JP2021137438A - Stress estimation program, stress estimation device, and stress estimation method - Google Patents

Abstract

To provide a stress estimation program, a stress estimation device, and a stress estimation method that make it possible to accurately and stably determine a stress state of an object person.SOLUTION: A stress estimation program makes processing be executed, the processing including: calculating a first index using a LF, a power spectrum in a low frequency band included in time series data on an object person's heartbeat interval, and a HF, a power spectrum in a high frequency band included in the time series data; calculating a second index using a SDNN, a standard deviation of heartbeat intervals included in the time series data, and a RMSSD, a square root of a mean value of squares of differences between two heartbeat intervals continuously included in the time series data; calculating a value representing a state of stress of the object person by using the calculated first index and second index; and outputting the calculated value.SELECTED DRAWING: Figure 3

Description

The present invention relates to a stress estimation program, a stress estimation device and a stress estimation method.

In general, stress not only causes illnesses such as depression, but also has effects such as fatigue, loss of appetite, fever, and insomnia. Therefore, in recent years, the importance of mental health has been recognized, and for example, there is increasing interest in methods for controlling one's own stress (Patent Documents 1 to 4 and Non-Patent Documents 1 to 2). reference).

Japanese Unexamined Patent Publication No. 2005-218595 Japanese Unexamined Patent Publication No. 2010-033148 Japanese Unexamined Patent Publication No. 2016-142553 Japanese Unexamined Patent Publication No. 2017-064216

Estimation of Stress dancing Car Race with Factor Analysis, Daisuke Tomoi et al. , 2015 International Symposium on Micro-nano Mechatronics and Human Science (MHS), Nagoya, 2015, pp. 1-5. Unbiased entropy stimates in stress: A parameter study, Aleksandar Boskovic et al. , Computers in Biologic and Medicine, vol. 42, 2012, pp. 667-679.

Here, the stress state in the human body may be determined by, for example, subjective evaluation using a psychological index based on the results of interviews, questionnaires, and the like.

However, in the judgment by subjective evaluation, it may take a long time for the diagnosis by a doctor or the like, and the objective judgment may not be made.

On the other hand, for example, the stress state as described above may be determined by using the objective evaluation used as a physiological index based on the measurement results such as electroencephalogram and electrocardiogram.

However, in the judgment by objective evaluation, if the accuracy and sensitivity of the physiological index used for the judgment are not sufficient, it may not be possible to stably obtain a highly accurate judgment result.

Therefore, it is an object of the present invention to provide a stress estimation program, a stress estimation device, and a stress estimation method that enable a highly accurate and stable determination of a stress state of a subject.

The stress estimation program in the present invention for achieving the above object includes LF (Low frequency), which is a low frequency band power spectrum included in the time series data of the heartbeat interval of the subject, and high, which is included in the time series data. The first index is calculated using the power spectrum of the frequency band, HF (High frequency), and SDNN (Standard deviation of NN intervals), which is the standard deviation of the heartbeat interval included in the time series data, and the time series. The second index was calculated using RMSD (Root mean square of success differentials), which is the square root of the squared value of the difference between the two heartbeat intervals included in the data in succession, and the calculated first index and the first index. By using the second index, a value indicating the stress state of the subject is calculated, and the calculated value is output, which is characterized by causing a computer to execute a process.

Further, in one aspect of the stress estimation program in the present invention for achieving the above object, the first index is a value calculated by dividing the LF by the HF, and the second index is a value calculated by dividing the LF by the HF. , The value is calculated by dividing the SDNN by the RMSDD.

Further, the stress estimation program in the present invention for achieving the above object is characterized in that, in one embodiment, the value is calculated by multiplying the first index and the second index.

Further, the stress estimation program in the present invention for achieving the above object calculates, in one embodiment, a third index showing the sample entropy indicating the unpredictability (irregularity) of the fluctuation of the time series data. The value is calculated by multiplying the first index and the second index and dividing by the third index.

Further, in one embodiment, the stress estimation program in the present invention for achieving the above object specifies the maximum value and the minimum value from the values calculated in the past for the subject, and the specified maximum value and the said maximum value. It is characterized in that the value is normalized by using the minimum value and the normalized value is output.

Further, in one embodiment, the stress estimation program in the present invention for achieving the above object subtracts the minimum value from the value calculated by the calculation process, and subtracts the minimum value from the maximum value. By dividing by the calculated value, the value calculated by the calculation process is normalized.

