a biorhythm curve. A characteristic parameter input means 2 for inputting a characteristic parameter of a biorhythm curve, an environment information input means 3 for inputting environment information, a physical condition information input means 4 for inputting physical condition information, and input means 1, 2, 3 , 4, a database 5 configured based on the input information, a search unit 6 that searches the database 5 to evaluate a biorhythm curve, and an output unit 7 that outputs the evaluation result of the search unit 6.
And a database correction means 8 having a learning function.

As shown in FIG. 3, a neural net in which a plurality of neurons are connected is used, and the database 5 is constructed as a set of connection degrees of the neurons.
The searching means 6 may be constituted by a forward signal transmission function from the input layer of the neural network to the output layer, and the database correction means 8 may be constituted by a backward learning function of the neural network from the output layer to the input layer. .

[0014]

In the present invention, the biorhythm curve inputted by the rhythm curve input means 1, the characteristic parameter of the biorhythm curve inputted by the characteristic parameter input means 2, and the environmental information inputted by the environment information input means 3 When,
The database 5 is configured on the basis of the physical condition information input by the physical condition information input means 4. When the biological rhythm curve of one day and its characteristic parameters and environmental information are input, the search means 6 searches the database 5. Then, you can judge the physical condition of the day. This evaluation result is output by the output means 7. Further, the database 5 is corrected by the database correcting unit 8 so that the error between the actual physical condition information of the day and the physical condition information obtained by the output unit 7 is converged to zero. As a result, the learning function can be provided, and the accuracy of the evaluation gradually improves as the evaluation device is used.

[0015]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention. In the figure, reference numeral 1 denotes a rhythm curve input means for inputting biorhythm curve data measured based on deep body temperature measurement data and electrocardiogram measurement data. Reference numeral 2 is a characteristic parameter input means for inputting a characteristic parameter of a biorhythm curve. The characteristic parameters include, for example, the three elements of the cycle, amplitude, and phase of the biorhythm curve, the duty ratio, the spectrum, the rising / falling slope, the number of minimum values, and the number of maximum values. Reference numeral 3 denotes an environmental information input means, which inputs environmental information such as the outside air temperature, the activity of the wrist, and the ambient illuminance as objective measurement data. Reference numeral 4 is a physical condition information input means for inputting the physical condition of the subject on that day. The physical condition information may be based on subjective evaluation or objective evaluation (performance test, etc.), but either may be used. 5 is a database,
The input information from each input means 1, 2, 3, 4 is accumulated. Reference numeral 6 denotes a search means, which is configured to search for the corresponding physical condition information by using the rhythm curve, its characteristic parameters, and environment information as the key to the search. Reference numeral 7 denotes an output unit that outputs the physical condition information found by the search unit 6. Reference numeral 8 denotes a database correction means, which compares the physical condition information output from the output means 7 with the actual physical condition information of the day, and if there is an evaluation error, the database 5 is set in a direction in which the evaluation error converges to zero. to correct. With this feedback, the learning function is obtained, and the evaluation accuracy gradually improves as the evaluation device is used.

Next, the structure of the database 5 will be described. Here, it is assumed that the physical condition information is expressed as a function of the characteristic parameter of the biological rhythm curve or the environmental information, and the physical condition = a × period + b × phase + c × amplitude + d × stability + e × spectrum + f × outside temperature + g Define a function such as × activity ... Then, by accumulating the daily input information, the coefficients a, b, c, ... Are obtained as the solution of the simultaneous equations, and the coefficient is updated by further accumulating the information, and the accuracy of the evaluation gradually increases. go.

Further, instead of the function, it is assumed that the physical condition information is determined for each of a plurality of regions in the multidimensional plane consisting of the characteristic parameters of the biorhythm curve and the environmental information, and the input information of the day is close to which region. By determining whether or not it is possible to evaluate the physical condition of the day. Then, by accumulating daily input information, the area in the feature plane is updated, and the accuracy of evaluation gradually increases.