Further, the stress estimation device in the present invention for achieving the above object has an LF which is a power spectrum of a low frequency band included in the time series data of the heartbeat interval of the subject and a high frequency band included in the time series data. The first index is calculated using HF, which is the power spectrum of the above, and the difference between SDNN, which is the standard deviation of the heartbeat interval included in the time series data, and two heartbeat intervals, which are continuously included in the time series data. By using the index acquisition unit that calculates the second index using the RMSSD, which is the square root of the squared value of the square, and the calculated first index and the second index, the stress state of the subject. It is characterized by having an information calculation unit for calculating a value indicating the above value and an information output unit for outputting the calculated value.

Further, the stress estimation method in the present invention for achieving the above object is LF, which is a power spectrum of a low frequency band included in the time series data of the heartbeat interval of the subject, and a high frequency band included in the time series data. The first index is calculated using HF, which is the power spectrum of the above, and the difference between SDNN, which is the standard deviation of the heartbeat interval included in the time series data, and two heartbeat intervals, which are continuously included in the time series data. The second index is calculated using the RMSSD, which is the square root of the mean value of the square of, and the calculated first index and the second index are used to obtain a value indicating the stress state of the subject. It is characterized in that a computer executes a process of calculating and outputting the calculated value.

According to the stress estimation program, the stress estimation device, and the stress estimation method in the present invention, it is possible to determine the stress state of the subject with high accuracy and stability.

FIG. 1 is a diagram showing a configuration example of the stress estimation device 1 according to the embodiment of the present invention. FIG. 2 is a diagram illustrating a specific example when the stress state is estimated. FIG. 3 is a diagram illustrating a stress estimation process in the outline of the first embodiment. FIG. 4 is a flowchart illustrating the details of the stress estimation process according to the first embodiment. FIG. 5 is a flowchart illustrating the details of the stress estimation process according to the first embodiment. FIG. 6 is a diagram illustrating details of the stress estimation process according to the first embodiment. FIG. 7 is a diagram illustrating an output example in the processing of S24. FIG. 8 is a diagram illustrating an output example in the processing of S24. FIG. 9 is a diagram illustrating an output example in the processing of S24. FIG. 10 is a diagram illustrating an output example in the processing of S24. FIG. 11 is a diagram illustrating an output example in the processing of S24. FIG. 12 is a diagram illustrating an output example in the processing of S24. FIG. 13 is a diagram illustrating an output example in the processing of S24. FIG. 14 is a diagram illustrating an output example in the processing of S24. FIG. 15 is a diagram illustrating an output example in the processing of S24. FIG. 16 is a diagram illustrating an output example in the processing of S24. FIG. 17 is a diagram illustrating an output example in the processing of S24. FIG. 18 is a diagram illustrating an output example in the processing of S24.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, such embodiments do not limit the technical scope of the invention.

FIG. 1 is a diagram showing a configuration example of the stress estimation device 1 according to the embodiment of the present invention. The stress estimation device 1 may be a computer device, for example, a general-purpose PC (Personal Computer). Further, the stress estimation device 1 may be in any form such as a stationary type, a node book type, and a tablet type.

The stress estimation device 1 has a hardware configuration of a general-purpose computer device, and includes, for example, as shown in FIG. 1, a CPU 101 which is a processor, a memory 102, a network interface 103, and a storage medium 104. The parts are connected to each other via the bus 105.

The storage medium 104 has, for example, a program storage area (not shown) for storing a program (not shown) for performing a process of estimating the stress state of each subject (hereinafter, also referred to as a stress estimation process). Further, the storage medium 104 has, for example, a storage area 110 for storing information used when performing stress estimation processing. The storage medium 104 may be, for example, an HDD (Hard Disk Drive) or an SSD (Solid State Drive).

The CPU 101 executes a program loaded from the storage medium 104 (storage area 110) into the memory 102 to perform stress estimation processing.

Further, the network interface 103 communicates with, for example, an operation terminal 5 operated by an operator. The operation terminal 5 may be a computer device, for example, a general-purpose PC, a mobile terminal such as a smartphone or a tablet.

[Estimation of stress state]
Next, a specific example when the stress state is estimated will be described. FIG. 2 is a diagram illustrating a specific example when the stress state is estimated.

As shown in FIG. 2, when the subject is in a high stress state, the hypothalamus of the subject enhances the sympathetic nerve function and suppresses the parasympathetic nerve function. As a result, the subject's heart is affected by a decrease in heart rate variability and the like.

Therefore, in the determination of the stress state by the objective evaluation, for example, a method of estimating the stress state of the subject from the stress index based on the electrocardiographic signal measured in the heart of the subject is used.

[Outline of the first embodiment]
Next, the outline of the first embodiment will be described. FIG. 3 is a diagram illustrating a stress estimation process in the outline of the first embodiment.

First, the index acquisition unit 11 of the stress estimation device 1 calculates a plurality of indexes used for determining the stress state of the subject.

Specifically, the index acquisition unit 11 is, for example, a low frequency band (frequency band of 0.04 to 0.15 Hz) included in time series data (hereinafter, also simply referred to as time series data) regarding the heartbeat interval of the subject. An index (hereinafter, also referred to as a first index) is obtained by using LF, which is the power spectrum of the above, and HF, which is the power spectrum of the high frequency band (frequency band of 0.15 Hz to 0.4 Hz) included in the time series data. calculate.

Further, the index acquisition unit 11 is, for example, the square root of the average value of the square of the difference between the SDNN, which is the standard deviation of the heartbeat intervals included in the time series data, and the difference between the two heartbeat intervals continuously included in the time series data. An index (hereinafter, also referred to as a second index) is calculated using a certain RMSSD.

The time-series data may be stored in the storage area 110 in advance, or may be input via the operation terminal 5 as needed.

Then, the information calculation unit 12 of the stress estimation device 1 calculates a value indicating the stress state from the plurality of indexes acquired by the index acquisition unit 11. After that, the information output unit 15 of the stress estimation device 1 outputs, for example, the value calculated by the information calculation unit 12 to the operation terminal 5.

The normalization unit 14 of the stress estimation device 1 normalizes the value calculated by the information calculation unit 12 by using each reference value specified by the information identification unit 13 of the stress estimation device 1. May be good. Then, the information output unit 15 may output the value normalized by the normalization unit 14.

That is, each index used for estimating the stress state has advantages and disadvantages in terms of accuracy, sensitivity, and the like. Therefore, in the stress estimation device 1 of the present embodiment, the stress state is estimated while canceling the demerits of each index by using an index that combines a plurality of indexes having different merits and demerits.

As a result, the stress estimation device 1 in the present embodiment can perform the stress state of the subject with high accuracy and stability.

[Details of the first embodiment]
Next, the details of the first embodiment will be described. 4 and 5 are flowcharts illustrating the details of the stress estimation process according to the first embodiment. 6 to 18 are views for explaining the details of the stress estimation process according to the first embodiment.

When the stress estimation process 1 (index acquisition unit 11) starts the stress estimation process, the stress estimation device 1 (index acquisition unit 11) calculates a first index calculated using LF and HF as shown in FIG. 4 (S11). Specifically, the index acquisition unit 11 calculates the first index by, for example, dividing LF by HF.

The index acquisition unit 11 calculates the first index by dividing the LF corrected using the following formula (1) by the HF corrected using the following formula (2) in the processing of S11. It may be what you do.

In the above equations (1) and (2), Total Power indicates the total power of the power spectrum included in the time-series data, and VLF (Very Low Frequency) indicates the ultra-low frequency band (0) included in the time-series data. The power spectrum of the frequency band of .0033 to 0.04 Hz) is shown, and ULF (Ultra Low Frequency) shows the power spectrum of the extremely low frequency band (frequency band of ~ 0.0033 Hz) included in the time series data. ..

The index acquisition unit 11 may start the execution of the processing after S11 in response to the time series data measured from the target person being input via the operation terminal 5, for example.

In addition, the index acquisition unit 11 calculates the second index using SDNN and RMSSD (S12). Specifically, the index acquisition unit 11 calculates the second index by, for example, dividing SDNN by RMSSD.

Further, the index acquisition unit 11 calculates a third index indicating the sample entropy, which is the unpredictability (irregularity) of the fluctuation of the time series data (S13). Hereinafter, a specific example of the processing of S13 will be described.