Further, by using the neural network as shown in FIG. 2, the database 5, the searching means 6 and the database correcting means 8 can be realized at the same time. The neural network has a hierarchical structure of an input layer, an intermediate layer, and an output layer. This middle layer is "hidden layer"
It is called (hidden layer) and plays an important role. The structure of each unit is shown in FIG. Units are connected from the input layer to the output layer. There is no bond within each layer. The input pattern p is transmitted from the input layer to the output layer through the intermediate layer. This constitutes the search means 6. On the other hand, learning proceeds from the output layer to the input layer. This constitutes the database correction means 8. Learning is to change the strength of the coupling so as to reduce the error between the actual output and the desired output value. The database 5 is constituted by the set of the strength of the bond.

The neural network will be further described below. FIG. 3 shows a model of the configuration of the neuron. The structure is simple and consists of a part that receives input from other units, a part that converts the input according to a certain rule, and a part that outputs the result. A variable weight Wij is attached to each coupling portion with another unit. This is a coefficient that represents the strength of the bond. Changing this value changes the network structure. Learning the network means changing this value. Wij takes positive, zero, and negative values. Zero means no bond. When a unit receives input from multiple units, the total sum is used as the input value. The total sum "net" of the inputs is given by the following equation.

[0020]

[Equation 1]

Each unit applies the sum of inputs net to the function f and transforms it. This input / output characteristic may be different for each unit. However, in general, the threshold function or si
Often uses the gmoid function. In the threshold function,
The output is 1 if the total "net" of the inputs exceeds the predetermined threshold value .theta., And the output is 0 otherwise. Also, sig
The moid function is a pseudo-linear function that can be differentiated, and is defined by the following equation.

[0022]

[Equation 2]

The value of the sigmoid function is in the range of 0 to 1, and approaches 1 as the input value increases and approaches 0 as the input value decreases. The input is 0
When, the function value is 0.5. The following function may be used by adding a predetermined threshold value θ to the input value.

[0024]

[Equation 3]

The network has a hierarchical structure as shown in FIG. When each unit of the input layer is given with characteristic parameters such as period, phase, amplitude, stability, spectrum of the biorhythm curve and various environmental information such as outside temperature, wrist activity and ambient illuminance, these The signal of is converted by each unit, transmitted through the intermediate layer, and finally the physical condition information is obtained from the output layer. The physical condition information output from this neural network is compared with the actual physical condition information of the day, and if there is an error, the strength of the connection is changed so that the error converges to zero. As a result, the accuracy of evaluation gradually improves as the device is used.

[0026]

According to the biorhythm curve evaluation apparatus of the first aspect, the database is constructed based on the biorhythm curve, its characteristic parameters, environment information and physical condition information, and the database can be searched by the searching means. There is an effect that it becomes possible to evaluate the biorhythm curve after considering information and physical condition information.

According to the biological rhythm curve evaluation device of claim 2,
By using the neural network, the database, the search means, and the database correction means are integrated, and there is an effect that highly accurate evaluation can be performed with a simple configuration.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of the present invention.

FIG. 2 is a diagram showing a configuration of a neural network used in the present invention.

FIG. 3 is a diagram showing a configuration of a neuron used in the present invention.

[Explanation of symbols]

1 Rhythm curve input means 2 Feature parameter input means 3 Environmental information input means 4 Physical condition information input means 5 database 6 Search method 7 Output means 8 Database correction means

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁵ Identification number Office reference number FI technical display location G06F 15/42 B 7060-5L

Claims (2)

[Claims] 1. Rhythm curve input means for inputting a biorhythm curve, characteristic parameter input means for inputting characteristic parameters of a biorhythm curve, environment information input means for inputting environment information, and physical condition information for inputting physical condition information. An input unit, a database configured on the basis of input information of each input unit, a search unit that searches the database to evaluate a biorhythm curve, an output unit that outputs an evaluation result by the search unit, and a learning function. A biological rhythm curve evaluation device, comprising: a database correction means having the same. 2. A neural network comprising a plurality of neurons connected to each other, said database being configured as a set of coupling degrees of each neuron, and said search means having a function of transmitting a forward signal from an input layer to an output layer of the neural network. 2. The biological rhythm curve evaluation device according to claim 1, wherein the database correction means is configured by a backward learning function from the output layer to the input layer of the neural network.