[Specific example of processing of S13]
FIG. 6 is a diagram illustrating a specific example of the process of S13. Specifically, FIG. 6 is a diagram illustrating a specific example of the pseudo time series data DT when it has a predetermined regularity. In the graph shown in FIG. 6, the horizontal axis represents the measurement point i, and the vertical axis represents the measurement data x at each measurement point. Further, in the graph shown in FIG. 6, the number of measurement points N is 50, and the measurement data x ₁ to x _N at each measurement point are "1001", "1002", "1003", "1004", and "1005", respectively. , "1001", "1002", "1003", "1004", "1005", "1001" ... "1005". Hereinafter, a set of _{x N} is also referred to as _{S N} from the measurement data _{x 1.}

_{First, the index acquisition unit 11 uses the predicted correlation pattern length m to generate a pattern P m} (i) of measurement data x corresponding to each of the m points starting from the measurement point i. do. Hereinafter, the description will be made assuming that the length m is 5. Further, hereinafter, _{the number of patterns P m} (i) is 46, which is the number calculated by subtracting 5 which is the value of length m from 50 which is the value of the number of measurement points N and adding 1 to the number calculated. It will be explained as.

Specifically, index acquiring unit 11, for example, "1001", "1002", "1003", the pattern _P 5 (1) including "1004" and "1005", "1002", "1003", " 1004 ", to generate a pattern _P 5 (2) including" 1005 "and" 1001 ". Also, the index acquiring unit 11, for example, "1003", "1004", "1005", the pattern _P 5 including "1001" and "1002" (3), "1004", "1005", "1001" , it generates the pattern _P 5 (4) including "1002" and "1003". Furthermore, the index acquiring unit 11, for example, "1005", "1001", "1002", the pattern _P 5 including "1003" and "1004" (5) "1001", "1002", "1003" , it generates the pattern _P 5 (6) including "1004" and "1005". Similarly, index acquiring unit 11 generates the pattern _P 5 from the pattern _P 5 (7) (46), respectively.

Next, the index acquisition unit 11 uses a distance function such as the Chebyshev distance or the Euclidean distance to determine the correlation of the time-series data DT, so that the measurement data x included in the _{two patterns Pm (i).} The error d of is calculated for all combinations. Then, the index acquisition unit 11 _{calculates the values C i, m} (r), which is the ratio of _{the pattern P m} (i) in which the calculated error d exceeds the permissible value r for each pattern P _{m (i).} After that, the index acquisition unit 11 calculates _{the average value of the calculated values Ci and m} (r) as the value B ^m (r).

Similarly, index acquiring unit 11 generates a pattern P _{m + 1} measurement data x corresponding to each of the m + 1 points starting at the measurement point i _(i), for each pattern P _{m + 1 (i),} was calculated _{The values C i and m + 1} (r), which are the ratios of the _{patterns P m + 1} (i) in which the error d exceeds the permissible value r, are calculated. Then, the index acquisition unit 11 calculates _{the average value of the calculated values Ci and m + 1} (r) as the value ^Am (r).

After that, the index acquisition unit 11 ^{applies the calculated values Am} (r) and value B ^m (r) to the following equation (3) to obtain unpredictability (irregularity) of fluctuations in the time series data. A third index indicating a sample entropy is calculated.

In the above equation (3), SampEn (m, r, N) indicates a third index calculated by using _{the length m, the permissible value r, and the set SN.} In the following description, it is assumed that the permissible value r is 2.

Specifically, for example, the _{first measurement data x included in the pattern P 5} (1) is “1001”, and the _{first measurement data x included in the pattern P 5} (2) is “1002”. Therefore, the index acquisition unit 11 calculates "1" as the absolute value of the difference between the measurement data x included first in _{the pattern P 5} (1) and the pattern P _{5 (2).} Similarly, the index acquisition unit 11 calculates "1" as the absolute value of the difference between the measurement data x included second in _{the pattern P 5} (1) and the pattern P _{5 (2), and the measurement included in the third pattern.} "1" is calculated as the absolute value of the difference of the data x, "1" is calculated as the absolute value of the difference of the measurement data x included in the fourth, and the absolute value of the difference of the measurement data x included in the fifth is calculated. Calculate "4". Then, the index acquisition unit 11 calculates the maximum value "4" of the calculated values as an error d, and determines whether or not the calculated value "4" exceeds the permissible value r.

After that, the index acquisition unit 11 performs the same processing for the relationship between _{the pattern P 5} (1) and the pattern P _{5 (i) other than the pattern P 5 (2), for example, the pattern P 5} ( If the number is "9" on 1) the pattern _P 5 the error d exceeds the allowable value r of (i), it calculates the "9/45" as the value _{C 1, 5} (2).

The index acquisition unit 11, the average is also calculated values _{C i,} 5 (2) for each of the calculated values _{C i,} 5 (2) of the pattern _P 5 (1) except for the pattern _P 5 (i) By doing so, the value B ⁵ (2) is calculated.

Similarly, the index acquisition unit 11 ^{calculates the value A 5} (2) _{by averaging the values Ci and 6} (2) calculated for each of _{the patterns P 6 (i).}

Further, the index acquisition unit 11 calculates ^{SampEn (5, 2, 50) by applying the calculated values B 5} (2) and A 5 (2) to the equation (3).

As a result, the index acquisition unit 11 can calculate a third index that takes a small value when the periodicity of the time series data is high and a large value when the periodicity is low.

Returning to FIG. 4, the information calculation unit 12 uses the first index calculated in the process of S11, the second index calculated in the process of S12, and the third index calculated in the process of S13, and thereby the target person. A value indicating the state of stress is calculated (S14).

Specifically, the information calculation unit 12 calculates the value by, for example, following the following equation (4).

In the above equation (4), "LF / HF" corresponds to the first index, "SDNN / RMSSD" corresponds to the second index, and "SampEn" corresponds to the third index. Therefore, in this case, the information calculation unit 12 uses the value calculated by multiplying the first index calculated in the process of S11 and the second index calculated in the process of S12 as the third index calculated in the process of S13. By dividing, a value indicating the stress state of the subject is calculated.

That is, the first index has a merit of high sensitivity and a demerit of low stability. On the other hand, the second index and the third index have a merit of high stability and a demerit of low sensitivity. In addition, the stability of each of the first index and the third index may decrease depending on the shortness of the time series data.

Therefore, the stress estimation device 1 in the present embodiment cancels the demerits of each index by estimating the stress state of the subject using an index that combines the first index, the second index, and the third index. Will be possible. Therefore, the stress estimation device 1 can determine the stress state of the subject with high accuracy and stability.

Subsequently, as shown in FIG. 5, the information specifying unit 13 specifies the maximum value among the values indicating the past stress state of the subject (S21). Specifically, the information identification unit 13 refers to the storage area 110 that has accumulated the values calculated in the stress estimation process executed in the past, and is calculated in the past for the same target person as the stress estimation process executed this time. Specify the maximum value among the values. The value indicating the past stress state of the subject may be, for example, a value indicating the stress state of the subject for the past 30 days.

In addition, the information specifying unit 13 specifies the minimum value among the values indicating the past stress state of the subject (S22). Specifically, the information identification unit 13 refers to the storage area 110 that has accumulated the values calculated in the stress estimation process executed in the past, and is calculated in the past for the same target person as the stress estimation process executed this time. Identify the minimum of the values.

The information specifying unit 13 specifies the maximum value and the minimum value from the values indicating the past stress state, which are calculated using the time series data measured in the same time zone. It may be there.

Specifically, for example, when the value calculated in the process of S14 is a value calculated using the time series data measured from the subject at 9:00 am, the information specifying unit 13 is a value indicating a past stress state. Of these, the maximum value and the minimum value may be specified from the values calculated using the time series data measured from the subject around 9:00 am.

After that, the normalization unit 14 normalizes the value calculated in the process of S14 by using the maximum value specified in the process of S21 and the minimum value specified in the process of S22 (S23).

As a result, the stress estimation device 1 can compare the stress states between different subjects by keeping the value indicating the stress state of each subject between 0 and 1. That is, by using a common standard (value normalized by the processing of S23), the stress estimation device 1 makes it possible to compare stress states even among subjects having individual differences.

Specifically, the normalization unit 14 normalizes the value by, for example, following the following equation (5).

In the above equation (5), "SI" corresponds to the value calculated in the processing of S14, "SI _max " corresponds to the maximum value calculated in the processing of S21, and "SI _min " is calculated in the processing of S22. and corresponds to a minimum value, corresponding to a value obtained by normalizing the value calculated by the processing of the "S _n" is S14. Therefore, in this case, the normalization unit 14 subtracts the minimum value calculated in the processing of S22 from the value calculated in the processing of S14, and subtracts the minimum value calculated in the processing of S22 from the maximum value calculated in the processing of S21. Is divided by the subtracted value to normalize the value calculated in the process of S14.

In the process of S21, the information specifying unit 13 may specify the maximum value among the values indicating the past stress state of the target person and excluding the outliers. Further, the information specifying unit 13 may specify the minimum value among the values indicating the past stress state of the target person and excluding the outliers in the processing of S22.

Specifically, for example, the SI of the subject for the past 30 days is "0.14", "0.1", "1.48", "0.3", "0.3", "0.54". , "1.65", "0.1", "0.26", "0.06", "0.16", "0.47", "1.04", "0.56", "0.23", "1.62", "0.2", "1.91", "0.46", "1.16", "1.15", "0.69", "0" When it is ".32", "0.27", "0.41", "0.14", "0.3", "0.19", "0.16" and "0.18" The information identification unit 13 calculates "0.7775", "0.96", and "0.7175" as the first quadrant, the third quadrant, and the quadrant range of these data, respectively. do.

Then, the information specifying unit 13 calculates "1.8537", which is a value calculated by adding the third quartile to the value calculated by multiplying the quartile range by 1.5, as the upper limit threshold value. From the SI of the subject for the past 30 days, the maximum value "1.65" among the values smaller than the calculated upper limit threshold value is specified.

Further, the information specifying unit 13 calculates the value calculated by multiplying the interquartile range by 1.5 and subtracting the value calculated by subtracting it from the first quartile, using "1.8537" as the lower limit threshold value. From the SI of the subject for the past 30 days, the minimum value "0.06" among the values larger than the calculated lower limit threshold value is specified.

As a result, the stress estimation device 1 can more accurately determine the stress state of the subject.

Then, the information output unit 15 outputs, for example, the value normalized by the process of S23 to the operation terminal 5 (S24). Hereinafter, an output example in the processing of S24 will be described.

[Output example in S24 processing]
7 to 18 are diagrams for explaining an output example in the process of S24.

[Output example when using a scale]
First, an output example when using a scale will be described. FIG. 7 is an output example when a scale is used.

In the output example OP11a shown in FIG. 7A, a scale indicating that the degree of “stress” and the degree of “relaxation” are each “0.5” is displayed. Then, in the output example OP11a shown in FIG. 7A, the character string "in a balanced state" is displayed.

Further, in the output example OP11b shown in FIG. 7B, a scale indicating that the degree of “stress” is “0.9” and the degree of “relaxation” is “0.1” is displayed. .. Then, in the output example OP11b shown in FIG. 7B, the character string "It is in a very dangerous state" is displayed.

[Output example when using a tachometer]
Next, an output example when a tachometer is used will be described. FIG. 8 is an output example when a tachometer is used.

In the output example OP12 shown in FIG. 8, the needle indicates “0.4” in the region corresponding to relaxation. Then, in the output example OP12 shown in FIG. 8, the character string "a little relaxed" is displayed.

[Output example when using colors and characters]
Next, an output example when colors and characters are used will be described. FIG. 9 is an output example when colors and characters are used.

In the output example OP13 (A) shown in FIG. 9, the character strings “high stress” and “average value: 0.5” are displayed together with the character string “0.9”. A color (for example, red) corresponding to a high stress state is used inside the circle displaying the character string.

Further, in the output example OP13 (B) shown in FIG. 9, the character strings “balance” and “average value: 0.5” are displayed together with the character string “0.3”. A color corresponding to the balanced state (for example, orange) is used inside the circle displaying the character string.

Further, in the output example OP13 (C) shown in FIG. 9, the character strings “relax” and “average value: 0.5” are displayed together with the character string “0.1”. A color (for example, blue) corresponding to the relaxed state is used inside the circle displaying the character string.

[Output example when using bars and stars]
Next, an output example when a bar and a star mark are used will be described. 10 and 11 are output examples when a bar and a star mark are used.

In the output example OP14 shown in FIG. 10, a horizontal bar corresponding to the stress state is displayed on the left side and the horizontal bar corresponds to the relaxed state on the right side. Then, in the output example OP14 shown in FIG. 10, a star indicating the stress state of the subject is displayed at the position corresponding to the relaxed state in the horizontal bar (the position on the right side in the horizontal bar).

Further, in the output example OP15 shown in FIG. 11, four horizontal bars are displayed on the left side corresponding to the stress state and on the right side corresponding to the relaxed state. The uppermost horizontal bar indicates, for example, the stress state estimated using the time series data measured at 9:00 am on September 21, 2018, and the position corresponding to the relaxed state in the horizontal bar ( A star mark is displayed at the position on the right side of the horizontal bar). The second horizontal bar from the top shows, for example, the stress state estimated using the time series data measured at 10 am on September 21, 2018, and corresponds to the stress state in the horizontal bar. An asterisk is displayed at the position (the position on the left side of the horizontal bar).

That is, the stress estimation device 1 may perform the processes S11 to S24 a plurality of times every day. Then, in this case, the stress estimation device 1 may output each estimation result separately in the process of S25.

As a result, the stress estimation device 1 can output a more detailed determination result regarding the stress state of the subject.

[Output example when using chart representation]
Next, an output example when the chart representation is used will be described. 12 and 13 are output examples when the chart representation is used.

In the output example OP16 shown in FIG. 12, seven vertical bars corresponding to the stress state on the upper side and the relaxed state on the lower side are displayed. Then, in the output example OP16 shown in FIG. 12, for example, the leftmost vertical bar corresponds to "0.9" indicating a high stress state, indicating that the subject is in a high stress state. There is. Further, in the output example OP16 shown in FIG. 12, for example, the third vertical bar from the left corresponds to “0.1” indicating the relaxed state, indicating that the subject is in the relaxed state. In the chart in the output example OP16 shown in FIG. 12, the horizontal axis corresponds to time, and each of the vertical bars corresponds to the stress state estimated using the time series data measured at different times. ..

Further, in the output example OP17 shown in FIG. 13, the upper chart corresponds to the stress state estimated using the time series data measured in the morning time zone (3 o'clock to 11 o'clock), and the middle chart corresponds to the daytime. Corresponding to the stress state estimated using the time series data measured in the time zone (11:00 to 19:00), when the lower chart is measured in the night time zone (19:00 to 3:00 the next day) It corresponds to the stress state estimated using the series data. Then, in the chart of each stage in the output example OP17 shown in FIG. 13, a plurality of vertical bars corresponding to the stress state on the upper side and the relaxed state on the lower side are displayed. In the chart in the output example OP17 shown in FIG. 13, the horizontal axis corresponds to the date, and for example, the leftmost vertical bar in each column is the time series data measured in each time zone on the 16th (Tuesday). The stress state estimated using is shown, and the second vertical bar from the left in each stage shows the stress state estimated using the time series data measured in each time zone on Wednesday, 17th. ing.

[Output example when using heart rate]
Next, an output example when the heart rate is used will be described. FIG. 14 is an output example when the heart rate is used.

In the output example OP18 shown in FIG. 14, the horizontal axis represents the value of the stress state of the subject, and the vertical axis represents the heart rate of the subject. Then, in the output example OP18 shown in FIG. 14, in addition to the points corresponding to the stress state estimated this time, for example, the points corresponding to the stress state estimated at the time of executing the stress estimation processing for the past seven times are displayed. There is. Further, in the output example OP18 shown in FIG. 14, a locus connecting each of the past points and the latest point is displayed in chronological order.

[Output example when using icon expression]
Next, an output example when the icon expression is used will be described. FIG. 15 is an output example when the icon representation is used.

In the output example OP19 shown in FIG. 15, the upper side corresponds to the stress state and the lower side corresponds to the relaxed state, as in the case described with reference to FIG. Then, in the output example OP19 shown in FIG. 15, face icons corresponding to the stress state of the subject are displayed at each position. Specifically, in the output example OP19 shown in FIG. 15, a face corresponding to "relaxation", a face corresponding to "slightly relaxing", a face corresponding to "balance", and "high stress (high stress state)" are supported. Each of the face to be displayed and the face corresponding to "very high stress (very high stress)" is displayed.

[Output example when comparing with the stress state of others]
Next, an output example when comparing with the stress state of another person (average of the tissue) will be described. 16 to 18 are output examples when comparing with the stress state of another person (average of the tissue).

In the output example OP21 shown in FIG. 16, a horizontal bar corresponding to the stress state is displayed on the left side and the horizontal bar corresponding to the relaxed state is displayed on the right side, as in the case described with reference to FIG. Then, in the output example OP21 shown in FIG. 16, a star indicating the stress state of the subject is displayed at the position corresponding to the high stress state in the horizontal bar (the position on the left side in the horizontal bar), and the horizontal bar shows the stress state. At the position corresponding to the relaxed state (the position on the right side of the horizontal bar), an asterisk indicating the average value of the stress state values in the tissue including the subject (hereinafter, also referred to as the average stress state of the tissue) is displayed. ..

Further, in the output example OP22 shown in FIG. 17, a circle including a value “0.8” indicating the stress state of the subject and a circle including a value “0.6” indicating the average stress state of the tissue. And are displayed respectively. In the output example OP22 shown in FIG. 17, the circle including the stress state value of the subject is larger than the circle including the average stress state value of the tissue.

Further, in the output example OP23 shown in FIG. 18, vertical bars corresponding to the stress state on the upper side and the relaxed state on the lower side are displayed as in the case described with reference to FIG. Then, in the output example OP23 shown in FIG. 18, the vertical bar on the left indicates the stress state of the subject, and corresponds to “0.8” indicating the high stress state. Further, in the output example OP23 shown in FIG. 18, the vertical bar on the right indicates the average stress state of the tissue, and corresponds to “0.3” indicating the relaxed state.

That is, as shown in FIGS. 7 to 18, the information output unit 15 generates and outputs an output screen according to the intended use and the like from the values normalized by the processing of S24.

As a result, the stress estimation device 1 makes it possible for each subject to intuitively understand the stress state. Therefore, the stress estimation device 1 can reduce the burden on each subject due to the confirmation of the stress state.

1: Stress estimation device 5: Operation terminal 101: CPU
102: Memory 103: Network interface 104: Storage medium 105: Bus

Japanese Unexamined Patent Publication No. 2005-218595 Japanese Unexamined Patent Publication No. 2011-167323 Japanese Unexamined Patent Publication No. 2016-142553 JP-A-2017-176833

Claims (8)

Using LF (Low frequency), which is a low frequency band power spectrum included in the time series data of the heartbeat interval of the subject, and HF (High frequency), which is a high frequency band power spectrum included in the time series data. Calculate the first index
It is the square root of the mean square of the difference between SDNN (Standard deviation of NN intervals), which is the standard deviation of the heartbeat interval included in the time-series data, and the difference between two heartbeat intervals continuously included in the time-series data. The second index was calculated using RMSD (Root mean square of success deviation).
By using the calculated first index and the second index, a value indicating the stress state of the subject is calculated.
Output the calculated value,
A stress estimation program characterized by having a computer perform processing. In claim 1,
The first index is a value calculated by dividing the LF by the HF.
The second index is a value calculated by dividing the SDNN by the RMSSD.
A stress estimation program characterized by this. In claim 1,
In the process of calculating the value, the value is calculated by multiplying the first index and the second index.
A stress estimation program characterized by this. In claim 3, further
A third index showing the sample entropy representing the unpredictability of the fluctuation of the time series data is calculated.
Let the computer do the work
In the process of calculating the value, the value is calculated by multiplying the first index and the second index and dividing by the third index.
A stress estimation program characterized by this. In claim 1, further
The maximum value and the minimum value are specified from the above-mentioned values calculated in the past for the target person, and the maximum value and the minimum value are specified.
Normalization of the value is performed by using the specified maximum value and the minimum value.
Let the computer do the work
In the output process,
Output the normalized value,
A stress estimation program characterized by this. In claim 5,
In the process of performing the normalization, the value obtained by subtracting the minimum value from the value calculated in the calculated process is divided by the value obtained by subtracting the minimum value from the maximum value, thereby calculating in the calculated process. Normalize the above values
A stress estimation program characterized by this. The first index was calculated using LF, which is a low frequency band power spectrum included in the time series data of the heartbeat interval of the subject, and HF, which is a high frequency band power spectrum included in the time series data. A second using SDNN, which is the standard deviation of the heartbeat interval included in the time series data, and RMSSD, which is the square root of the square of the square of the difference between the two heartbeat intervals included in the time series data in succession. The index acquisition department that calculates the index and
By using the calculated first index and the second index, an information calculation unit that calculates a value indicating a stress state of the subject, and an information calculation unit.
It has an information output unit that outputs the calculated value.
A stress estimator characterized by this. The first index is calculated using LF, which is a low frequency band power spectrum included in the time series data of the heartbeat interval of the subject, and HF, which is a high frequency band power spectrum included in the time series data.
A second using SDNN, which is the standard deviation of the heartbeat interval included in the time series data, and RMSSD, which is the square root of the square root of the square of the difference between the two heartbeat intervals included in the time series data in succession. Calculate the index,
By using the calculated first index and the second index, a value indicating the stress state of the subject is calculated.
Output the calculated value,
A stress estimation method characterized by having a computer perform processing.

Images